{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.870583564906709,
  "eval_steps": 3000,
  "global_step": 39000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.742559337615967,
      "epoch": 0.0004962286621675268,
      "grad_norm": 5.21875,
      "learning_rate": 2e-06,
      "loss": 10.7925,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11448.0,
      "step": 5
    },
    {
      "entropy": 10.742586231231689,
      "epoch": 0.0009924573243350536,
      "grad_norm": 5.0,
      "learning_rate": 4.5e-06,
      "loss": 10.7825,
      "mean_token_accuracy": 0.0,
      "num_tokens": 22646.0,
      "step": 10
    },
    {
      "entropy": 10.742585563659668,
      "epoch": 0.0014886859865025804,
      "grad_norm": 4.75,
      "learning_rate": 7e-06,
      "loss": 10.7537,
      "mean_token_accuracy": 0.00020905662677250803,
      "num_tokens": 34011.0,
      "step": 15
    },
    {
      "entropy": 10.742612171173096,
      "epoch": 0.001984914648670107,
      "grad_norm": 4.6875,
      "learning_rate": 9.5e-06,
      "loss": 10.6968,
      "mean_token_accuracy": 0.0005319510004483163,
      "num_tokens": 44566.0,
      "step": 20
    },
    {
      "entropy": 10.742612838745117,
      "epoch": 0.002481143310837634,
      "grad_norm": 4.3125,
      "learning_rate": 1.2e-05,
      "loss": 10.6222,
      "mean_token_accuracy": 0.004951164737576619,
      "num_tokens": 56274.0,
      "step": 25
    },
    {
      "entropy": 10.742339897155762,
      "epoch": 0.0029773719730051607,
      "grad_norm": 3.875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.4853,
      "mean_token_accuracy": 0.024180484376847743,
      "num_tokens": 66584.0,
      "step": 30
    },
    {
      "entropy": 10.741417789459229,
      "epoch": 0.0034736006351726877,
      "grad_norm": 3.0625,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3696,
      "mean_token_accuracy": 0.035252702049911024,
      "num_tokens": 76948.0,
      "step": 35
    },
    {
      "entropy": 10.738713264465332,
      "epoch": 0.003969829297340214,
      "grad_norm": 2.296875,
      "learning_rate": 1.95e-05,
      "loss": 10.2243,
      "mean_token_accuracy": 0.03839292954653502,
      "num_tokens": 87253.0,
      "step": 40
    },
    {
      "entropy": 10.734406661987304,
      "epoch": 0.004466057959507741,
      "grad_norm": 2.3125,
      "learning_rate": 2.2e-05,
      "loss": 10.1158,
      "mean_token_accuracy": 0.04111976623535156,
      "num_tokens": 98273.0,
      "step": 45
    },
    {
      "entropy": 10.730075454711914,
      "epoch": 0.004962286621675268,
      "grad_norm": 2.09375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 10.0069,
      "mean_token_accuracy": 0.04187291283160448,
      "num_tokens": 107788.0,
      "step": 50
    },
    {
      "entropy": 10.727989768981933,
      "epoch": 0.0054585152838427945,
      "grad_norm": 1.984375,
      "learning_rate": 2.7e-05,
      "loss": 9.9971,
      "mean_token_accuracy": 0.03726662136614323,
      "num_tokens": 118632.0,
      "step": 55
    },
    {
      "entropy": 10.727395725250243,
      "epoch": 0.0059547439460103215,
      "grad_norm": 1.984375,
      "learning_rate": 2.95e-05,
      "loss": 9.8873,
      "mean_token_accuracy": 0.04101524651050568,
      "num_tokens": 128649.0,
      "step": 60
    },
    {
      "entropy": 10.725363159179688,
      "epoch": 0.0064509726081778485,
      "grad_norm": 1.6953125,
      "learning_rate": 3.2e-05,
      "loss": 9.8548,
      "mean_token_accuracy": 0.04273284543305635,
      "num_tokens": 138958.0,
      "step": 65
    },
    {
      "entropy": 10.721133422851562,
      "epoch": 0.0069472012703453755,
      "grad_norm": 1.8203125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.8217,
      "mean_token_accuracy": 0.04243846274912357,
      "num_tokens": 151445.0,
      "step": 70
    },
    {
      "entropy": 10.715934658050537,
      "epoch": 0.007443429932512902,
      "grad_norm": 1.6953125,
      "learning_rate": 3.7e-05,
      "loss": 9.7492,
      "mean_token_accuracy": 0.043674266710877416,
      "num_tokens": 162239.0,
      "step": 75
    },
    {
      "entropy": 10.70990047454834,
      "epoch": 0.007939658594680429,
      "grad_norm": 1.8203125,
      "learning_rate": 3.95e-05,
      "loss": 9.7155,
      "mean_token_accuracy": 0.03882305920124054,
      "num_tokens": 173261.0,
      "step": 80
    },
    {
      "entropy": 10.70549554824829,
      "epoch": 0.008435887256847956,
      "grad_norm": 1.7734375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.6222,
      "mean_token_accuracy": 0.037595166079699996,
      "num_tokens": 183857.0,
      "step": 85
    },
    {
      "entropy": 10.701297950744628,
      "epoch": 0.008932115919015483,
      "grad_norm": 1.671875,
      "learning_rate": 4.45e-05,
      "loss": 9.646,
      "mean_token_accuracy": 0.03864660933613777,
      "num_tokens": 196635.0,
      "step": 90
    },
    {
      "entropy": 10.692024230957031,
      "epoch": 0.00942834458118301,
      "grad_norm": 1.75,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.4935,
      "mean_token_accuracy": 0.04372961949557066,
      "num_tokens": 208700.0,
      "step": 95
    },
    {
      "entropy": 10.67728853225708,
      "epoch": 0.009924573243350537,
      "grad_norm": 1.828125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.3747,
      "mean_token_accuracy": 0.04096841104328632,
      "num_tokens": 218915.0,
      "step": 100
    },
    {
      "entropy": 10.659735870361327,
      "epoch": 0.010420801905518062,
      "grad_norm": 1.765625,
      "learning_rate": 5.2e-05,
      "loss": 9.2746,
      "mean_token_accuracy": 0.04906293600797653,
      "num_tokens": 230211.0,
      "step": 105
    },
    {
      "entropy": 10.6349609375,
      "epoch": 0.010917030567685589,
      "grad_norm": 1.8046875,
      "learning_rate": 5.45e-05,
      "loss": 9.1754,
      "mean_token_accuracy": 0.05304342210292816,
      "num_tokens": 241174.0,
      "step": 110
    },
    {
      "entropy": 10.604046154022218,
      "epoch": 0.011413259229853116,
      "grad_norm": 1.65625,
      "learning_rate": 5.7e-05,
      "loss": 9.1068,
      "mean_token_accuracy": 0.04802950657904148,
      "num_tokens": 251708.0,
      "step": 115
    },
    {
      "entropy": 10.571563243865967,
      "epoch": 0.011909487892020643,
      "grad_norm": 1.5234375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 9.1035,
      "mean_token_accuracy": 0.04602174609899521,
      "num_tokens": 263287.0,
      "step": 120
    },
    {
      "entropy": 10.524694061279297,
      "epoch": 0.01240571655418817,
      "grad_norm": 1.46875,
      "learning_rate": 6.2e-05,
      "loss": 8.9308,
      "mean_token_accuracy": 0.04466274194419384,
      "num_tokens": 275179.0,
      "step": 125
    },
    {
      "entropy": 10.459578609466552,
      "epoch": 0.012901945216355697,
      "grad_norm": 1.671875,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.731,
      "mean_token_accuracy": 0.049411849677562715,
      "num_tokens": 287223.0,
      "step": 130
    },
    {
      "entropy": 10.371681499481202,
      "epoch": 0.013398173878523224,
      "grad_norm": 1.5078125,
      "learning_rate": 6.7e-05,
      "loss": 8.6072,
      "mean_token_accuracy": 0.04893956743180752,
      "num_tokens": 297868.0,
      "step": 135
    },
    {
      "entropy": 10.280974102020263,
      "epoch": 0.013894402540690751,
      "grad_norm": 1.3984375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.5167,
      "mean_token_accuracy": 0.05017189048230648,
      "num_tokens": 309702.0,
      "step": 140
    },
    {
      "entropy": 10.183215141296387,
      "epoch": 0.014390631202858276,
      "grad_norm": 1.3515625,
      "learning_rate": 7.2e-05,
      "loss": 8.483,
      "mean_token_accuracy": 0.04685804210603237,
      "num_tokens": 321613.0,
      "step": 145
    },
    {
      "entropy": 10.04955472946167,
      "epoch": 0.014886859865025803,
      "grad_norm": 1.2890625,
      "learning_rate": 7.45e-05,
      "loss": 8.2837,
      "mean_token_accuracy": 0.053469978272914886,
      "num_tokens": 332668.0,
      "step": 150
    },
    {
      "entropy": 9.918639755249023,
      "epoch": 0.01538308852719333,
      "grad_norm": 1.203125,
      "learning_rate": 7.7e-05,
      "loss": 8.2455,
      "mean_token_accuracy": 0.054456012323498726,
      "num_tokens": 342579.0,
      "step": 155
    },
    {
      "entropy": 9.765514373779297,
      "epoch": 0.015879317189360857,
      "grad_norm": 1.359375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.0663,
      "mean_token_accuracy": 0.054721662402153017,
      "num_tokens": 352271.0,
      "step": 160
    },
    {
      "entropy": 9.54785737991333,
      "epoch": 0.016375545851528384,
      "grad_norm": 1.1328125,
      "learning_rate": 8.2e-05,
      "loss": 8.0234,
      "mean_token_accuracy": 0.0642885323613882,
      "num_tokens": 364033.0,
      "step": 165
    },
    {
      "entropy": 9.348752498626709,
      "epoch": 0.01687177451369591,
      "grad_norm": 1.1015625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.9279,
      "mean_token_accuracy": 0.06007566004991531,
      "num_tokens": 373788.0,
      "step": 170
    },
    {
      "entropy": 9.17256269454956,
      "epoch": 0.01736800317586344,
      "grad_norm": 1.03125,
      "learning_rate": 8.7e-05,
      "loss": 7.8747,
      "mean_token_accuracy": 0.06030374839901924,
      "num_tokens": 384883.0,
      "step": 175
    },
    {
      "entropy": 8.98403024673462,
      "epoch": 0.017864231838030965,
      "grad_norm": 1.0234375,
      "learning_rate": 8.95e-05,
      "loss": 7.7558,
      "mean_token_accuracy": 0.06696737073361873,
      "num_tokens": 396044.0,
      "step": 180
    },
    {
      "entropy": 8.754393672943115,
      "epoch": 0.018360460500198492,
      "grad_norm": 0.9765625,
      "learning_rate": 9.2e-05,
      "loss": 7.8178,
      "mean_token_accuracy": 0.06663747020065784,
      "num_tokens": 406995.0,
      "step": 185
    },
    {
      "entropy": 8.613575267791749,
      "epoch": 0.01885668916236602,
      "grad_norm": 0.8671875,
      "learning_rate": 9.45e-05,
      "loss": 7.6515,
      "mean_token_accuracy": 0.06824805773794651,
      "num_tokens": 417238.0,
      "step": 190
    },
    {
      "entropy": 8.565249347686768,
      "epoch": 0.019352917824533546,
      "grad_norm": 0.96484375,
      "learning_rate": 9.7e-05,
      "loss": 7.7144,
      "mean_token_accuracy": 0.06267430260777473,
      "num_tokens": 427263.0,
      "step": 195
    },
    {
      "entropy": 8.521870040893555,
      "epoch": 0.019849146486701073,
      "grad_norm": 0.98046875,
      "learning_rate": 9.95e-05,
      "loss": 7.6025,
      "mean_token_accuracy": 0.06824213974177837,
      "num_tokens": 438357.0,
      "step": 200
    },
    {
      "entropy": 8.434043884277344,
      "epoch": 0.0203453751488686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000102,
      "loss": 7.6098,
      "mean_token_accuracy": 0.06504783928394317,
      "num_tokens": 448738.0,
      "step": 205
    },
    {
      "entropy": 8.3219557762146,
      "epoch": 0.020841603811036124,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.6115,
      "mean_token_accuracy": 0.07722424305975437,
      "num_tokens": 457873.0,
      "step": 210
    },
    {
      "entropy": 8.32616958618164,
      "epoch": 0.02133783247320365,
      "grad_norm": 0.8125,
      "learning_rate": 0.000107,
      "loss": 7.5469,
      "mean_token_accuracy": 0.0731696669012308,
      "num_tokens": 468640.0,
      "step": 215
    },
    {
      "entropy": 8.25720796585083,
      "epoch": 0.021834061135371178,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001095,
      "loss": 7.5443,
      "mean_token_accuracy": 0.07317286394536496,
      "num_tokens": 479922.0,
      "step": 220
    },
    {
      "entropy": 8.193887042999268,
      "epoch": 0.022330289797538705,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000112,
      "loss": 7.5043,
      "mean_token_accuracy": 0.06673288382589818,
      "num_tokens": 491056.0,
      "step": 225
    },
    {
      "entropy": 8.124149227142334,
      "epoch": 0.022826518459706232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001145,
      "loss": 7.5857,
      "mean_token_accuracy": 0.07121467962861061,
      "num_tokens": 504365.0,
      "step": 230
    },
    {
      "entropy": 8.186816024780274,
      "epoch": 0.02332274712187376,
      "grad_norm": 1.28125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.4841,
      "mean_token_accuracy": 0.06768220961093903,
      "num_tokens": 514347.0,
      "step": 235
    },
    {
      "entropy": 8.151187849044799,
      "epoch": 0.023818975784041286,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.4982,
      "mean_token_accuracy": 0.07583453841507434,
      "num_tokens": 525912.0,
      "step": 240
    },
    {
      "entropy": 8.13270378112793,
      "epoch": 0.024315204446208813,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000122,
      "loss": 7.3789,
      "mean_token_accuracy": 0.0720100063830614,
      "num_tokens": 535566.0,
      "step": 245
    },
    {
      "entropy": 8.082850551605224,
      "epoch": 0.02481143310837634,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0001245,
      "loss": 7.4273,
      "mean_token_accuracy": 0.07373461574316025,
      "num_tokens": 546474.0,
      "step": 250
    },
    {
      "entropy": 8.037993764877319,
      "epoch": 0.025307661770543867,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000127,
      "loss": 7.512,
      "mean_token_accuracy": 0.06475581601262093,
      "num_tokens": 558132.0,
      "step": 255
    },
    {
      "entropy": 8.116750812530517,
      "epoch": 0.025803890432711394,
      "grad_norm": 1.109375,
      "learning_rate": 0.0001295,
      "loss": 7.4896,
      "mean_token_accuracy": 0.07219315581023693,
      "num_tokens": 568309.0,
      "step": 260
    },
    {
      "entropy": 7.967530822753906,
      "epoch": 0.02630011909487892,
      "grad_norm": 1.015625,
      "learning_rate": 0.000132,
      "loss": 7.3878,
      "mean_token_accuracy": 0.07423482574522496,
      "num_tokens": 579701.0,
      "step": 265
    },
    {
      "entropy": 8.024831199645996,
      "epoch": 0.026796347757046448,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.3601,
      "mean_token_accuracy": 0.0777493104338646,
      "num_tokens": 590832.0,
      "step": 270
    },
    {
      "entropy": 7.985195302963257,
      "epoch": 0.027292576419213975,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3025,
      "mean_token_accuracy": 0.07602775730192661,
      "num_tokens": 601094.0,
      "step": 275
    },
    {
      "entropy": 7.902724742889404,
      "epoch": 0.027788805081381502,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001395,
      "loss": 7.3129,
      "mean_token_accuracy": 0.08000940680503846,
      "num_tokens": 612027.0,
      "step": 280
    },
    {
      "entropy": 7.878976964950562,
      "epoch": 0.02828503374354903,
      "grad_norm": 0.875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.3466,
      "mean_token_accuracy": 0.07674156315624714,
      "num_tokens": 623698.0,
      "step": 285
    },
    {
      "entropy": 7.936695146560669,
      "epoch": 0.028781262405716553,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0001445,
      "loss": 7.359,
      "mean_token_accuracy": 0.07654249593615532,
      "num_tokens": 635631.0,
      "step": 290
    },
    {
      "entropy": 7.894163751602173,
      "epoch": 0.02927749106788408,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000147,
      "loss": 7.3397,
      "mean_token_accuracy": 0.0750736203044653,
      "num_tokens": 647663.0,
      "step": 295
    },
    {
      "entropy": 7.829724407196045,
      "epoch": 0.029773719730051607,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0001495,
      "loss": 7.3044,
      "mean_token_accuracy": 0.07420375496149063,
      "num_tokens": 659352.0,
      "step": 300
    },
    {
      "entropy": 7.924322032928467,
      "epoch": 0.030269948392219134,
      "grad_norm": 0.921875,
      "learning_rate": 0.000152,
      "loss": 7.2742,
      "mean_token_accuracy": 0.08252886533737183,
      "num_tokens": 669379.0,
      "step": 305
    },
    {
      "entropy": 7.784340810775757,
      "epoch": 0.03076617705438666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.2618,
      "mean_token_accuracy": 0.0796934574842453,
      "num_tokens": 680144.0,
      "step": 310
    },
    {
      "entropy": 7.935956048965454,
      "epoch": 0.03126240571655419,
      "grad_norm": 1.015625,
      "learning_rate": 0.000157,
      "loss": 7.2152,
      "mean_token_accuracy": 0.07189621031284332,
      "num_tokens": 691165.0,
      "step": 315
    },
    {
      "entropy": 7.792111110687256,
      "epoch": 0.031758634378721715,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001595,
      "loss": 7.2931,
      "mean_token_accuracy": 0.07880425304174424,
      "num_tokens": 703767.0,
      "step": 320
    },
    {
      "entropy": 7.7711742401123045,
      "epoch": 0.032254863040889245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000162,
      "loss": 7.2808,
      "mean_token_accuracy": 0.08353963568806648,
      "num_tokens": 714298.0,
      "step": 325
    },
    {
      "entropy": 7.765182256698608,
      "epoch": 0.03275109170305677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.2558,
      "mean_token_accuracy": 0.08778629601001739,
      "num_tokens": 724135.0,
      "step": 330
    },
    {
      "entropy": 7.761576843261719,
      "epoch": 0.03324732036522429,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.3084,
      "mean_token_accuracy": 0.07605698890984058,
      "num_tokens": 735050.0,
      "step": 335
    },
    {
      "entropy": 7.796260786056519,
      "epoch": 0.03374354902739182,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.2195,
      "mean_token_accuracy": 0.0794803261756897,
      "num_tokens": 746073.0,
      "step": 340
    },
    {
      "entropy": 7.7471351146698,
      "epoch": 0.034239777689559346,
      "grad_norm": 0.921875,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.2147,
      "mean_token_accuracy": 0.07722761482000351,
      "num_tokens": 758998.0,
      "step": 345
    },
    {
      "entropy": 7.7211285591125485,
      "epoch": 0.03473600635172688,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.1516,
      "mean_token_accuracy": 0.07914387956261634,
      "num_tokens": 771608.0,
      "step": 350
    },
    {
      "entropy": 7.7531922340393065,
      "epoch": 0.0352322350138944,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000177,
      "loss": 7.2202,
      "mean_token_accuracy": 0.07867058292031288,
      "num_tokens": 782394.0,
      "step": 355
    },
    {
      "entropy": 7.697312593460083,
      "epoch": 0.03572846367606193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0001795,
      "loss": 7.122,
      "mean_token_accuracy": 0.08720502257347107,
      "num_tokens": 792389.0,
      "step": 360
    },
    {
      "entropy": 7.740397548675537,
      "epoch": 0.036224692338229454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000182,
      "loss": 7.1616,
      "mean_token_accuracy": 0.08682659715414047,
      "num_tokens": 803219.0,
      "step": 365
    },
    {
      "entropy": 7.670851850509644,
      "epoch": 0.036720921000396985,
      "grad_norm": 0.984375,
      "learning_rate": 0.0001845,
      "loss": 7.2297,
      "mean_token_accuracy": 0.08005850054323674,
      "num_tokens": 814249.0,
      "step": 370
    },
    {
      "entropy": 7.705906867980957,
      "epoch": 0.03721714966256451,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000187,
      "loss": 7.1186,
      "mean_token_accuracy": 0.08939927890896797,
      "num_tokens": 825138.0,
      "step": 375
    },
    {
      "entropy": 7.6558843612670895,
      "epoch": 0.03771337832473204,
      "grad_norm": 1.0,
      "learning_rate": 0.0001895,
      "loss": 7.1626,
      "mean_token_accuracy": 0.08431188724935054,
      "num_tokens": 836024.0,
      "step": 380
    },
    {
      "entropy": 7.675263166427612,
      "epoch": 0.03820960698689956,
      "grad_norm": 0.90625,
      "learning_rate": 0.000192,
      "loss": 7.1514,
      "mean_token_accuracy": 0.08133338317275048,
      "num_tokens": 847765.0,
      "step": 385
    },
    {
      "entropy": 7.624440860748291,
      "epoch": 0.03870583564906709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001945,
      "loss": 7.1454,
      "mean_token_accuracy": 0.08270616680383683,
      "num_tokens": 859980.0,
      "step": 390
    },
    {
      "entropy": 7.753679609298706,
      "epoch": 0.039202064311234616,
      "grad_norm": 1.09375,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.1381,
      "mean_token_accuracy": 0.08271029070019723,
      "num_tokens": 871514.0,
      "step": 395
    },
    {
      "entropy": 7.651297092437744,
      "epoch": 0.03969829297340215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.1347,
      "mean_token_accuracy": 0.08268468752503395,
      "num_tokens": 881984.0,
      "step": 400
    },
    {
      "entropy": 7.617497539520263,
      "epoch": 0.04019452163556967,
      "grad_norm": 1.046875,
      "learning_rate": 0.000202,
      "loss": 7.0907,
      "mean_token_accuracy": 0.0867445409297943,
      "num_tokens": 894198.0,
      "step": 405
    },
    {
      "entropy": 7.598975372314453,
      "epoch": 0.0406907502977372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.1054,
      "mean_token_accuracy": 0.08495021611452103,
      "num_tokens": 905394.0,
      "step": 410
    },
    {
      "entropy": 7.668603801727295,
      "epoch": 0.041186978959904724,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000207,
      "loss": 7.1566,
      "mean_token_accuracy": 0.07967745214700699,
      "num_tokens": 917010.0,
      "step": 415
    },
    {
      "entropy": 7.577554368972779,
      "epoch": 0.04168320762207225,
      "grad_norm": 1.125,
      "learning_rate": 0.0002095,
      "loss": 6.988,
      "mean_token_accuracy": 0.09206651523709297,
      "num_tokens": 926761.0,
      "step": 420
    },
    {
      "entropy": 7.479721593856811,
      "epoch": 0.04217943628423978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000212,
      "loss": 6.9942,
      "mean_token_accuracy": 0.08749158009886741,
      "num_tokens": 938385.0,
      "step": 425
    },
    {
      "entropy": 7.532217884063721,
      "epoch": 0.0426756649464073,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002145,
      "loss": 7.0277,
      "mean_token_accuracy": 0.08903890922665596,
      "num_tokens": 949709.0,
      "step": 430
    },
    {
      "entropy": 7.584686422348023,
      "epoch": 0.04317189360857483,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.0589,
      "mean_token_accuracy": 0.0841300755739212,
      "num_tokens": 960837.0,
      "step": 435
    },
    {
      "entropy": 7.509622049331665,
      "epoch": 0.043668122270742356,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002195,
      "loss": 6.8948,
      "mean_token_accuracy": 0.09134786874055863,
      "num_tokens": 971036.0,
      "step": 440
    },
    {
      "entropy": 7.528602838516235,
      "epoch": 0.044164350932909886,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000222,
      "loss": 7.0396,
      "mean_token_accuracy": 0.08592119216918945,
      "num_tokens": 983057.0,
      "step": 445
    },
    {
      "entropy": 7.493656635284424,
      "epoch": 0.04466057959507741,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002245,
      "loss": 6.9887,
      "mean_token_accuracy": 0.09124278873205185,
      "num_tokens": 994303.0,
      "step": 450
    },
    {
      "entropy": 7.4633971691131595,
      "epoch": 0.04515680825724494,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.9794,
      "mean_token_accuracy": 0.09112128540873528,
      "num_tokens": 1005033.0,
      "step": 455
    },
    {
      "entropy": 7.55339617729187,
      "epoch": 0.045653036919412464,
      "grad_norm": 0.953125,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.0539,
      "mean_token_accuracy": 0.0886804811656475,
      "num_tokens": 1015940.0,
      "step": 460
    },
    {
      "entropy": 7.4837071895599365,
      "epoch": 0.046149265581579994,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.0009,
      "mean_token_accuracy": 0.0916803628206253,
      "num_tokens": 1027582.0,
      "step": 465
    },
    {
      "entropy": 7.472231149673462,
      "epoch": 0.04664549424374752,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.9737,
      "mean_token_accuracy": 0.08888428881764412,
      "num_tokens": 1040046.0,
      "step": 470
    },
    {
      "entropy": 7.467594766616822,
      "epoch": 0.04714172290591505,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000237,
      "loss": 7.0423,
      "mean_token_accuracy": 0.08253751546144486,
      "num_tokens": 1050636.0,
      "step": 475
    },
    {
      "entropy": 7.522167110443116,
      "epoch": 0.04763795156808257,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002395,
      "loss": 6.9296,
      "mean_token_accuracy": 0.09664900973439217,
      "num_tokens": 1060748.0,
      "step": 480
    },
    {
      "entropy": 7.500424242019653,
      "epoch": 0.0481341802302501,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000242,
      "loss": 6.9719,
      "mean_token_accuracy": 0.08677386716008187,
      "num_tokens": 1071940.0,
      "step": 485
    },
    {
      "entropy": 7.417007827758789,
      "epoch": 0.048630408892417626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002445,
      "loss": 6.9827,
      "mean_token_accuracy": 0.08790183737874031,
      "num_tokens": 1082962.0,
      "step": 490
    },
    {
      "entropy": 7.444004106521606,
      "epoch": 0.04912663755458515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000247,
      "loss": 6.96,
      "mean_token_accuracy": 0.09731238633394242,
      "num_tokens": 1093082.0,
      "step": 495
    },
    {
      "entropy": 7.518671083450317,
      "epoch": 0.04962286621675268,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002495,
      "loss": 7.0303,
      "mean_token_accuracy": 0.09159182384610176,
      "num_tokens": 1103528.0,
      "step": 500
    },
    {
      "entropy": 7.388113021850586,
      "epoch": 0.0501190948789202,
      "grad_norm": 1.015625,
      "learning_rate": 0.000252,
      "loss": 6.9133,
      "mean_token_accuracy": 0.08871109336614609,
      "num_tokens": 1114813.0,
      "step": 505
    },
    {
      "entropy": 7.351367425918579,
      "epoch": 0.050615323541087734,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002545,
      "loss": 6.9058,
      "mean_token_accuracy": 0.08876910582184791,
      "num_tokens": 1125345.0,
      "step": 510
    },
    {
      "entropy": 7.353222036361695,
      "epoch": 0.05111155220325526,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000257,
      "loss": 6.9205,
      "mean_token_accuracy": 0.09047875106334687,
      "num_tokens": 1134985.0,
      "step": 515
    },
    {
      "entropy": 7.4535609722137455,
      "epoch": 0.05160778086542279,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002595,
      "loss": 6.9339,
      "mean_token_accuracy": 0.09264354780316353,
      "num_tokens": 1145796.0,
      "step": 520
    },
    {
      "entropy": 7.47781138420105,
      "epoch": 0.05210400952759031,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000262,
      "loss": 7.0635,
      "mean_token_accuracy": 0.08175969608128071,
      "num_tokens": 1156936.0,
      "step": 525
    },
    {
      "entropy": 7.3812775135040285,
      "epoch": 0.05260023818975784,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.9329,
      "mean_token_accuracy": 0.09363158792257309,
      "num_tokens": 1167526.0,
      "step": 530
    },
    {
      "entropy": 7.422448301315308,
      "epoch": 0.053096466851925365,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.9746,
      "mean_token_accuracy": 0.08542074412107467,
      "num_tokens": 1177647.0,
      "step": 535
    },
    {
      "entropy": 7.395486497879029,
      "epoch": 0.053592695514092896,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.9292,
      "mean_token_accuracy": 0.09526413604617119,
      "num_tokens": 1188255.0,
      "step": 540
    },
    {
      "entropy": 7.320084714889527,
      "epoch": 0.05408892417626042,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.8533,
      "mean_token_accuracy": 0.09879972338676453,
      "num_tokens": 1199433.0,
      "step": 545
    },
    {
      "entropy": 7.263059377670288,
      "epoch": 0.05458515283842795,
      "grad_norm": 0.875,
      "learning_rate": 0.0002745,
      "loss": 6.9302,
      "mean_token_accuracy": 0.0863654799759388,
      "num_tokens": 1210067.0,
      "step": 550
    },
    {
      "entropy": 7.307140684127807,
      "epoch": 0.055081381500595473,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000277,
      "loss": 6.8185,
      "mean_token_accuracy": 0.0939325712621212,
      "num_tokens": 1220639.0,
      "step": 555
    },
    {
      "entropy": 7.349351215362549,
      "epoch": 0.055577610162763004,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002795,
      "loss": 6.861,
      "mean_token_accuracy": 0.09410361275076866,
      "num_tokens": 1233158.0,
      "step": 560
    },
    {
      "entropy": 7.274934434890747,
      "epoch": 0.05607383882493053,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.8381,
      "mean_token_accuracy": 0.08947066888213158,
      "num_tokens": 1244552.0,
      "step": 565
    },
    {
      "entropy": 7.424350881576538,
      "epoch": 0.05657006748709806,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002845,
      "loss": 6.9567,
      "mean_token_accuracy": 0.08916749060153961,
      "num_tokens": 1255478.0,
      "step": 570
    },
    {
      "entropy": 7.213148832321167,
      "epoch": 0.05706629614926558,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000287,
      "loss": 6.7664,
      "mean_token_accuracy": 0.09938948079943657,
      "num_tokens": 1267507.0,
      "step": 575
    },
    {
      "entropy": 7.141163396835327,
      "epoch": 0.057562524811433105,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002895,
      "loss": 6.7759,
      "mean_token_accuracy": 0.09177131280303001,
      "num_tokens": 1279377.0,
      "step": 580
    },
    {
      "entropy": 7.411661386489868,
      "epoch": 0.058058753473600636,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000292,
      "loss": 6.8635,
      "mean_token_accuracy": 0.10038365721702576,
      "num_tokens": 1290671.0,
      "step": 585
    },
    {
      "entropy": 7.1382372856140135,
      "epoch": 0.05855498213576816,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002945,
      "loss": 6.8309,
      "mean_token_accuracy": 0.10017411112785339,
      "num_tokens": 1302792.0,
      "step": 590
    },
    {
      "entropy": 7.324809122085571,
      "epoch": 0.05905121079793569,
      "grad_norm": 1.09375,
      "learning_rate": 0.000297,
      "loss": 6.7832,
      "mean_token_accuracy": 0.08957966193556785,
      "num_tokens": 1313315.0,
      "step": 595
    },
    {
      "entropy": 7.144257354736328,
      "epoch": 0.05954743946010321,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002995,
      "loss": 6.8465,
      "mean_token_accuracy": 0.0927205927670002,
      "num_tokens": 1326358.0,
      "step": 600
    },
    {
      "entropy": 7.283497524261475,
      "epoch": 0.060043668122270744,
      "grad_norm": 1.046875,
      "learning_rate": 0.000302,
      "loss": 6.7205,
      "mean_token_accuracy": 0.09314653351902961,
      "num_tokens": 1336266.0,
      "step": 605
    },
    {
      "entropy": 7.164820337295533,
      "epoch": 0.06053989678443827,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003045,
      "loss": 6.6683,
      "mean_token_accuracy": 0.09883529841899871,
      "num_tokens": 1347610.0,
      "step": 610
    },
    {
      "entropy": 7.1178053855896,
      "epoch": 0.0610361254466058,
      "grad_norm": 1.078125,
      "learning_rate": 0.000307,
      "loss": 6.6337,
      "mean_token_accuracy": 0.10055704191327094,
      "num_tokens": 1358317.0,
      "step": 615
    },
    {
      "entropy": 7.149615907669068,
      "epoch": 0.06153235410877332,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003095,
      "loss": 6.7493,
      "mean_token_accuracy": 0.09263555109500884,
      "num_tokens": 1370126.0,
      "step": 620
    },
    {
      "entropy": 7.162906694412231,
      "epoch": 0.06202858277094085,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000312,
      "loss": 6.7092,
      "mean_token_accuracy": 0.09986358359456063,
      "num_tokens": 1381571.0,
      "step": 625
    },
    {
      "entropy": 7.126469230651855,
      "epoch": 0.06252481143310838,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003145,
      "loss": 6.7397,
      "mean_token_accuracy": 0.09763761013746261,
      "num_tokens": 1392486.0,
      "step": 630
    },
    {
      "entropy": 7.15671968460083,
      "epoch": 0.0630210400952759,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000317,
      "loss": 6.751,
      "mean_token_accuracy": 0.09868052825331688,
      "num_tokens": 1403086.0,
      "step": 635
    },
    {
      "entropy": 7.13517165184021,
      "epoch": 0.06351726875744343,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003195,
      "loss": 6.7055,
      "mean_token_accuracy": 0.09947170913219452,
      "num_tokens": 1414537.0,
      "step": 640
    },
    {
      "entropy": 7.06722993850708,
      "epoch": 0.06401349741961096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000322,
      "loss": 6.726,
      "mean_token_accuracy": 0.09579051956534386,
      "num_tokens": 1426998.0,
      "step": 645
    },
    {
      "entropy": 7.146437740325927,
      "epoch": 0.06450972608177849,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.7544,
      "mean_token_accuracy": 0.09660795778036117,
      "num_tokens": 1437667.0,
      "step": 650
    },
    {
      "entropy": 7.114442014694214,
      "epoch": 0.065005954743946,
      "grad_norm": 0.921875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6146,
      "mean_token_accuracy": 0.10034202635288239,
      "num_tokens": 1448280.0,
      "step": 655
    },
    {
      "entropy": 7.085196685791016,
      "epoch": 0.06550218340611354,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.6777,
      "mean_token_accuracy": 0.09931168034672737,
      "num_tokens": 1458608.0,
      "step": 660
    },
    {
      "entropy": 7.150270652770996,
      "epoch": 0.06599841206828107,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.8578,
      "mean_token_accuracy": 0.09408899918198585,
      "num_tokens": 1470435.0,
      "step": 665
    },
    {
      "entropy": 7.095894765853882,
      "epoch": 0.06649464073044858,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.7049,
      "mean_token_accuracy": 0.098654755204916,
      "num_tokens": 1481959.0,
      "step": 670
    },
    {
      "entropy": 7.1083564281463625,
      "epoch": 0.06699086939261611,
      "grad_norm": 1.015625,
      "learning_rate": 0.000337,
      "loss": 6.7581,
      "mean_token_accuracy": 0.09214193895459175,
      "num_tokens": 1493798.0,
      "step": 675
    },
    {
      "entropy": 7.06939730644226,
      "epoch": 0.06748709805478365,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003395,
      "loss": 6.5739,
      "mean_token_accuracy": 0.09953450039029121,
      "num_tokens": 1502969.0,
      "step": 680
    },
    {
      "entropy": 7.057254981994629,
      "epoch": 0.06798332671695118,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000342,
      "loss": 6.7168,
      "mean_token_accuracy": 0.09586154073476791,
      "num_tokens": 1513825.0,
      "step": 685
    },
    {
      "entropy": 7.048639583587646,
      "epoch": 0.06847955537911869,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.6118,
      "mean_token_accuracy": 0.10043377801775932,
      "num_tokens": 1525328.0,
      "step": 690
    },
    {
      "entropy": 7.052638721466065,
      "epoch": 0.06897578404128622,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000347,
      "loss": 6.6426,
      "mean_token_accuracy": 0.09776772186160088,
      "num_tokens": 1537383.0,
      "step": 695
    },
    {
      "entropy": 7.164321613311768,
      "epoch": 0.06947201270345375,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003495,
      "loss": 6.7385,
      "mean_token_accuracy": 0.1011982411146164,
      "num_tokens": 1549671.0,
      "step": 700
    },
    {
      "entropy": 6.959526395797729,
      "epoch": 0.06996824136562128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000352,
      "loss": 6.623,
      "mean_token_accuracy": 0.09803913906216621,
      "num_tokens": 1559974.0,
      "step": 705
    },
    {
      "entropy": 7.045153188705444,
      "epoch": 0.0704644700277888,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003545,
      "loss": 6.725,
      "mean_token_accuracy": 0.09499528259038925,
      "num_tokens": 1570890.0,
      "step": 710
    },
    {
      "entropy": 7.018730640411377,
      "epoch": 0.07096069868995633,
      "grad_norm": 1.15625,
      "learning_rate": 0.000357,
      "loss": 6.622,
      "mean_token_accuracy": 0.10477519556879997,
      "num_tokens": 1581434.0,
      "step": 715
    },
    {
      "entropy": 7.0257142066955565,
      "epoch": 0.07145692735212386,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595,
      "loss": 6.6871,
      "mean_token_accuracy": 0.09542187601327896,
      "num_tokens": 1593710.0,
      "step": 720
    },
    {
      "entropy": 7.100641775131225,
      "epoch": 0.07195315601429139,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000362,
      "loss": 6.6523,
      "mean_token_accuracy": 0.10131741389632225,
      "num_tokens": 1604635.0,
      "step": 725
    },
    {
      "entropy": 7.126769018173218,
      "epoch": 0.07244938467645891,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003645,
      "loss": 6.7134,
      "mean_token_accuracy": 0.10091805458068848,
      "num_tokens": 1615258.0,
      "step": 730
    },
    {
      "entropy": 6.9821710109710695,
      "epoch": 0.07294561333862644,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000367,
      "loss": 6.6178,
      "mean_token_accuracy": 0.09964116811752319,
      "num_tokens": 1625942.0,
      "step": 735
    },
    {
      "entropy": 6.951982069015503,
      "epoch": 0.07344184200079397,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003695,
      "loss": 6.5998,
      "mean_token_accuracy": 0.105032167583704,
      "num_tokens": 1636270.0,
      "step": 740
    },
    {
      "entropy": 6.968512439727784,
      "epoch": 0.07393807066296149,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000372,
      "loss": 6.5799,
      "mean_token_accuracy": 0.10235915258526802,
      "num_tokens": 1647029.0,
      "step": 745
    },
    {
      "entropy": 7.01955018043518,
      "epoch": 0.07443429932512902,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003745,
      "loss": 6.7457,
      "mean_token_accuracy": 0.0975896142423153,
      "num_tokens": 1659483.0,
      "step": 750
    },
    {
      "entropy": 6.89774866104126,
      "epoch": 0.07493052798729655,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000377,
      "loss": 6.6003,
      "mean_token_accuracy": 0.0963432714343071,
      "num_tokens": 1669425.0,
      "step": 755
    },
    {
      "entropy": 7.036856985092163,
      "epoch": 0.07542675664946408,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003795,
      "loss": 6.6063,
      "mean_token_accuracy": 0.09760197177529335,
      "num_tokens": 1680426.0,
      "step": 760
    },
    {
      "entropy": 6.869618892669678,
      "epoch": 0.0759229853116316,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000382,
      "loss": 6.6246,
      "mean_token_accuracy": 0.09728450551629067,
      "num_tokens": 1692803.0,
      "step": 765
    },
    {
      "entropy": 7.04201602935791,
      "epoch": 0.07641921397379912,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0003845,
      "loss": 6.6441,
      "mean_token_accuracy": 0.09765328615903854,
      "num_tokens": 1703919.0,
      "step": 770
    },
    {
      "entropy": 6.969884634017944,
      "epoch": 0.07691544263596665,
      "grad_norm": 0.875,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6147,
      "mean_token_accuracy": 0.10524557158350945,
      "num_tokens": 1714487.0,
      "step": 775
    },
    {
      "entropy": 6.979475402832032,
      "epoch": 0.07741167129813419,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6626,
      "mean_token_accuracy": 0.09930093437433243,
      "num_tokens": 1726601.0,
      "step": 780
    },
    {
      "entropy": 6.933769798278808,
      "epoch": 0.0779078999603017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.554,
      "mean_token_accuracy": 0.10119575038552284,
      "num_tokens": 1738743.0,
      "step": 785
    },
    {
      "entropy": 6.9056957244873045,
      "epoch": 0.07840412862246923,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.6345,
      "mean_token_accuracy": 0.10144512727856636,
      "num_tokens": 1748286.0,
      "step": 790
    },
    {
      "entropy": 6.91564908027649,
      "epoch": 0.07890035728463676,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.6636,
      "mean_token_accuracy": 0.09903955459594727,
      "num_tokens": 1758917.0,
      "step": 795
    },
    {
      "entropy": 6.979661178588867,
      "epoch": 0.0793965859468043,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003995,
      "loss": 6.4992,
      "mean_token_accuracy": 0.10489165559411048,
      "num_tokens": 1769179.0,
      "step": 800
    },
    {
      "entropy": 6.849152421951294,
      "epoch": 0.07989281460897181,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000402,
      "loss": 6.6226,
      "mean_token_accuracy": 0.09758250415325165,
      "num_tokens": 1782369.0,
      "step": 805
    },
    {
      "entropy": 6.871382761001587,
      "epoch": 0.08038904327113934,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004045,
      "loss": 6.6281,
      "mean_token_accuracy": 0.10324807316064835,
      "num_tokens": 1795992.0,
      "step": 810
    },
    {
      "entropy": 6.90522723197937,
      "epoch": 0.08088527193330687,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5238,
      "mean_token_accuracy": 0.10905546322464943,
      "num_tokens": 1806938.0,
      "step": 815
    },
    {
      "entropy": 6.812549114227295,
      "epoch": 0.0813815005954744,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004095,
      "loss": 6.4573,
      "mean_token_accuracy": 0.10598495304584503,
      "num_tokens": 1817663.0,
      "step": 820
    },
    {
      "entropy": 6.936794233322144,
      "epoch": 0.08187772925764192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000412,
      "loss": 6.6127,
      "mean_token_accuracy": 0.10083081871271134,
      "num_tokens": 1828254.0,
      "step": 825
    },
    {
      "entropy": 6.896241903305054,
      "epoch": 0.08237395791980945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004145,
      "loss": 6.5339,
      "mean_token_accuracy": 0.10185903310775757,
      "num_tokens": 1839001.0,
      "step": 830
    },
    {
      "entropy": 6.864541244506836,
      "epoch": 0.08287018658197698,
      "grad_norm": 0.90625,
      "learning_rate": 0.000417,
      "loss": 6.6313,
      "mean_token_accuracy": 0.09963079243898391,
      "num_tokens": 1849906.0,
      "step": 835
    },
    {
      "entropy": 6.797179365158081,
      "epoch": 0.0833664152441445,
      "grad_norm": 1.0,
      "learning_rate": 0.0004195,
      "loss": 6.5266,
      "mean_token_accuracy": 0.09986047223210334,
      "num_tokens": 1860982.0,
      "step": 840
    },
    {
      "entropy": 6.9268652439117435,
      "epoch": 0.08386264390631203,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000422,
      "loss": 6.4752,
      "mean_token_accuracy": 0.10353965610265732,
      "num_tokens": 1872177.0,
      "step": 845
    },
    {
      "entropy": 6.7672933578491214,
      "epoch": 0.08435887256847956,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004245,
      "loss": 6.455,
      "mean_token_accuracy": 0.10329471975564956,
      "num_tokens": 1882783.0,
      "step": 850
    },
    {
      "entropy": 6.962534761428833,
      "epoch": 0.08485510123064709,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000427,
      "loss": 6.6426,
      "mean_token_accuracy": 0.09471128433942795,
      "num_tokens": 1893706.0,
      "step": 855
    },
    {
      "entropy": 6.8037285804748535,
      "epoch": 0.0853513298928146,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004295,
      "loss": 6.5138,
      "mean_token_accuracy": 0.10021722540259362,
      "num_tokens": 1903852.0,
      "step": 860
    },
    {
      "entropy": 6.835017728805542,
      "epoch": 0.08584755855498213,
      "grad_norm": 1.046875,
      "learning_rate": 0.000432,
      "loss": 6.5342,
      "mean_token_accuracy": 0.10754364952445031,
      "num_tokens": 1914157.0,
      "step": 865
    },
    {
      "entropy": 6.85911717414856,
      "epoch": 0.08634378721714966,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004345,
      "loss": 6.4726,
      "mean_token_accuracy": 0.10122424364089966,
      "num_tokens": 1924684.0,
      "step": 870
    },
    {
      "entropy": 6.886322259902954,
      "epoch": 0.0868400158793172,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000437,
      "loss": 6.6263,
      "mean_token_accuracy": 0.09748353064060211,
      "num_tokens": 1935657.0,
      "step": 875
    },
    {
      "entropy": 6.7788018703460695,
      "epoch": 0.08733624454148471,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004395,
      "loss": 6.5028,
      "mean_token_accuracy": 0.10384351089596748,
      "num_tokens": 1946735.0,
      "step": 880
    },
    {
      "entropy": 6.872425079345703,
      "epoch": 0.08783247320365224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000442,
      "loss": 6.6691,
      "mean_token_accuracy": 0.09842884838581085,
      "num_tokens": 1957871.0,
      "step": 885
    },
    {
      "entropy": 6.7883913040161135,
      "epoch": 0.08832870186581977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004445,
      "loss": 6.5061,
      "mean_token_accuracy": 0.10142220854759217,
      "num_tokens": 1969032.0,
      "step": 890
    },
    {
      "entropy": 6.859635829925537,
      "epoch": 0.0888249305279873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000447,
      "loss": 6.5319,
      "mean_token_accuracy": 0.10258013382554054,
      "num_tokens": 1980212.0,
      "step": 895
    },
    {
      "entropy": 6.746867752075195,
      "epoch": 0.08932115919015482,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.4491,
      "mean_token_accuracy": 0.10580305159091949,
      "num_tokens": 1990333.0,
      "step": 900
    },
    {
      "entropy": 6.797193908691407,
      "epoch": 0.08981738785232235,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.5216,
      "mean_token_accuracy": 0.10975843593478203,
      "num_tokens": 2001486.0,
      "step": 905
    },
    {
      "entropy": 6.801945495605469,
      "epoch": 0.09031361651448988,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.5084,
      "mean_token_accuracy": 0.10368723571300506,
      "num_tokens": 2012535.0,
      "step": 910
    },
    {
      "entropy": 6.765622854232788,
      "epoch": 0.0908098451766574,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.4628,
      "mean_token_accuracy": 0.10697698220610619,
      "num_tokens": 2024116.0,
      "step": 915
    },
    {
      "entropy": 6.651122283935547,
      "epoch": 0.09130607383882493,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.5028,
      "mean_token_accuracy": 0.10319509655237198,
      "num_tokens": 2036341.0,
      "step": 920
    },
    {
      "entropy": 6.888702630996704,
      "epoch": 0.09180230250099246,
      "grad_norm": 0.953125,
      "learning_rate": 0.000462,
      "loss": 6.4747,
      "mean_token_accuracy": 0.10404421612620354,
      "num_tokens": 2046768.0,
      "step": 925
    },
    {
      "entropy": 6.6753270626068115,
      "epoch": 0.09229853116315999,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004645,
      "loss": 6.4364,
      "mean_token_accuracy": 0.10831291154026985,
      "num_tokens": 2057315.0,
      "step": 930
    },
    {
      "entropy": 6.732809257507324,
      "epoch": 0.0927947598253275,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000467,
      "loss": 6.4229,
      "mean_token_accuracy": 0.10807890966534614,
      "num_tokens": 2068267.0,
      "step": 935
    },
    {
      "entropy": 6.747051239013672,
      "epoch": 0.09329098848749504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004695,
      "loss": 6.4919,
      "mean_token_accuracy": 0.1026596762239933,
      "num_tokens": 2078271.0,
      "step": 940
    },
    {
      "entropy": 6.686890411376953,
      "epoch": 0.09378721714966257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000472,
      "loss": 6.4104,
      "mean_token_accuracy": 0.10279989019036292,
      "num_tokens": 2088381.0,
      "step": 945
    },
    {
      "entropy": 6.7383002758026125,
      "epoch": 0.0942834458118301,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004745,
      "loss": 6.3729,
      "mean_token_accuracy": 0.10531853288412094,
      "num_tokens": 2098899.0,
      "step": 950
    },
    {
      "entropy": 6.682248830795288,
      "epoch": 0.09477967447399761,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000477,
      "loss": 6.5398,
      "mean_token_accuracy": 0.10515285804867744,
      "num_tokens": 2110057.0,
      "step": 955
    },
    {
      "entropy": 6.719785499572754,
      "epoch": 0.09527590313616514,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004795,
      "loss": 6.4403,
      "mean_token_accuracy": 0.10757240056991577,
      "num_tokens": 2121758.0,
      "step": 960
    },
    {
      "entropy": 6.719501113891601,
      "epoch": 0.09577213179833267,
      "grad_norm": 0.921875,
      "learning_rate": 0.000482,
      "loss": 6.4557,
      "mean_token_accuracy": 0.10785956755280494,
      "num_tokens": 2131840.0,
      "step": 965
    },
    {
      "entropy": 6.616270351409912,
      "epoch": 0.0962683604605002,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845,
      "loss": 6.4294,
      "mean_token_accuracy": 0.10218973308801652,
      "num_tokens": 2142084.0,
      "step": 970
    },
    {
      "entropy": 6.70799412727356,
      "epoch": 0.09676458912266772,
      "grad_norm": 0.953125,
      "learning_rate": 0.000487,
      "loss": 6.4822,
      "mean_token_accuracy": 0.10346104353666305,
      "num_tokens": 2152785.0,
      "step": 975
    },
    {
      "entropy": 6.7580810546875,
      "epoch": 0.09726081778483525,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004895,
      "loss": 6.4715,
      "mean_token_accuracy": 0.10466889962553978,
      "num_tokens": 2163774.0,
      "step": 980
    },
    {
      "entropy": 6.600415515899658,
      "epoch": 0.09775704644700278,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000492,
      "loss": 6.3865,
      "mean_token_accuracy": 0.10628859028220176,
      "num_tokens": 2175378.0,
      "step": 985
    },
    {
      "entropy": 6.872743368148804,
      "epoch": 0.0982532751091703,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004945,
      "loss": 6.6002,
      "mean_token_accuracy": 0.10284787714481354,
      "num_tokens": 2186510.0,
      "step": 990
    },
    {
      "entropy": 6.675116109848022,
      "epoch": 0.09874950377133783,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000497,
      "loss": 6.3875,
      "mean_token_accuracy": 0.11210239455103874,
      "num_tokens": 2197482.0,
      "step": 995
    },
    {
      "entropy": 6.613697242736817,
      "epoch": 0.09924573243350536,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995,
      "loss": 6.3886,
      "mean_token_accuracy": 0.1066878505051136,
      "num_tokens": 2208329.0,
      "step": 1000
    },
    {
      "entropy": 6.565951251983643,
      "epoch": 0.09974196109567289,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999999982145551,
      "loss": 6.2974,
      "mean_token_accuracy": 0.10837465524673462,
      "num_tokens": 2218632.0,
      "step": 1005
    },
    {
      "entropy": 6.640917587280273,
      "epoch": 0.1002381897578404,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004999999909611855,
      "loss": 6.4173,
      "mean_token_accuracy": 0.10387595295906067,
      "num_tokens": 2231335.0,
      "step": 1010
    },
    {
      "entropy": 6.693672466278076,
      "epoch": 0.10073441842000794,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999999781283008,
      "loss": 6.3927,
      "mean_token_accuracy": 0.10786194428801536,
      "num_tokens": 2241704.0,
      "step": 1015
    },
    {
      "entropy": 6.652690649032593,
      "epoch": 0.10123064708217547,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999999597159017,
      "loss": 6.448,
      "mean_token_accuracy": 0.1010373942553997,
      "num_tokens": 2252310.0,
      "step": 1020
    },
    {
      "entropy": 6.650144433975219,
      "epoch": 0.101726875744343,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999999357239882,
      "loss": 6.3674,
      "mean_token_accuracy": 0.10643667429685592,
      "num_tokens": 2262270.0,
      "step": 1025
    },
    {
      "entropy": 6.674994850158692,
      "epoch": 0.10222310440651051,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999999061525612,
      "loss": 6.5075,
      "mean_token_accuracy": 0.10479943379759789,
      "num_tokens": 2273986.0,
      "step": 1030
    },
    {
      "entropy": 6.751373052597046,
      "epoch": 0.10271933306867805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999998710016214,
      "loss": 6.4006,
      "mean_token_accuracy": 0.10447262153029442,
      "num_tokens": 2283430.0,
      "step": 1035
    },
    {
      "entropy": 6.647673654556274,
      "epoch": 0.10321556173084558,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999998302711696,
      "loss": 6.4082,
      "mean_token_accuracy": 0.09957063123583794,
      "num_tokens": 2294201.0,
      "step": 1040
    },
    {
      "entropy": 6.568925905227661,
      "epoch": 0.1037117903930131,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999997839612069,
      "loss": 6.309,
      "mean_token_accuracy": 0.10835110917687416,
      "num_tokens": 2303876.0,
      "step": 1045
    },
    {
      "entropy": 6.604097175598144,
      "epoch": 0.10420801905518062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999997320717342,
      "loss": 6.3495,
      "mean_token_accuracy": 0.1118161328136921,
      "num_tokens": 2313713.0,
      "step": 1050
    },
    {
      "entropy": 6.559231519699097,
      "epoch": 0.10470424771734815,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999996746027531,
      "loss": 6.3595,
      "mean_token_accuracy": 0.10454496443271637,
      "num_tokens": 2324862.0,
      "step": 1055
    },
    {
      "entropy": 6.71983904838562,
      "epoch": 0.10520047637951568,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499999611554265,
      "loss": 6.431,
      "mean_token_accuracy": 0.10956530421972274,
      "num_tokens": 2334974.0,
      "step": 1060
    },
    {
      "entropy": 6.596372747421265,
      "epoch": 0.10569670504168321,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999995429262712,
      "loss": 6.4574,
      "mean_token_accuracy": 0.1073372520506382,
      "num_tokens": 2345945.0,
      "step": 1065
    },
    {
      "entropy": 6.653563451766968,
      "epoch": 0.10619293370385073,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999994687187737,
      "loss": 6.397,
      "mean_token_accuracy": 0.10704459249973297,
      "num_tokens": 2356749.0,
      "step": 1070
    },
    {
      "entropy": 6.710749959945678,
      "epoch": 0.10668916236601826,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999993889317742,
      "loss": 6.4721,
      "mean_token_accuracy": 0.1034046970307827,
      "num_tokens": 2368181.0,
      "step": 1075
    },
    {
      "entropy": 6.5815818309783936,
      "epoch": 0.10718539102818579,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999993035652746,
      "loss": 6.38,
      "mean_token_accuracy": 0.10739323794841767,
      "num_tokens": 2377737.0,
      "step": 1080
    },
    {
      "entropy": 6.599908828735352,
      "epoch": 0.10768161969035331,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999992126192771,
      "loss": 6.3006,
      "mean_token_accuracy": 0.10967165976762772,
      "num_tokens": 2388944.0,
      "step": 1085
    },
    {
      "entropy": 6.593209838867187,
      "epoch": 0.10817784835252084,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499999116093784,
      "loss": 6.3809,
      "mean_token_accuracy": 0.10844719260931016,
      "num_tokens": 2400758.0,
      "step": 1090
    },
    {
      "entropy": 6.568376064300537,
      "epoch": 0.10867407701468837,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999990139887976,
      "loss": 6.3747,
      "mean_token_accuracy": 0.11013728603720666,
      "num_tokens": 2411916.0,
      "step": 1095
    },
    {
      "entropy": 6.728429841995239,
      "epoch": 0.1091703056768559,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999989063043206,
      "loss": 6.4014,
      "mean_token_accuracy": 0.10465751141309738,
      "num_tokens": 2423310.0,
      "step": 1100
    },
    {
      "entropy": 6.530461025238037,
      "epoch": 0.10966653433902342,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999987930403555,
      "loss": 6.3331,
      "mean_token_accuracy": 0.10839267894625663,
      "num_tokens": 2432100.0,
      "step": 1105
    },
    {
      "entropy": 6.574756145477295,
      "epoch": 0.11016276300119095,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999986741969051,
      "loss": 6.4049,
      "mean_token_accuracy": 0.1044836975634098,
      "num_tokens": 2443563.0,
      "step": 1110
    },
    {
      "entropy": 6.649301338195801,
      "epoch": 0.11065899166335848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999985497739724,
      "loss": 6.4178,
      "mean_token_accuracy": 0.10569542646408081,
      "num_tokens": 2455435.0,
      "step": 1115
    },
    {
      "entropy": 6.570544004440308,
      "epoch": 0.11115522032552601,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999984197715606,
      "loss": 6.3493,
      "mean_token_accuracy": 0.10882077589631081,
      "num_tokens": 2466173.0,
      "step": 1120
    },
    {
      "entropy": 6.516915893554687,
      "epoch": 0.11165144898769352,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999982841896728,
      "loss": 6.3036,
      "mean_token_accuracy": 0.10695828124880791,
      "num_tokens": 2476349.0,
      "step": 1125
    },
    {
      "entropy": 6.61347975730896,
      "epoch": 0.11214767764986106,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999981430283123,
      "loss": 6.4251,
      "mean_token_accuracy": 0.11261754408478737,
      "num_tokens": 2487864.0,
      "step": 1130
    },
    {
      "entropy": 6.566452312469482,
      "epoch": 0.11264390631202859,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999979962874827,
      "loss": 6.361,
      "mean_token_accuracy": 0.10681881308555603,
      "num_tokens": 2499188.0,
      "step": 1135
    },
    {
      "entropy": 6.645647430419922,
      "epoch": 0.11314013497419612,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999978439671877,
      "loss": 6.3345,
      "mean_token_accuracy": 0.10881530791521073,
      "num_tokens": 2509813.0,
      "step": 1140
    },
    {
      "entropy": 6.521003484725952,
      "epoch": 0.11363636363636363,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999976860674309,
      "loss": 6.3642,
      "mean_token_accuracy": 0.10348583310842514,
      "num_tokens": 2520813.0,
      "step": 1145
    },
    {
      "entropy": 6.54542760848999,
      "epoch": 0.11413259229853116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999975225882164,
      "loss": 6.363,
      "mean_token_accuracy": 0.10692540258169174,
      "num_tokens": 2531802.0,
      "step": 1150
    },
    {
      "entropy": 6.528800964355469,
      "epoch": 0.1146288209606987,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499997353529548,
      "loss": 6.1676,
      "mean_token_accuracy": 0.12081934958696365,
      "num_tokens": 2541935.0,
      "step": 1155
    },
    {
      "entropy": 6.474112510681152,
      "epoch": 0.11512504962286621,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999971788914303,
      "loss": 6.2718,
      "mean_token_accuracy": 0.1114352934062481,
      "num_tokens": 2553224.0,
      "step": 1160
    },
    {
      "entropy": 6.483311462402344,
      "epoch": 0.11562127828503374,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999969986738672,
      "loss": 6.3156,
      "mean_token_accuracy": 0.10600073337554931,
      "num_tokens": 2565355.0,
      "step": 1165
    },
    {
      "entropy": 6.61416974067688,
      "epoch": 0.11611750694720127,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999968128768634,
      "loss": 6.4,
      "mean_token_accuracy": 0.10448111891746521,
      "num_tokens": 2575962.0,
      "step": 1170
    },
    {
      "entropy": 6.583769941329956,
      "epoch": 0.1166137356093688,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999966215004236,
      "loss": 6.4055,
      "mean_token_accuracy": 0.11200469061732292,
      "num_tokens": 2588136.0,
      "step": 1175
    },
    {
      "entropy": 6.557573032379151,
      "epoch": 0.11710996427153632,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999964245445522,
      "loss": 6.2965,
      "mean_token_accuracy": 0.11258405596017837,
      "num_tokens": 2597785.0,
      "step": 1180
    },
    {
      "entropy": 6.516626834869385,
      "epoch": 0.11760619293370385,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999962220092545,
      "loss": 6.337,
      "mean_token_accuracy": 0.11605330407619477,
      "num_tokens": 2609738.0,
      "step": 1185
    },
    {
      "entropy": 6.4262982368469235,
      "epoch": 0.11810242159587138,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999960138945351,
      "loss": 6.2232,
      "mean_token_accuracy": 0.11857966929674149,
      "num_tokens": 2620179.0,
      "step": 1190
    },
    {
      "entropy": 6.496797704696656,
      "epoch": 0.11859865025803891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999958002003996,
      "loss": 6.2665,
      "mean_token_accuracy": 0.11352286636829376,
      "num_tokens": 2631550.0,
      "step": 1195
    },
    {
      "entropy": 6.50876030921936,
      "epoch": 0.11909487892020643,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004999955809268529,
      "loss": 6.3027,
      "mean_token_accuracy": 0.10872863009572029,
      "num_tokens": 2642467.0,
      "step": 1200
    },
    {
      "entropy": 6.478094530105591,
      "epoch": 0.11959110758237396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999953560739008,
      "loss": 6.2868,
      "mean_token_accuracy": 0.11035665571689605,
      "num_tokens": 2652132.0,
      "step": 1205
    },
    {
      "entropy": 6.516570568084717,
      "epoch": 0.12008733624454149,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999951256415485,
      "loss": 6.3359,
      "mean_token_accuracy": 0.1081702619791031,
      "num_tokens": 2663336.0,
      "step": 1210
    },
    {
      "entropy": 6.608760070800781,
      "epoch": 0.12058356490670902,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000499994889629802,
      "loss": 6.4131,
      "mean_token_accuracy": 0.10076656267046928,
      "num_tokens": 2674201.0,
      "step": 1215
    },
    {
      "entropy": 6.543395376205444,
      "epoch": 0.12107979356887653,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499994648038667,
      "loss": 6.3385,
      "mean_token_accuracy": 0.11510411649942398,
      "num_tokens": 2685073.0,
      "step": 1220
    },
    {
      "entropy": 6.572003412246704,
      "epoch": 0.12157602223104406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999944008681496,
      "loss": 6.2812,
      "mean_token_accuracy": 0.11140692159533501,
      "num_tokens": 2695024.0,
      "step": 1225
    },
    {
      "entropy": 6.5019573211669925,
      "epoch": 0.1220722508932116,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999941481182559,
      "loss": 6.3588,
      "mean_token_accuracy": 0.10714156180620193,
      "num_tokens": 2706232.0,
      "step": 1230
    },
    {
      "entropy": 6.5704296112060545,
      "epoch": 0.12256847955537911,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999938897889921,
      "loss": 6.3373,
      "mean_token_accuracy": 0.11231422051787376,
      "num_tokens": 2715874.0,
      "step": 1235
    },
    {
      "entropy": 6.454069900512695,
      "epoch": 0.12306470821754664,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004999936258803647,
      "loss": 6.2863,
      "mean_token_accuracy": 0.11209510788321495,
      "num_tokens": 2727409.0,
      "step": 1240
    },
    {
      "entropy": 6.470958852767945,
      "epoch": 0.12356093687971417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999933563923802,
      "loss": 6.1948,
      "mean_token_accuracy": 0.11663963794708251,
      "num_tokens": 2737259.0,
      "step": 1245
    },
    {
      "entropy": 6.504468393325806,
      "epoch": 0.1240571655418817,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004999930813250451,
      "loss": 6.3227,
      "mean_token_accuracy": 0.11047348752617836,
      "num_tokens": 2748416.0,
      "step": 1250
    },
    {
      "entropy": 6.485745906829834,
      "epoch": 0.12455339420404922,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999928006783667,
      "loss": 6.2973,
      "mean_token_accuracy": 0.11183110401034355,
      "num_tokens": 2760470.0,
      "step": 1255
    },
    {
      "entropy": 6.535139369964599,
      "epoch": 0.12504962286621676,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999925144523515,
      "loss": 6.3279,
      "mean_token_accuracy": 0.10748472735285759,
      "num_tokens": 2770222.0,
      "step": 1260
    },
    {
      "entropy": 6.476829242706299,
      "epoch": 0.12554585152838427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999922226470066,
      "loss": 6.2505,
      "mean_token_accuracy": 0.11203066259622574,
      "num_tokens": 2780212.0,
      "step": 1265
    },
    {
      "entropy": 6.390298366546631,
      "epoch": 0.1260420801905518,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999919252623397,
      "loss": 6.2144,
      "mean_token_accuracy": 0.11430977880954743,
      "num_tokens": 2790239.0,
      "step": 1270
    },
    {
      "entropy": 6.52067379951477,
      "epoch": 0.12653830885271933,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999916222983577,
      "loss": 6.254,
      "mean_token_accuracy": 0.11566178128123283,
      "num_tokens": 2801902.0,
      "step": 1275
    },
    {
      "entropy": 6.4108202934265135,
      "epoch": 0.12703453751488686,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999913137550682,
      "loss": 6.2473,
      "mean_token_accuracy": 0.11208073645830155,
      "num_tokens": 2813268.0,
      "step": 1280
    },
    {
      "entropy": 6.478806877136231,
      "epoch": 0.1275307661770544,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004999909996324791,
      "loss": 6.32,
      "mean_token_accuracy": 0.10953322052955627,
      "num_tokens": 2824506.0,
      "step": 1285
    },
    {
      "entropy": 6.505747509002686,
      "epoch": 0.12802699483922192,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499990679930598,
      "loss": 6.2224,
      "mean_token_accuracy": 0.11515569612383843,
      "num_tokens": 2835144.0,
      "step": 1290
    },
    {
      "entropy": 6.419563961029053,
      "epoch": 0.12852322350138945,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999903546494327,
      "loss": 6.2472,
      "mean_token_accuracy": 0.11156807765364647,
      "num_tokens": 2845382.0,
      "step": 1295
    },
    {
      "entropy": 6.44638466835022,
      "epoch": 0.12901945216355698,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999900237889916,
      "loss": 6.2666,
      "mean_token_accuracy": 0.11455247327685356,
      "num_tokens": 2854697.0,
      "step": 1300
    },
    {
      "entropy": 6.497555685043335,
      "epoch": 0.12951568082572448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999896873492827,
      "loss": 6.2643,
      "mean_token_accuracy": 0.10822557806968688,
      "num_tokens": 2866216.0,
      "step": 1305
    },
    {
      "entropy": 6.422132730484009,
      "epoch": 0.130011909487892,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999893453303143,
      "loss": 6.2308,
      "mean_token_accuracy": 0.11659785062074661,
      "num_tokens": 2877734.0,
      "step": 1310
    },
    {
      "entropy": 6.505683135986328,
      "epoch": 0.13050813815005954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999889977320949,
      "loss": 6.289,
      "mean_token_accuracy": 0.1148169457912445,
      "num_tokens": 2889445.0,
      "step": 1315
    },
    {
      "entropy": 6.354604768753052,
      "epoch": 0.13100436681222707,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999886445546332,
      "loss": 6.2237,
      "mean_token_accuracy": 0.1161403276026249,
      "num_tokens": 2900485.0,
      "step": 1320
    },
    {
      "entropy": 6.544115304946899,
      "epoch": 0.1315005954743946,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499988285797938,
      "loss": 6.2087,
      "mean_token_accuracy": 0.11861399188637733,
      "num_tokens": 2909795.0,
      "step": 1325
    },
    {
      "entropy": 6.294916772842408,
      "epoch": 0.13199682413656214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499987921462018,
      "loss": 6.2314,
      "mean_token_accuracy": 0.12291175946593284,
      "num_tokens": 2920153.0,
      "step": 1330
    },
    {
      "entropy": 6.524097633361817,
      "epoch": 0.13249305279872967,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999875515468825,
      "loss": 6.2228,
      "mean_token_accuracy": 0.1149907335639,
      "num_tokens": 2930457.0,
      "step": 1335
    },
    {
      "entropy": 6.369509267807007,
      "epoch": 0.13298928146089717,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999871760525403,
      "loss": 6.2798,
      "mean_token_accuracy": 0.11520167663693429,
      "num_tokens": 2941305.0,
      "step": 1340
    },
    {
      "entropy": 6.499384498596191,
      "epoch": 0.1334855101230647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999867949790011,
      "loss": 6.2212,
      "mean_token_accuracy": 0.11666619777679443,
      "num_tokens": 2951798.0,
      "step": 1345
    },
    {
      "entropy": 6.377152729034424,
      "epoch": 0.13398173878523223,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999864083262743,
      "loss": 6.1253,
      "mean_token_accuracy": 0.11854363307356834,
      "num_tokens": 2962486.0,
      "step": 1350
    },
    {
      "entropy": 6.414649248123169,
      "epoch": 0.13447796744739976,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999860160943691,
      "loss": 6.2323,
      "mean_token_accuracy": 0.1163373351097107,
      "num_tokens": 2974317.0,
      "step": 1355
    },
    {
      "entropy": 6.405454921722412,
      "epoch": 0.1349741961095673,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999856182832958,
      "loss": 6.1428,
      "mean_token_accuracy": 0.11802050322294236,
      "num_tokens": 2985440.0,
      "step": 1360
    },
    {
      "entropy": 6.40319504737854,
      "epoch": 0.13547042477173482,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999852148930637,
      "loss": 6.222,
      "mean_token_accuracy": 0.11579161584377289,
      "num_tokens": 2997044.0,
      "step": 1365
    },
    {
      "entropy": 6.445984601974487,
      "epoch": 0.13596665343390235,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999848059236832,
      "loss": 6.2333,
      "mean_token_accuracy": 0.11098912060260772,
      "num_tokens": 3005965.0,
      "step": 1370
    },
    {
      "entropy": 6.295212173461914,
      "epoch": 0.13646288209606988,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999843913751643,
      "loss": 6.1123,
      "mean_token_accuracy": 0.11798060461878776,
      "num_tokens": 3015031.0,
      "step": 1375
    },
    {
      "entropy": 6.371025562286377,
      "epoch": 0.13695911075823738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999839712475174,
      "loss": 6.1708,
      "mean_token_accuracy": 0.11246290653944016,
      "num_tokens": 3026275.0,
      "step": 1380
    },
    {
      "entropy": 6.35689868927002,
      "epoch": 0.13745533942040492,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999835455407526,
      "loss": 6.2198,
      "mean_token_accuracy": 0.11414360404014587,
      "num_tokens": 3036778.0,
      "step": 1385
    },
    {
      "entropy": 6.4954687595367435,
      "epoch": 0.13795156808257245,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999831142548808,
      "loss": 6.3258,
      "mean_token_accuracy": 0.11213771849870682,
      "num_tokens": 3048149.0,
      "step": 1390
    },
    {
      "entropy": 6.3874505996704105,
      "epoch": 0.13844779674473998,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999826773899124,
      "loss": 6.1823,
      "mean_token_accuracy": 0.11371054351329804,
      "num_tokens": 3059149.0,
      "step": 1395
    },
    {
      "entropy": 6.409183120727539,
      "epoch": 0.1389440254069075,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004999822349458586,
      "loss": 6.2059,
      "mean_token_accuracy": 0.1147544652223587,
      "num_tokens": 3071613.0,
      "step": 1400
    },
    {
      "entropy": 6.439931964874267,
      "epoch": 0.13944025406907504,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999817869227301,
      "loss": 6.3355,
      "mean_token_accuracy": 0.11122817322611808,
      "num_tokens": 3084742.0,
      "step": 1405
    },
    {
      "entropy": 6.420901536941528,
      "epoch": 0.13993648273124257,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999813333205381,
      "loss": 6.2625,
      "mean_token_accuracy": 0.11637563481926919,
      "num_tokens": 3095356.0,
      "step": 1410
    },
    {
      "entropy": 6.433026123046875,
      "epoch": 0.14043271139341007,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999808741392937,
      "loss": 6.2222,
      "mean_token_accuracy": 0.10983924940228462,
      "num_tokens": 3107717.0,
      "step": 1415
    },
    {
      "entropy": 6.381127834320068,
      "epoch": 0.1409289400555776,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004999804093790086,
      "loss": 6.2182,
      "mean_token_accuracy": 0.11328766494989395,
      "num_tokens": 3119205.0,
      "step": 1420
    },
    {
      "entropy": 6.441397428512573,
      "epoch": 0.14142516871774513,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499979939039694,
      "loss": 6.2685,
      "mean_token_accuracy": 0.11661531254649163,
      "num_tokens": 3130564.0,
      "step": 1425
    },
    {
      "entropy": 6.354721212387085,
      "epoch": 0.14192139737991266,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999794631213618,
      "loss": 6.2024,
      "mean_token_accuracy": 0.11099046766757965,
      "num_tokens": 3141883.0,
      "step": 1430
    },
    {
      "entropy": 6.4592526912689205,
      "epoch": 0.1424176260420802,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999789816240236,
      "loss": 6.208,
      "mean_token_accuracy": 0.11153334975242615,
      "num_tokens": 3152817.0,
      "step": 1435
    },
    {
      "entropy": 6.309853506088257,
      "epoch": 0.14291385470424772,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999784945476915,
      "loss": 6.1224,
      "mean_token_accuracy": 0.12020867317914963,
      "num_tokens": 3163546.0,
      "step": 1440
    },
    {
      "entropy": 6.413205289840699,
      "epoch": 0.14341008336641525,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999780018923775,
      "loss": 6.1713,
      "mean_token_accuracy": 0.1094489723443985,
      "num_tokens": 3173899.0,
      "step": 1445
    },
    {
      "entropy": 6.2561510562896725,
      "epoch": 0.14390631202858278,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999775036580939,
      "loss": 6.0903,
      "mean_token_accuracy": 0.11627468764781952,
      "num_tokens": 3184433.0,
      "step": 1450
    },
    {
      "entropy": 6.487192440032959,
      "epoch": 0.1444025406907503,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999769998448529,
      "loss": 6.1973,
      "mean_token_accuracy": 0.1131520263850689,
      "num_tokens": 3196077.0,
      "step": 1455
    },
    {
      "entropy": 6.187653255462647,
      "epoch": 0.14489876935291782,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999764904526672,
      "loss": 6.1208,
      "mean_token_accuracy": 0.11931058466434478,
      "num_tokens": 3206645.0,
      "step": 1460
    },
    {
      "entropy": 6.462401914596557,
      "epoch": 0.14539499801508535,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999759754815492,
      "loss": 6.2386,
      "mean_token_accuracy": 0.11986212581396102,
      "num_tokens": 3216786.0,
      "step": 1465
    },
    {
      "entropy": 6.32322940826416,
      "epoch": 0.14589122667725288,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999754549315118,
      "loss": 6.1656,
      "mean_token_accuracy": 0.11642653867602348,
      "num_tokens": 3227384.0,
      "step": 1470
    },
    {
      "entropy": 6.288940668106079,
      "epoch": 0.1463874553394204,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499974928802568,
      "loss": 6.0217,
      "mean_token_accuracy": 0.1218691572546959,
      "num_tokens": 3238800.0,
      "step": 1475
    },
    {
      "entropy": 6.292658805847168,
      "epoch": 0.14688368400158794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999743970947306,
      "loss": 6.1753,
      "mean_token_accuracy": 0.11594840288162231,
      "num_tokens": 3248627.0,
      "step": 1480
    },
    {
      "entropy": 6.4044637203216555,
      "epoch": 0.14737991266375547,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499973859808013,
      "loss": 6.1955,
      "mean_token_accuracy": 0.11360678747296334,
      "num_tokens": 3258989.0,
      "step": 1485
    },
    {
      "entropy": 6.28885703086853,
      "epoch": 0.14787614132592297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999733169424284,
      "loss": 6.175,
      "mean_token_accuracy": 0.11919418796896934,
      "num_tokens": 3269474.0,
      "step": 1490
    },
    {
      "entropy": 6.379557180404663,
      "epoch": 0.1483723699880905,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999727684979904,
      "loss": 6.1567,
      "mean_token_accuracy": 0.11620205119252205,
      "num_tokens": 3280453.0,
      "step": 1495
    },
    {
      "entropy": 6.28340425491333,
      "epoch": 0.14886859865025803,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999722144747125,
      "loss": 6.0847,
      "mean_token_accuracy": 0.11536976397037506,
      "num_tokens": 3290234.0,
      "step": 1500
    },
    {
      "entropy": 6.280434465408325,
      "epoch": 0.14936482731242556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999716548726085,
      "loss": 6.0635,
      "mean_token_accuracy": 0.12484975904226303,
      "num_tokens": 3300943.0,
      "step": 1505
    },
    {
      "entropy": 6.274162673950196,
      "epoch": 0.1498610559745931,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999710896916921,
      "loss": 6.0263,
      "mean_token_accuracy": 0.11883052140474319,
      "num_tokens": 3311296.0,
      "step": 1510
    },
    {
      "entropy": 6.32963080406189,
      "epoch": 0.15035728463676062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999705189319774,
      "loss": 6.1816,
      "mean_token_accuracy": 0.11108686327934265,
      "num_tokens": 3322310.0,
      "step": 1515
    },
    {
      "entropy": 6.295844268798828,
      "epoch": 0.15085351329892815,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999699425934788,
      "loss": 6.1842,
      "mean_token_accuracy": 0.11696541234850884,
      "num_tokens": 3334176.0,
      "step": 1520
    },
    {
      "entropy": 6.382207250595092,
      "epoch": 0.15134974196109569,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999693606762103,
      "loss": 6.1791,
      "mean_token_accuracy": 0.11649519056081772,
      "num_tokens": 3344839.0,
      "step": 1525
    },
    {
      "entropy": 6.323039865493774,
      "epoch": 0.1518459706232632,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999687731801865,
      "loss": 6.1538,
      "mean_token_accuracy": 0.11218968853354454,
      "num_tokens": 3357046.0,
      "step": 1530
    },
    {
      "entropy": 6.362309741973877,
      "epoch": 0.15234219928543072,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999681801054217,
      "loss": 6.2222,
      "mean_token_accuracy": 0.11561866924166679,
      "num_tokens": 3367783.0,
      "step": 1535
    },
    {
      "entropy": 6.31308650970459,
      "epoch": 0.15283842794759825,
      "grad_norm": 1.0,
      "learning_rate": 0.000499967581451931,
      "loss": 6.0902,
      "mean_token_accuracy": 0.11377239897847176,
      "num_tokens": 3378529.0,
      "step": 1540
    },
    {
      "entropy": 6.354143381118774,
      "epoch": 0.15333465660976578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499966977219729,
      "loss": 6.2817,
      "mean_token_accuracy": 0.11187926903367043,
      "num_tokens": 3389318.0,
      "step": 1545
    },
    {
      "entropy": 6.434057950973511,
      "epoch": 0.1538308852719333,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999663674088306,
      "loss": 6.2759,
      "mean_token_accuracy": 0.11047122180461884,
      "num_tokens": 3400603.0,
      "step": 1550
    },
    {
      "entropy": 6.393321371078491,
      "epoch": 0.15432711393410084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999657520192511,
      "loss": 6.1748,
      "mean_token_accuracy": 0.1146856315433979,
      "num_tokens": 3410316.0,
      "step": 1555
    },
    {
      "entropy": 6.320789861679077,
      "epoch": 0.15482334259626837,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999651310510058,
      "loss": 6.1198,
      "mean_token_accuracy": 0.12020543217658997,
      "num_tokens": 3422155.0,
      "step": 1560
    },
    {
      "entropy": 6.328480195999146,
      "epoch": 0.15531957125843587,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999645045041099,
      "loss": 6.0594,
      "mean_token_accuracy": 0.11951424777507783,
      "num_tokens": 3432892.0,
      "step": 1565
    },
    {
      "entropy": 6.272742033004761,
      "epoch": 0.1558157999206034,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999638723785791,
      "loss": 6.1812,
      "mean_token_accuracy": 0.11743228211998939,
      "num_tokens": 3443967.0,
      "step": 1570
    },
    {
      "entropy": 6.284852838516235,
      "epoch": 0.15631202858277093,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499963234674429,
      "loss": 6.069,
      "mean_token_accuracy": 0.11911690831184388,
      "num_tokens": 3454354.0,
      "step": 1575
    },
    {
      "entropy": 6.222921228408813,
      "epoch": 0.15680825724493846,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999625913916754,
      "loss": 6.0872,
      "mean_token_accuracy": 0.11979979053139686,
      "num_tokens": 3465232.0,
      "step": 1580
    },
    {
      "entropy": 6.206456661224365,
      "epoch": 0.157304485907106,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999619425303343,
      "loss": 6.0632,
      "mean_token_accuracy": 0.12077862471342087,
      "num_tokens": 3475443.0,
      "step": 1585
    },
    {
      "entropy": 6.376963663101196,
      "epoch": 0.15780071456927353,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999612880904219,
      "loss": 6.1005,
      "mean_token_accuracy": 0.11797050833702087,
      "num_tokens": 3486643.0,
      "step": 1590
    },
    {
      "entropy": 6.291903114318847,
      "epoch": 0.15829694323144106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999606280719542,
      "loss": 6.1495,
      "mean_token_accuracy": 0.11660699099302292,
      "num_tokens": 3497557.0,
      "step": 1595
    },
    {
      "entropy": 6.405364274978638,
      "epoch": 0.1587931718936086,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999599624749475,
      "loss": 6.1583,
      "mean_token_accuracy": 0.1143525168299675,
      "num_tokens": 3509790.0,
      "step": 1600
    },
    {
      "entropy": 6.194210481643677,
      "epoch": 0.1592894005557761,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999592912994187,
      "loss": 6.077,
      "mean_token_accuracy": 0.11786877140402793,
      "num_tokens": 3521577.0,
      "step": 1605
    },
    {
      "entropy": 6.348882055282592,
      "epoch": 0.15978562921794362,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999586145453842,
      "loss": 6.1151,
      "mean_token_accuracy": 0.11298025399446487,
      "num_tokens": 3532126.0,
      "step": 1610
    },
    {
      "entropy": 6.160671663284302,
      "epoch": 0.16028185788011115,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999579322128608,
      "loss": 5.9537,
      "mean_token_accuracy": 0.11721618697047234,
      "num_tokens": 3542560.0,
      "step": 1615
    },
    {
      "entropy": 6.2272850513458256,
      "epoch": 0.16077808654227868,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999572443018654,
      "loss": 6.1069,
      "mean_token_accuracy": 0.1176838219165802,
      "num_tokens": 3553863.0,
      "step": 1620
    },
    {
      "entropy": 6.241751861572266,
      "epoch": 0.1612743152044462,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999565508124151,
      "loss": 6.1353,
      "mean_token_accuracy": 0.11932577788829804,
      "num_tokens": 3564953.0,
      "step": 1625
    },
    {
      "entropy": 6.313349819183349,
      "epoch": 0.16177054386661374,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000499955851744527,
      "loss": 6.1226,
      "mean_token_accuracy": 0.12102637365460396,
      "num_tokens": 3576247.0,
      "step": 1630
    },
    {
      "entropy": 6.362412357330323,
      "epoch": 0.16226677252878127,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999551470982185,
      "loss": 6.1641,
      "mean_token_accuracy": 0.11501411274075508,
      "num_tokens": 3587600.0,
      "step": 1635
    },
    {
      "entropy": 6.3714934349060055,
      "epoch": 0.1627630011909488,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999544368735073,
      "loss": 6.1295,
      "mean_token_accuracy": 0.12266664281487465,
      "num_tokens": 3597727.0,
      "step": 1640
    },
    {
      "entropy": 6.2838012218475345,
      "epoch": 0.1632592298531163,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999537210704106,
      "loss": 6.1149,
      "mean_token_accuracy": 0.1231493391096592,
      "num_tokens": 3608066.0,
      "step": 1645
    },
    {
      "entropy": 6.3129997730255125,
      "epoch": 0.16375545851528384,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999529996889463,
      "loss": 6.1644,
      "mean_token_accuracy": 0.11525594145059585,
      "num_tokens": 3618263.0,
      "step": 1650
    },
    {
      "entropy": 6.383852910995484,
      "epoch": 0.16425168717745137,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999522727291325,
      "loss": 6.1189,
      "mean_token_accuracy": 0.12352551445364952,
      "num_tokens": 3629465.0,
      "step": 1655
    },
    {
      "entropy": 6.246583700180054,
      "epoch": 0.1647479158396189,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499951540190987,
      "loss": 6.1389,
      "mean_token_accuracy": 0.11551668420433998,
      "num_tokens": 3639657.0,
      "step": 1660
    },
    {
      "entropy": 6.313644647598267,
      "epoch": 0.16524414450178643,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999508020745281,
      "loss": 6.1039,
      "mean_token_accuracy": 0.11597631126642227,
      "num_tokens": 3652374.0,
      "step": 1665
    },
    {
      "entropy": 6.2801806926727295,
      "epoch": 0.16574037316395396,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999500583797739,
      "loss": 6.0802,
      "mean_token_accuracy": 0.11924066990613938,
      "num_tokens": 3664083.0,
      "step": 1670
    },
    {
      "entropy": 6.2540826320648195,
      "epoch": 0.1662366018261215,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999493091067431,
      "loss": 6.1179,
      "mean_token_accuracy": 0.12284742221236229,
      "num_tokens": 3675314.0,
      "step": 1675
    },
    {
      "entropy": 6.218377304077149,
      "epoch": 0.166732830488289,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004999485542554541,
      "loss": 6.081,
      "mean_token_accuracy": 0.11232990324497223,
      "num_tokens": 3686019.0,
      "step": 1680
    },
    {
      "entropy": 6.390085935592651,
      "epoch": 0.16722905915045652,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999477938259256,
      "loss": 6.1781,
      "mean_token_accuracy": 0.11307597011327744,
      "num_tokens": 3698170.0,
      "step": 1685
    },
    {
      "entropy": 6.218471336364746,
      "epoch": 0.16772528781262405,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004999470278181766,
      "loss": 6.1603,
      "mean_token_accuracy": 0.11149280816316605,
      "num_tokens": 3709243.0,
      "step": 1690
    },
    {
      "entropy": 6.361375761032105,
      "epoch": 0.16822151647479158,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999462562322261,
      "loss": 6.1193,
      "mean_token_accuracy": 0.11822729557752609,
      "num_tokens": 3720196.0,
      "step": 1695
    },
    {
      "entropy": 6.330177640914917,
      "epoch": 0.1687177451369591,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499945479068093,
      "loss": 6.1702,
      "mean_token_accuracy": 0.1214469164609909,
      "num_tokens": 3732714.0,
      "step": 1700
    },
    {
      "entropy": 6.278452777862549,
      "epoch": 0.16921397379912664,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004999446963257969,
      "loss": 6.14,
      "mean_token_accuracy": 0.11547001376748085,
      "num_tokens": 3746474.0,
      "step": 1705
    },
    {
      "entropy": 6.218746757507324,
      "epoch": 0.16971020246129417,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999439080053571,
      "loss": 6.0096,
      "mean_token_accuracy": 0.12439236715435982,
      "num_tokens": 3757528.0,
      "step": 1710
    },
    {
      "entropy": 6.3749387741088865,
      "epoch": 0.1702064311234617,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999431141067928,
      "loss": 6.1542,
      "mean_token_accuracy": 0.11521807983517647,
      "num_tokens": 3768360.0,
      "step": 1715
    },
    {
      "entropy": 6.213152885437012,
      "epoch": 0.1707026597856292,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999423146301242,
      "loss": 6.016,
      "mean_token_accuracy": 0.12060007154941559,
      "num_tokens": 3779809.0,
      "step": 1720
    },
    {
      "entropy": 6.188921022415161,
      "epoch": 0.17119888844779674,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999415095753709,
      "loss": 6.0182,
      "mean_token_accuracy": 0.12153950557112694,
      "num_tokens": 3791455.0,
      "step": 1725
    },
    {
      "entropy": 6.190573024749756,
      "epoch": 0.17169511710996427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999406989425527,
      "loss": 6.051,
      "mean_token_accuracy": 0.11963336691260337,
      "num_tokens": 3803209.0,
      "step": 1730
    },
    {
      "entropy": 6.1931469440460205,
      "epoch": 0.1721913457721318,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999398827316901,
      "loss": 6.0878,
      "mean_token_accuracy": 0.11741367131471633,
      "num_tokens": 3813553.0,
      "step": 1735
    },
    {
      "entropy": 6.292501640319824,
      "epoch": 0.17268757443429933,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499939060942803,
      "loss": 6.1305,
      "mean_token_accuracy": 0.1166291169822216,
      "num_tokens": 3824856.0,
      "step": 1740
    },
    {
      "entropy": 6.269052934646607,
      "epoch": 0.17318380309646686,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999382335759119,
      "loss": 6.2027,
      "mean_token_accuracy": 0.11211237907409669,
      "num_tokens": 3835489.0,
      "step": 1745
    },
    {
      "entropy": 6.338535118103027,
      "epoch": 0.1736800317586344,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999374006310372,
      "loss": 6.0367,
      "mean_token_accuracy": 0.11971709057688713,
      "num_tokens": 3844838.0,
      "step": 1750
    },
    {
      "entropy": 6.158118391036988,
      "epoch": 0.1741762604208019,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999365621081999,
      "loss": 6.0031,
      "mean_token_accuracy": 0.11688855215907097,
      "num_tokens": 3855288.0,
      "step": 1755
    },
    {
      "entropy": 6.210778141021729,
      "epoch": 0.17467248908296942,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999357180074204,
      "loss": 6.0566,
      "mean_token_accuracy": 0.12111689299345016,
      "num_tokens": 3865585.0,
      "step": 1760
    },
    {
      "entropy": 6.272298288345337,
      "epoch": 0.17516871774513695,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999348683287199,
      "loss": 6.0413,
      "mean_token_accuracy": 0.11825228407979012,
      "num_tokens": 3877694.0,
      "step": 1765
    },
    {
      "entropy": 6.2135413646697994,
      "epoch": 0.17566494640730448,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004999340130721192,
      "loss": 6.0028,
      "mean_token_accuracy": 0.13018031790852547,
      "num_tokens": 3888802.0,
      "step": 1770
    },
    {
      "entropy": 6.1698943138122555,
      "epoch": 0.17616117506947201,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999331522376396,
      "loss": 5.9874,
      "mean_token_accuracy": 0.12037508189678192,
      "num_tokens": 3900294.0,
      "step": 1775
    },
    {
      "entropy": 6.191412830352784,
      "epoch": 0.17665740373163955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999322858253027,
      "loss": 6.0106,
      "mean_token_accuracy": 0.12115705609321595,
      "num_tokens": 3910865.0,
      "step": 1780
    },
    {
      "entropy": 6.2171179294586185,
      "epoch": 0.17715363239380708,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999314138351297,
      "loss": 6.0731,
      "mean_token_accuracy": 0.11963929235935211,
      "num_tokens": 3921960.0,
      "step": 1785
    },
    {
      "entropy": 6.269604063034057,
      "epoch": 0.1776498610559746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999305362671424,
      "loss": 6.0819,
      "mean_token_accuracy": 0.11684257239103317,
      "num_tokens": 3934085.0,
      "step": 1790
    },
    {
      "entropy": 6.16135573387146,
      "epoch": 0.1781460897181421,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999296531213624,
      "loss": 5.9371,
      "mean_token_accuracy": 0.1304635725915432,
      "num_tokens": 3944532.0,
      "step": 1795
    },
    {
      "entropy": 6.181241798400879,
      "epoch": 0.17864231838030964,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999287643978117,
      "loss": 5.9725,
      "mean_token_accuracy": 0.12921013534069062,
      "num_tokens": 3954995.0,
      "step": 1800
    },
    {
      "entropy": 6.115298318862915,
      "epoch": 0.17913854704247717,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999278700965124,
      "loss": 6.0287,
      "mean_token_accuracy": 0.12031830176711082,
      "num_tokens": 3966621.0,
      "step": 1805
    },
    {
      "entropy": 6.2137250900268555,
      "epoch": 0.1796347757046447,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999269702174866,
      "loss": 6.0895,
      "mean_token_accuracy": 0.12122552394866944,
      "num_tokens": 3977568.0,
      "step": 1810
    },
    {
      "entropy": 6.222836112976074,
      "epoch": 0.18013100436681223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999260647607565,
      "loss": 6.1064,
      "mean_token_accuracy": 0.11863134428858757,
      "num_tokens": 3988571.0,
      "step": 1815
    },
    {
      "entropy": 6.24027271270752,
      "epoch": 0.18062723302897976,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999251537263448,
      "loss": 5.9314,
      "mean_token_accuracy": 0.12621500641107558,
      "num_tokens": 3998380.0,
      "step": 1820
    },
    {
      "entropy": 6.1338170051574705,
      "epoch": 0.1811234616911473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999242371142739,
      "loss": 6.1195,
      "mean_token_accuracy": 0.12032051682472229,
      "num_tokens": 4010017.0,
      "step": 1825
    },
    {
      "entropy": 6.292723035812378,
      "epoch": 0.1816196903533148,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999233149245665,
      "loss": 6.0814,
      "mean_token_accuracy": 0.11958788856863975,
      "num_tokens": 4020922.0,
      "step": 1830
    },
    {
      "entropy": 6.293724489212036,
      "epoch": 0.18211591901548232,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999223871572456,
      "loss": 6.0459,
      "mean_token_accuracy": 0.11999811977148056,
      "num_tokens": 4030927.0,
      "step": 1835
    },
    {
      "entropy": 6.2004194259643555,
      "epoch": 0.18261214767764986,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999214538123342,
      "loss": 6.122,
      "mean_token_accuracy": 0.11295817345380783,
      "num_tokens": 4041375.0,
      "step": 1840
    },
    {
      "entropy": 6.214535427093506,
      "epoch": 0.18310837633981739,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999205148898554,
      "loss": 6.1229,
      "mean_token_accuracy": 0.11756806150078773,
      "num_tokens": 4053213.0,
      "step": 1845
    },
    {
      "entropy": 6.2128990650177,
      "epoch": 0.18360460500198492,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999195703898325,
      "loss": 5.9532,
      "mean_token_accuracy": 0.12472454011440277,
      "num_tokens": 4064536.0,
      "step": 1850
    },
    {
      "entropy": 6.201371765136718,
      "epoch": 0.18410083366415245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999186203122889,
      "loss": 6.0295,
      "mean_token_accuracy": 0.12741652950644494,
      "num_tokens": 4075282.0,
      "step": 1855
    },
    {
      "entropy": 6.092880630493164,
      "epoch": 0.18459706232631998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999176646572482,
      "loss": 6.0337,
      "mean_token_accuracy": 0.12471091002225876,
      "num_tokens": 4085501.0,
      "step": 1860
    },
    {
      "entropy": 6.233616590499878,
      "epoch": 0.1850932909884875,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499916703424734,
      "loss": 5.9813,
      "mean_token_accuracy": 0.11598324999213219,
      "num_tokens": 4095697.0,
      "step": 1865
    },
    {
      "entropy": 6.167621278762818,
      "epoch": 0.185589519650655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999157366147702,
      "loss": 6.0469,
      "mean_token_accuracy": 0.12036259993910789,
      "num_tokens": 4106459.0,
      "step": 1870
    },
    {
      "entropy": 6.21561450958252,
      "epoch": 0.18608574831282254,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999147642273809,
      "loss": 5.9731,
      "mean_token_accuracy": 0.12678501680493354,
      "num_tokens": 4118320.0,
      "step": 1875
    },
    {
      "entropy": 6.1109779357910154,
      "epoch": 0.18658197697499007,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049991378626259,
      "loss": 5.995,
      "mean_token_accuracy": 0.13212494775652886,
      "num_tokens": 4131249.0,
      "step": 1880
    },
    {
      "entropy": 6.219920682907104,
      "epoch": 0.1870782056371576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999128027204219,
      "loss": 6.0718,
      "mean_token_accuracy": 0.12198005616664886,
      "num_tokens": 4141480.0,
      "step": 1885
    },
    {
      "entropy": 6.236648035049439,
      "epoch": 0.18757443429932513,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499911813600901,
      "loss": 6.1177,
      "mean_token_accuracy": 0.11755191907286644,
      "num_tokens": 4152344.0,
      "step": 1890
    },
    {
      "entropy": 6.261018133163452,
      "epoch": 0.18807066296149266,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999108189040516,
      "loss": 6.0806,
      "mean_token_accuracy": 0.11942842453718186,
      "num_tokens": 4164772.0,
      "step": 1895
    },
    {
      "entropy": 6.152463722229004,
      "epoch": 0.1885668916236602,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999098186298986,
      "loss": 6.0359,
      "mean_token_accuracy": 0.11918087005615234,
      "num_tokens": 4175511.0,
      "step": 1900
    },
    {
      "entropy": 6.117107248306274,
      "epoch": 0.1890631202858277,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999088127784667,
      "loss": 6.0012,
      "mean_token_accuracy": 0.12256012931466102,
      "num_tokens": 4187445.0,
      "step": 1905
    },
    {
      "entropy": 6.237027502059936,
      "epoch": 0.18955934894799523,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499907801349781,
      "loss": 5.9748,
      "mean_token_accuracy": 0.1233195923268795,
      "num_tokens": 4198124.0,
      "step": 1910
    },
    {
      "entropy": 6.142972564697265,
      "epoch": 0.19005557761016276,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999067843438664,
      "loss": 6.0176,
      "mean_token_accuracy": 0.12327835708856583,
      "num_tokens": 4208954.0,
      "step": 1915
    },
    {
      "entropy": 6.2328629970550535,
      "epoch": 0.1905518062723303,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999057617607482,
      "loss": 6.0782,
      "mean_token_accuracy": 0.12020684257149697,
      "num_tokens": 4218896.0,
      "step": 1920
    },
    {
      "entropy": 6.11851167678833,
      "epoch": 0.19104803493449782,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999047336004517,
      "loss": 5.9858,
      "mean_token_accuracy": 0.11666620373725892,
      "num_tokens": 4228851.0,
      "step": 1925
    },
    {
      "entropy": 6.121028709411621,
      "epoch": 0.19154426359666535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999036998630024,
      "loss": 5.9408,
      "mean_token_accuracy": 0.12234764248132705,
      "num_tokens": 4240377.0,
      "step": 1930
    },
    {
      "entropy": 6.136601066589355,
      "epoch": 0.19204049225883288,
      "grad_norm": 0.875,
      "learning_rate": 0.000499902660548426,
      "loss": 5.9324,
      "mean_token_accuracy": 0.12323587760329247,
      "num_tokens": 4251341.0,
      "step": 1935
    },
    {
      "entropy": 6.131585884094238,
      "epoch": 0.1925367209210004,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999016156567482,
      "loss": 6.039,
      "mean_token_accuracy": 0.12415843605995178,
      "num_tokens": 4261206.0,
      "step": 1940
    },
    {
      "entropy": 6.232227182388305,
      "epoch": 0.1930329495831679,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499900565187995,
      "loss": 6.0428,
      "mean_token_accuracy": 0.11449379473924637,
      "num_tokens": 4272662.0,
      "step": 1945
    },
    {
      "entropy": 6.217731094360351,
      "epoch": 0.19352917824533544,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998995091421924,
      "loss": 6.0248,
      "mean_token_accuracy": 0.12198412641882897,
      "num_tokens": 4283302.0,
      "step": 1950
    },
    {
      "entropy": 6.128766965866089,
      "epoch": 0.19402540690750297,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998984475193665,
      "loss": 5.9794,
      "mean_token_accuracy": 0.12389139384031296,
      "num_tokens": 4296087.0,
      "step": 1955
    },
    {
      "entropy": 6.144579887390137,
      "epoch": 0.1945216355696705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998973803195439,
      "loss": 5.9912,
      "mean_token_accuracy": 0.12408131062984466,
      "num_tokens": 4305918.0,
      "step": 1960
    },
    {
      "entropy": 6.167060136795044,
      "epoch": 0.19501786423183803,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998963075427508,
      "loss": 6.0475,
      "mean_token_accuracy": 0.12526998519897461,
      "num_tokens": 4316507.0,
      "step": 1965
    },
    {
      "entropy": 6.1785789966583256,
      "epoch": 0.19551409289400556,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998952291890138,
      "loss": 5.9762,
      "mean_token_accuracy": 0.12871691659092904,
      "num_tokens": 4329091.0,
      "step": 1970
    },
    {
      "entropy": 6.164827680587768,
      "epoch": 0.1960103215561731,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998941452583599,
      "loss": 6.0421,
      "mean_token_accuracy": 0.12194729819893838,
      "num_tokens": 4339266.0,
      "step": 1975
    },
    {
      "entropy": 6.106686210632324,
      "epoch": 0.1965065502183406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998930557508156,
      "loss": 5.9713,
      "mean_token_accuracy": 0.12659700810909272,
      "num_tokens": 4349300.0,
      "step": 1980
    },
    {
      "entropy": 6.051992893218994,
      "epoch": 0.19700277888050813,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998919606664082,
      "loss": 5.8926,
      "mean_token_accuracy": 0.13368902131915092,
      "num_tokens": 4360314.0,
      "step": 1985
    },
    {
      "entropy": 6.151434469223022,
      "epoch": 0.19749900754267566,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998908600051648,
      "loss": 5.9413,
      "mean_token_accuracy": 0.12665844783186914,
      "num_tokens": 4370077.0,
      "step": 1990
    },
    {
      "entropy": 6.119269704818725,
      "epoch": 0.1979952362048432,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998897537671126,
      "loss": 5.951,
      "mean_token_accuracy": 0.12383900433778763,
      "num_tokens": 4381000.0,
      "step": 1995
    },
    {
      "entropy": 6.143065547943115,
      "epoch": 0.19849146486701072,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998886419522792,
      "loss": 6.1605,
      "mean_token_accuracy": 0.11577931866049766,
      "num_tokens": 4393485.0,
      "step": 2000
    },
    {
      "entropy": 6.28572907447815,
      "epoch": 0.19898769352917825,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998875245606919,
      "loss": 5.9903,
      "mean_token_accuracy": 0.12323976159095765,
      "num_tokens": 4404354.0,
      "step": 2005
    },
    {
      "entropy": 6.138240671157837,
      "epoch": 0.19948392219134578,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998864015923788,
      "loss": 6.1308,
      "mean_token_accuracy": 0.12045308947563171,
      "num_tokens": 4414818.0,
      "step": 2010
    },
    {
      "entropy": 6.158795976638794,
      "epoch": 0.1999801508535133,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998852730473674,
      "loss": 6.0079,
      "mean_token_accuracy": 0.11828904002904891,
      "num_tokens": 4426271.0,
      "step": 2015
    },
    {
      "entropy": 6.15863299369812,
      "epoch": 0.2004763795156808,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998841389256859,
      "loss": 6.0545,
      "mean_token_accuracy": 0.12046427428722381,
      "num_tokens": 4436981.0,
      "step": 2020
    },
    {
      "entropy": 6.142259788513184,
      "epoch": 0.20097260817784834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998829992273622,
      "loss": 5.8818,
      "mean_token_accuracy": 0.12613770440220834,
      "num_tokens": 4446494.0,
      "step": 2025
    },
    {
      "entropy": 6.0647642612457275,
      "epoch": 0.20146883684001587,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998818539524248,
      "loss": 5.9738,
      "mean_token_accuracy": 0.1264504000544548,
      "num_tokens": 4458535.0,
      "step": 2030
    },
    {
      "entropy": 6.1495154857635494,
      "epoch": 0.2019650655021834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499880703100902,
      "loss": 5.9488,
      "mean_token_accuracy": 0.1266962394118309,
      "num_tokens": 4469025.0,
      "step": 2035
    },
    {
      "entropy": 6.134205341339111,
      "epoch": 0.20246129416435094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998795466728223,
      "loss": 5.9595,
      "mean_token_accuracy": 0.12471545413136483,
      "num_tokens": 4478431.0,
      "step": 2040
    },
    {
      "entropy": 6.171464109420777,
      "epoch": 0.20295752282651847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998783846682144,
      "loss": 6.0815,
      "mean_token_accuracy": 0.12394077926874161,
      "num_tokens": 4490215.0,
      "step": 2045
    },
    {
      "entropy": 6.105607700347901,
      "epoch": 0.203453751488686,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004998772170871072,
      "loss": 5.963,
      "mean_token_accuracy": 0.1244163066148758,
      "num_tokens": 4502895.0,
      "step": 2050
    },
    {
      "entropy": 6.12325873374939,
      "epoch": 0.20394998015085353,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998760439295295,
      "loss": 5.9871,
      "mean_token_accuracy": 0.12174513414502144,
      "num_tokens": 4514468.0,
      "step": 2055
    },
    {
      "entropy": 6.173321723937988,
      "epoch": 0.20444620881302103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998748651955105,
      "loss": 6.007,
      "mean_token_accuracy": 0.12350882440805436,
      "num_tokens": 4525787.0,
      "step": 2060
    },
    {
      "entropy": 6.1624654769897464,
      "epoch": 0.20494243747518856,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004998736808850795,
      "loss": 6.0868,
      "mean_token_accuracy": 0.11808518916368485,
      "num_tokens": 4537536.0,
      "step": 2065
    },
    {
      "entropy": 6.173154306411743,
      "epoch": 0.2054386661373561,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998724909982656,
      "loss": 6.0371,
      "mean_token_accuracy": 0.12236888855695724,
      "num_tokens": 4549619.0,
      "step": 2070
    },
    {
      "entropy": 6.107865715026856,
      "epoch": 0.20593489479952362,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998712955350986,
      "loss": 5.9509,
      "mean_token_accuracy": 0.12667887806892394,
      "num_tokens": 4560248.0,
      "step": 2075
    },
    {
      "entropy": 6.182792615890503,
      "epoch": 0.20643112346169115,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004998700944956081,
      "loss": 6.0598,
      "mean_token_accuracy": 0.12352480441331863,
      "num_tokens": 4571888.0,
      "step": 2080
    },
    {
      "entropy": 6.149948072433472,
      "epoch": 0.20692735212385868,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998688878798237,
      "loss": 5.9082,
      "mean_token_accuracy": 0.13024957329034806,
      "num_tokens": 4583234.0,
      "step": 2085
    },
    {
      "entropy": 6.173293352127075,
      "epoch": 0.2074235807860262,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998676756877753,
      "loss": 6.1228,
      "mean_token_accuracy": 0.11778421998023987,
      "num_tokens": 4595252.0,
      "step": 2090
    },
    {
      "entropy": 6.100701141357422,
      "epoch": 0.20791980944819372,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998664579194934,
      "loss": 5.9016,
      "mean_token_accuracy": 0.12487494945526123,
      "num_tokens": 4607091.0,
      "step": 2095
    },
    {
      "entropy": 6.117470121383667,
      "epoch": 0.20841603811036125,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998652345750077,
      "loss": 5.8611,
      "mean_token_accuracy": 0.12737861275672913,
      "num_tokens": 4618003.0,
      "step": 2100
    },
    {
      "entropy": 6.059277963638306,
      "epoch": 0.20891226677252878,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998640056543487,
      "loss": 5.8767,
      "mean_token_accuracy": 0.1214037574827671,
      "num_tokens": 4628429.0,
      "step": 2105
    },
    {
      "entropy": 6.125901508331299,
      "epoch": 0.2094084954346963,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998627711575469,
      "loss": 5.9361,
      "mean_token_accuracy": 0.13086853697896003,
      "num_tokens": 4639414.0,
      "step": 2110
    },
    {
      "entropy": 6.141415405273437,
      "epoch": 0.20990472409686384,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000499861531084633,
      "loss": 6.022,
      "mean_token_accuracy": 0.12312455773353577,
      "num_tokens": 4652594.0,
      "step": 2115
    },
    {
      "entropy": 6.039756822586059,
      "epoch": 0.21040095275903137,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998602854356376,
      "loss": 5.8612,
      "mean_token_accuracy": 0.12929336726665497,
      "num_tokens": 4662981.0,
      "step": 2120
    },
    {
      "entropy": 6.109541177749634,
      "epoch": 0.2108971814211989,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998590342105917,
      "loss": 5.9639,
      "mean_token_accuracy": 0.1239814892411232,
      "num_tokens": 4672469.0,
      "step": 2125
    },
    {
      "entropy": 6.082083654403687,
      "epoch": 0.21139341008336643,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998577774095262,
      "loss": 5.9456,
      "mean_token_accuracy": 0.12228829637169839,
      "num_tokens": 4683640.0,
      "step": 2130
    },
    {
      "entropy": 6.100391864776611,
      "epoch": 0.21188963874553393,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998565150324724,
      "loss": 5.8938,
      "mean_token_accuracy": 0.12874594032764436,
      "num_tokens": 4694032.0,
      "step": 2135
    },
    {
      "entropy": 5.9822530269622805,
      "epoch": 0.21238586740770146,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998552470794616,
      "loss": 5.7792,
      "mean_token_accuracy": 0.1315452888607979,
      "num_tokens": 4704253.0,
      "step": 2140
    },
    {
      "entropy": 6.03460578918457,
      "epoch": 0.212882096069869,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004998539735505251,
      "loss": 5.9305,
      "mean_token_accuracy": 0.1236830122768879,
      "num_tokens": 4715327.0,
      "step": 2145
    },
    {
      "entropy": 6.228457880020142,
      "epoch": 0.21337832473203652,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998526944456946,
      "loss": 5.9449,
      "mean_token_accuracy": 0.12152760252356529,
      "num_tokens": 4726450.0,
      "step": 2150
    },
    {
      "entropy": 6.073284006118774,
      "epoch": 0.21387455339420405,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998514097650018,
      "loss": 5.9736,
      "mean_token_accuracy": 0.12102751210331916,
      "num_tokens": 4736630.0,
      "step": 2155
    },
    {
      "entropy": 6.08151159286499,
      "epoch": 0.21437078205637158,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998501195084786,
      "loss": 5.8635,
      "mean_token_accuracy": 0.12787755504250525,
      "num_tokens": 4746592.0,
      "step": 2160
    },
    {
      "entropy": 6.100288772583008,
      "epoch": 0.21486701071853911,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998488236761569,
      "loss": 6.0417,
      "mean_token_accuracy": 0.12549156099557876,
      "num_tokens": 4757824.0,
      "step": 2165
    },
    {
      "entropy": 6.139098644256592,
      "epoch": 0.21536323938070662,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998475222680688,
      "loss": 5.9766,
      "mean_token_accuracy": 0.12785990089178084,
      "num_tokens": 4768453.0,
      "step": 2170
    },
    {
      "entropy": 5.9953094005584715,
      "epoch": 0.21585946804287415,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998462152842468,
      "loss": 5.956,
      "mean_token_accuracy": 0.1242900937795639,
      "num_tokens": 4779664.0,
      "step": 2175
    },
    {
      "entropy": 6.162226676940918,
      "epoch": 0.21635569670504168,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499844902724723,
      "loss": 5.969,
      "mean_token_accuracy": 0.1265193648636341,
      "num_tokens": 4791012.0,
      "step": 2180
    },
    {
      "entropy": 6.099493169784546,
      "epoch": 0.2168519253672092,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998435845895302,
      "loss": 5.841,
      "mean_token_accuracy": 0.1277057059109211,
      "num_tokens": 4800578.0,
      "step": 2185
    },
    {
      "entropy": 6.023715972900391,
      "epoch": 0.21734815402937674,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499842260878701,
      "loss": 5.8514,
      "mean_token_accuracy": 0.12200021669268608,
      "num_tokens": 4810940.0,
      "step": 2190
    },
    {
      "entropy": 6.081690168380737,
      "epoch": 0.21784438269154427,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998409315922683,
      "loss": 5.9947,
      "mean_token_accuracy": 0.12390753328800201,
      "num_tokens": 4822696.0,
      "step": 2195
    },
    {
      "entropy": 6.113512086868286,
      "epoch": 0.2183406113537118,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998395967302649,
      "loss": 5.8774,
      "mean_token_accuracy": 0.13284592926502228,
      "num_tokens": 4833669.0,
      "step": 2200
    },
    {
      "entropy": 6.167998266220093,
      "epoch": 0.21883684001587933,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499838256292724,
      "loss": 6.0529,
      "mean_token_accuracy": 0.12145362943410873,
      "num_tokens": 4844021.0,
      "step": 2205
    },
    {
      "entropy": 5.9783563137054445,
      "epoch": 0.21933306867804683,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998369102796787,
      "loss": 5.9243,
      "mean_token_accuracy": 0.12736267521977424,
      "num_tokens": 4855336.0,
      "step": 2210
    },
    {
      "entropy": 6.101631212234497,
      "epoch": 0.21982929734021436,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004998355586911627,
      "loss": 5.9299,
      "mean_token_accuracy": 0.1268913999199867,
      "num_tokens": 4867713.0,
      "step": 2215
    },
    {
      "entropy": 6.054467105865479,
      "epoch": 0.2203255260023819,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004998342015272093,
      "loss": 5.9026,
      "mean_token_accuracy": 0.12375652343034745,
      "num_tokens": 4879134.0,
      "step": 2220
    },
    {
      "entropy": 6.114092779159546,
      "epoch": 0.22082175466454942,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998328387878521,
      "loss": 6.0599,
      "mean_token_accuracy": 0.12487743720412255,
      "num_tokens": 4891755.0,
      "step": 2225
    },
    {
      "entropy": 6.203889560699463,
      "epoch": 0.22131798332671695,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998314704731249,
      "loss": 6.0604,
      "mean_token_accuracy": 0.11887223795056343,
      "num_tokens": 4903264.0,
      "step": 2230
    },
    {
      "entropy": 6.066045093536377,
      "epoch": 0.22181421198888449,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998300965830618,
      "loss": 5.8843,
      "mean_token_accuracy": 0.12301613837480545,
      "num_tokens": 4912533.0,
      "step": 2235
    },
    {
      "entropy": 5.971320104598999,
      "epoch": 0.22231044065105202,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998287171176968,
      "loss": 5.876,
      "mean_token_accuracy": 0.12986253574490547,
      "num_tokens": 4923288.0,
      "step": 2240
    },
    {
      "entropy": 6.049497413635254,
      "epoch": 0.22280666931321952,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998273320770641,
      "loss": 5.8896,
      "mean_token_accuracy": 0.12646428495645523,
      "num_tokens": 4934759.0,
      "step": 2245
    },
    {
      "entropy": 6.087690019607544,
      "epoch": 0.22330289797538705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499825941461198,
      "loss": 5.9523,
      "mean_token_accuracy": 0.12756217569112777,
      "num_tokens": 4947024.0,
      "step": 2250
    },
    {
      "entropy": 6.019892168045044,
      "epoch": 0.22379912663755458,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998245452701329,
      "loss": 5.8962,
      "mean_token_accuracy": 0.1269862487912178,
      "num_tokens": 4959281.0,
      "step": 2255
    },
    {
      "entropy": 6.124982738494873,
      "epoch": 0.2242953552997221,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998231435039036,
      "loss": 5.9385,
      "mean_token_accuracy": 0.11895076110959053,
      "num_tokens": 4970466.0,
      "step": 2260
    },
    {
      "entropy": 6.050164318084716,
      "epoch": 0.22479158396188964,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998217361625449,
      "loss": 5.8953,
      "mean_token_accuracy": 0.13044259697198868,
      "num_tokens": 4981135.0,
      "step": 2265
    },
    {
      "entropy": 6.087223243713379,
      "epoch": 0.22528781262405717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998203232460915,
      "loss": 5.9574,
      "mean_token_accuracy": 0.13070524409413337,
      "num_tokens": 4991292.0,
      "step": 2270
    },
    {
      "entropy": 6.04678053855896,
      "epoch": 0.2257840412862247,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998189047545785,
      "loss": 5.9151,
      "mean_token_accuracy": 0.12900153622031213,
      "num_tokens": 5002806.0,
      "step": 2275
    },
    {
      "entropy": 6.015051984786988,
      "epoch": 0.22628026994839223,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998174806880411,
      "loss": 5.8519,
      "mean_token_accuracy": 0.1365997314453125,
      "num_tokens": 5013474.0,
      "step": 2280
    },
    {
      "entropy": 6.111403083801269,
      "epoch": 0.22677649861055973,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998160510465146,
      "loss": 5.895,
      "mean_token_accuracy": 0.13188145235180854,
      "num_tokens": 5025151.0,
      "step": 2285
    },
    {
      "entropy": 5.943822288513184,
      "epoch": 0.22727272727272727,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004998146158300345,
      "loss": 5.8967,
      "mean_token_accuracy": 0.13392265886068344,
      "num_tokens": 5037166.0,
      "step": 2290
    },
    {
      "entropy": 6.117561721801758,
      "epoch": 0.2277689559348948,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998131750386364,
      "loss": 5.9386,
      "mean_token_accuracy": 0.12657614946365356,
      "num_tokens": 5047452.0,
      "step": 2295
    },
    {
      "entropy": 6.033935308456421,
      "epoch": 0.22826518459706233,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998117286723559,
      "loss": 5.9561,
      "mean_token_accuracy": 0.1220039039850235,
      "num_tokens": 5058501.0,
      "step": 2300
    },
    {
      "entropy": 6.084085893630982,
      "epoch": 0.22876141325922986,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499810276731229,
      "loss": 5.9417,
      "mean_token_accuracy": 0.12538052052259446,
      "num_tokens": 5069716.0,
      "step": 2305
    },
    {
      "entropy": 6.188976240158081,
      "epoch": 0.2292576419213974,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998088192152918,
      "loss": 5.9239,
      "mean_token_accuracy": 0.12268051952123642,
      "num_tokens": 5078509.0,
      "step": 2310
    },
    {
      "entropy": 6.029656267166137,
      "epoch": 0.22975387058356492,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998073561245801,
      "loss": 5.9546,
      "mean_token_accuracy": 0.12687286511063575,
      "num_tokens": 5090355.0,
      "step": 2315
    },
    {
      "entropy": 6.091899156570435,
      "epoch": 0.23025009924573242,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998058874591305,
      "loss": 5.8536,
      "mean_token_accuracy": 0.13038059920072556,
      "num_tokens": 5100582.0,
      "step": 2320
    },
    {
      "entropy": 6.0615569114685055,
      "epoch": 0.23074632790789995,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004998044132189793,
      "loss": 5.9184,
      "mean_token_accuracy": 0.12220206931233406,
      "num_tokens": 5112176.0,
      "step": 2325
    },
    {
      "entropy": 6.061944389343262,
      "epoch": 0.23124255657006748,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499802933404163,
      "loss": 5.8503,
      "mean_token_accuracy": 0.12976043820381164,
      "num_tokens": 5122840.0,
      "step": 2330
    },
    {
      "entropy": 6.06622462272644,
      "epoch": 0.231738785232235,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998014480147183,
      "loss": 5.83,
      "mean_token_accuracy": 0.1355508543550968,
      "num_tokens": 5132926.0,
      "step": 2335
    },
    {
      "entropy": 5.981128120422364,
      "epoch": 0.23223501389440254,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997999570506823,
      "loss": 5.8163,
      "mean_token_accuracy": 0.12751471549272536,
      "num_tokens": 5143394.0,
      "step": 2340
    },
    {
      "entropy": 6.014122343063354,
      "epoch": 0.23273124255657007,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004997984605120917,
      "loss": 5.8538,
      "mean_token_accuracy": 0.12407195791602135,
      "num_tokens": 5155710.0,
      "step": 2345
    },
    {
      "entropy": 5.8755535125732425,
      "epoch": 0.2332274712187376,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997969583989835,
      "loss": 5.7797,
      "mean_token_accuracy": 0.12736281827092172,
      "num_tokens": 5167065.0,
      "step": 2350
    },
    {
      "entropy": 6.026074457168579,
      "epoch": 0.23372369988090513,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997954507113952,
      "loss": 5.875,
      "mean_token_accuracy": 0.12745090574026108,
      "num_tokens": 5177535.0,
      "step": 2355
    },
    {
      "entropy": 6.040788745880127,
      "epoch": 0.23421992854307264,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997939374493642,
      "loss": 5.8063,
      "mean_token_accuracy": 0.13261814862489701,
      "num_tokens": 5188324.0,
      "step": 2360
    },
    {
      "entropy": 5.968220281600952,
      "epoch": 0.23471615720524017,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499792418612928,
      "loss": 5.8842,
      "mean_token_accuracy": 0.1319134347140789,
      "num_tokens": 5199223.0,
      "step": 2365
    },
    {
      "entropy": 6.008479738235474,
      "epoch": 0.2352123858674077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997908942021241,
      "loss": 5.8617,
      "mean_token_accuracy": 0.13253337219357492,
      "num_tokens": 5210080.0,
      "step": 2370
    },
    {
      "entropy": 6.121827793121338,
      "epoch": 0.23570861452957523,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997893642169904,
      "loss": 5.9825,
      "mean_token_accuracy": 0.12802610248327256,
      "num_tokens": 5220867.0,
      "step": 2375
    },
    {
      "entropy": 6.06226658821106,
      "epoch": 0.23620484319174276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997878286575648,
      "loss": 5.8316,
      "mean_token_accuracy": 0.1282677799463272,
      "num_tokens": 5231451.0,
      "step": 2380
    },
    {
      "entropy": 5.996644639968872,
      "epoch": 0.2367010718539103,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997862875238855,
      "loss": 5.9339,
      "mean_token_accuracy": 0.12390372604131698,
      "num_tokens": 5244643.0,
      "step": 2385
    },
    {
      "entropy": 6.055671977996826,
      "epoch": 0.23719730051607782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997847408159907,
      "loss": 5.9503,
      "mean_token_accuracy": 0.12081070691347122,
      "num_tokens": 5256279.0,
      "step": 2390
    },
    {
      "entropy": 6.094895601272583,
      "epoch": 0.23769352917824532,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997831885339186,
      "loss": 5.9201,
      "mean_token_accuracy": 0.12661861553788184,
      "num_tokens": 5268498.0,
      "step": 2395
    },
    {
      "entropy": 6.0079608917236325,
      "epoch": 0.23818975784041285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997816306777077,
      "loss": 5.8873,
      "mean_token_accuracy": 0.13006770759820938,
      "num_tokens": 5280869.0,
      "step": 2400
    },
    {
      "entropy": 6.153380107879639,
      "epoch": 0.23868598650258038,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997800672473969,
      "loss": 5.9915,
      "mean_token_accuracy": 0.1220635175704956,
      "num_tokens": 5293082.0,
      "step": 2405
    },
    {
      "entropy": 6.005958080291748,
      "epoch": 0.2391822151647479,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004997784982430247,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1395465165376663,
      "num_tokens": 5304039.0,
      "step": 2410
    },
    {
      "entropy": 5.969437026977539,
      "epoch": 0.23967844382691544,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997769236646301,
      "loss": 5.8503,
      "mean_token_accuracy": 0.13086041882634164,
      "num_tokens": 5315016.0,
      "step": 2415
    },
    {
      "entropy": 6.144852113723755,
      "epoch": 0.24017467248908297,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997753435122522,
      "loss": 5.9808,
      "mean_token_accuracy": 0.12935518026351928,
      "num_tokens": 5326090.0,
      "step": 2420
    },
    {
      "entropy": 5.986780261993408,
      "epoch": 0.2406709011512505,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049977375778593,
      "loss": 5.778,
      "mean_token_accuracy": 0.12515551522374152,
      "num_tokens": 5336423.0,
      "step": 2425
    },
    {
      "entropy": 6.017197036743164,
      "epoch": 0.24116712981341804,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997721664857032,
      "loss": 5.9411,
      "mean_token_accuracy": 0.1286918543279171,
      "num_tokens": 5346982.0,
      "step": 2430
    },
    {
      "entropy": 6.0123132228851315,
      "epoch": 0.24166335847558554,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997705696116108,
      "loss": 5.829,
      "mean_token_accuracy": 0.13723039478063584,
      "num_tokens": 5359210.0,
      "step": 2435
    },
    {
      "entropy": 6.017412328720093,
      "epoch": 0.24215958713775307,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997689671636926,
      "loss": 5.8603,
      "mean_token_accuracy": 0.12343929260969162,
      "num_tokens": 5369891.0,
      "step": 2440
    },
    {
      "entropy": 6.015021753311157,
      "epoch": 0.2426558157999206,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004997673591419884,
      "loss": 5.8397,
      "mean_token_accuracy": 0.12764633148908616,
      "num_tokens": 5380850.0,
      "step": 2445
    },
    {
      "entropy": 6.030062103271485,
      "epoch": 0.24315204446208813,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004997657455465381,
      "loss": 5.8985,
      "mean_token_accuracy": 0.12984421476721764,
      "num_tokens": 5391903.0,
      "step": 2450
    },
    {
      "entropy": 6.048585891723633,
      "epoch": 0.24364827312425566,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997641263773815,
      "loss": 5.9466,
      "mean_token_accuracy": 0.12775701135396958,
      "num_tokens": 5402517.0,
      "step": 2455
    },
    {
      "entropy": 6.002290296554565,
      "epoch": 0.2441445017864232,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499762501634559,
      "loss": 5.8711,
      "mean_token_accuracy": 0.128414898365736,
      "num_tokens": 5414110.0,
      "step": 2460
    },
    {
      "entropy": 6.061224746704101,
      "epoch": 0.24464073044859072,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997608713181108,
      "loss": 5.8641,
      "mean_token_accuracy": 0.13066134005784988,
      "num_tokens": 5424466.0,
      "step": 2465
    },
    {
      "entropy": 5.991046953201294,
      "epoch": 0.24513695911075822,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997592354280772,
      "loss": 5.8833,
      "mean_token_accuracy": 0.1335628919303417,
      "num_tokens": 5434461.0,
      "step": 2470
    },
    {
      "entropy": 5.993779897689819,
      "epoch": 0.24563318777292575,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499757593964499,
      "loss": 5.8595,
      "mean_token_accuracy": 0.12100402787327766,
      "num_tokens": 5445921.0,
      "step": 2475
    },
    {
      "entropy": 5.966112470626831,
      "epoch": 0.24612941643509328,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997559469274166,
      "loss": 5.8419,
      "mean_token_accuracy": 0.12944890931248665,
      "num_tokens": 5456283.0,
      "step": 2480
    },
    {
      "entropy": 6.00761137008667,
      "epoch": 0.24662564509726081,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997542943168712,
      "loss": 5.8772,
      "mean_token_accuracy": 0.12770011499524117,
      "num_tokens": 5467616.0,
      "step": 2485
    },
    {
      "entropy": 5.989387559890747,
      "epoch": 0.24712187375942835,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997526361329035,
      "loss": 5.8305,
      "mean_token_accuracy": 0.13280585035681725,
      "num_tokens": 5478050.0,
      "step": 2490
    },
    {
      "entropy": 6.101069402694702,
      "epoch": 0.24761810242159588,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997509723755547,
      "loss": 5.8426,
      "mean_token_accuracy": 0.13170866221189498,
      "num_tokens": 5487185.0,
      "step": 2495
    },
    {
      "entropy": 5.979424858093262,
      "epoch": 0.2481143310837634,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997493030448662,
      "loss": 5.8727,
      "mean_token_accuracy": 0.1261827602982521,
      "num_tokens": 5498802.0,
      "step": 2500
    },
    {
      "entropy": 5.974308633804322,
      "epoch": 0.24861055974593094,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997476281408793,
      "loss": 5.8782,
      "mean_token_accuracy": 0.12776481211185456,
      "num_tokens": 5510285.0,
      "step": 2505
    },
    {
      "entropy": 6.030932426452637,
      "epoch": 0.24910678840809844,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997459476636354,
      "loss": 5.7601,
      "mean_token_accuracy": 0.1376047044992447,
      "num_tokens": 5520028.0,
      "step": 2510
    },
    {
      "entropy": 5.963772487640381,
      "epoch": 0.24960301707026597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997442616131763,
      "loss": 5.8871,
      "mean_token_accuracy": 0.1257816217839718,
      "num_tokens": 5529271.0,
      "step": 2515
    },
    {
      "entropy": 6.042953777313232,
      "epoch": 0.25009924573243353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997425699895438,
      "loss": 5.8057,
      "mean_token_accuracy": 0.12424818351864815,
      "num_tokens": 5539396.0,
      "step": 2520
    },
    {
      "entropy": 6.0163270950317385,
      "epoch": 0.25059547439460106,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004997408727927798,
      "loss": 5.9428,
      "mean_token_accuracy": 0.12824917510151862,
      "num_tokens": 5551674.0,
      "step": 2525
    },
    {
      "entropy": 6.025873517990112,
      "epoch": 0.25109170305676853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997391700229264,
      "loss": 5.8561,
      "mean_token_accuracy": 0.1312730759382248,
      "num_tokens": 5562749.0,
      "step": 2530
    },
    {
      "entropy": 6.001038408279419,
      "epoch": 0.25158793171893606,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499737461680026,
      "loss": 5.8772,
      "mean_token_accuracy": 0.13404013216495514,
      "num_tokens": 5573478.0,
      "step": 2535
    },
    {
      "entropy": 5.929586935043335,
      "epoch": 0.2520841603811036,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997357477641207,
      "loss": 5.8682,
      "mean_token_accuracy": 0.135566084086895,
      "num_tokens": 5584546.0,
      "step": 2540
    },
    {
      "entropy": 5.922780466079712,
      "epoch": 0.2525803890432711,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997340282752532,
      "loss": 5.7439,
      "mean_token_accuracy": 0.13249244838953017,
      "num_tokens": 5594004.0,
      "step": 2545
    },
    {
      "entropy": 6.023516273498535,
      "epoch": 0.25307661770543866,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499732303213466,
      "loss": 5.8103,
      "mean_token_accuracy": 0.12287582755088806,
      "num_tokens": 5606095.0,
      "step": 2550
    },
    {
      "entropy": 6.03965277671814,
      "epoch": 0.2535728463676062,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004997305725788019,
      "loss": 5.8433,
      "mean_token_accuracy": 0.12945516034960747,
      "num_tokens": 5615596.0,
      "step": 2555
    },
    {
      "entropy": 5.911628723144531,
      "epoch": 0.2540690750297737,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499728836371304,
      "loss": 5.7889,
      "mean_token_accuracy": 0.13019108325242995,
      "num_tokens": 5626753.0,
      "step": 2560
    },
    {
      "entropy": 5.890804100036621,
      "epoch": 0.25456530369194125,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997270945910151,
      "loss": 5.8127,
      "mean_token_accuracy": 0.1270523451268673,
      "num_tokens": 5636135.0,
      "step": 2565
    },
    {
      "entropy": 6.078077507019043,
      "epoch": 0.2550615323541088,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997253472379785,
      "loss": 5.8847,
      "mean_token_accuracy": 0.12609419897198676,
      "num_tokens": 5647074.0,
      "step": 2570
    },
    {
      "entropy": 6.022387552261352,
      "epoch": 0.2555577610162763,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004997235943122375,
      "loss": 5.9047,
      "mean_token_accuracy": 0.12348231971263886,
      "num_tokens": 5658600.0,
      "step": 2575
    },
    {
      "entropy": 5.92790265083313,
      "epoch": 0.25605398967844384,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997218358138356,
      "loss": 5.8006,
      "mean_token_accuracy": 0.12867250964045523,
      "num_tokens": 5668453.0,
      "step": 2580
    },
    {
      "entropy": 6.022593784332275,
      "epoch": 0.25655021834061137,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997200717428165,
      "loss": 5.8552,
      "mean_token_accuracy": 0.12913168221712112,
      "num_tokens": 5678855.0,
      "step": 2585
    },
    {
      "entropy": 5.977180624008179,
      "epoch": 0.2570464470027789,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997183020992237,
      "loss": 5.9078,
      "mean_token_accuracy": 0.12221302613615989,
      "num_tokens": 5690063.0,
      "step": 2590
    },
    {
      "entropy": 6.008505201339721,
      "epoch": 0.25754267566494643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997165268831014,
      "loss": 5.8082,
      "mean_token_accuracy": 0.1332797534763813,
      "num_tokens": 5701487.0,
      "step": 2595
    },
    {
      "entropy": 5.929228925704956,
      "epoch": 0.25803890432711396,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997147460944933,
      "loss": 5.8037,
      "mean_token_accuracy": 0.129875248670578,
      "num_tokens": 5713034.0,
      "step": 2600
    },
    {
      "entropy": 6.02344069480896,
      "epoch": 0.25853513298928144,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004997129597334438,
      "loss": 5.8529,
      "mean_token_accuracy": 0.13128544464707376,
      "num_tokens": 5725469.0,
      "step": 2605
    },
    {
      "entropy": 5.925415468215943,
      "epoch": 0.25903136165144897,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004997111677999971,
      "loss": 5.8374,
      "mean_token_accuracy": 0.1300521619617939,
      "num_tokens": 5737099.0,
      "step": 2610
    },
    {
      "entropy": 5.973599290847778,
      "epoch": 0.2595275903136165,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004997093702941976,
      "loss": 5.8495,
      "mean_token_accuracy": 0.12512658536434174,
      "num_tokens": 5748756.0,
      "step": 2615
    },
    {
      "entropy": 6.009793472290039,
      "epoch": 0.260023818975784,
      "grad_norm": 1.0,
      "learning_rate": 0.00049970756721609,
      "loss": 5.8378,
      "mean_token_accuracy": 0.13358309268951415,
      "num_tokens": 5759598.0,
      "step": 2620
    },
    {
      "entropy": 5.976476240158081,
      "epoch": 0.26052004763795156,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499705758565719,
      "loss": 5.8817,
      "mean_token_accuracy": 0.12957399263978003,
      "num_tokens": 5770397.0,
      "step": 2625
    },
    {
      "entropy": 6.043560457229614,
      "epoch": 0.2610162763001191,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997039443431292,
      "loss": 5.9561,
      "mean_token_accuracy": 0.12363751754164695,
      "num_tokens": 5780935.0,
      "step": 2630
    },
    {
      "entropy": 6.038046789169312,
      "epoch": 0.2615125049622866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997021245483659,
      "loss": 5.8193,
      "mean_token_accuracy": 0.13055009618401528,
      "num_tokens": 5791557.0,
      "step": 2635
    },
    {
      "entropy": 5.90911078453064,
      "epoch": 0.26200873362445415,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499700299181474,
      "loss": 5.8476,
      "mean_token_accuracy": 0.1279944323003292,
      "num_tokens": 5802233.0,
      "step": 2640
    },
    {
      "entropy": 5.943614435195923,
      "epoch": 0.2625049622866217,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996984682424989,
      "loss": 5.7656,
      "mean_token_accuracy": 0.13612023070454599,
      "num_tokens": 5813424.0,
      "step": 2645
    },
    {
      "entropy": 5.928361797332764,
      "epoch": 0.2630011909487892,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000499696631731486,
      "loss": 5.7455,
      "mean_token_accuracy": 0.1320380948483944,
      "num_tokens": 5825652.0,
      "step": 2650
    },
    {
      "entropy": 6.011695718765258,
      "epoch": 0.26349741961095674,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996947896484809,
      "loss": 5.8357,
      "mean_token_accuracy": 0.13260062411427498,
      "num_tokens": 5836603.0,
      "step": 2655
    },
    {
      "entropy": 5.954362440109253,
      "epoch": 0.26399364827312427,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499692941993529,
      "loss": 5.8881,
      "mean_token_accuracy": 0.13098345547914506,
      "num_tokens": 5849365.0,
      "step": 2660
    },
    {
      "entropy": 6.022516059875488,
      "epoch": 0.2644898769352918,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996910887666765,
      "loss": 5.8082,
      "mean_token_accuracy": 0.13563585206866263,
      "num_tokens": 5860305.0,
      "step": 2665
    },
    {
      "entropy": 5.918445682525634,
      "epoch": 0.26498610559745933,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499689229967969,
      "loss": 5.7518,
      "mean_token_accuracy": 0.13243596255779266,
      "num_tokens": 5872272.0,
      "step": 2670
    },
    {
      "entropy": 5.965515995025635,
      "epoch": 0.26548233425962686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996873655974529,
      "loss": 5.7807,
      "mean_token_accuracy": 0.12951171174645423,
      "num_tokens": 5884107.0,
      "step": 2675
    },
    {
      "entropy": 5.860142564773559,
      "epoch": 0.26597856292179434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996854956551743,
      "loss": 5.7719,
      "mean_token_accuracy": 0.1379713624715805,
      "num_tokens": 5894060.0,
      "step": 2680
    },
    {
      "entropy": 5.913548135757447,
      "epoch": 0.26647479158396187,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996836201411794,
      "loss": 5.7915,
      "mean_token_accuracy": 0.13487917557358742,
      "num_tokens": 5904804.0,
      "step": 2685
    },
    {
      "entropy": 6.072554445266723,
      "epoch": 0.2669710202461294,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499681739055515,
      "loss": 5.8487,
      "mean_token_accuracy": 0.13210411742329597,
      "num_tokens": 5916496.0,
      "step": 2690
    },
    {
      "entropy": 5.941082620620728,
      "epoch": 0.26746724890829693,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996798523982276,
      "loss": 5.7944,
      "mean_token_accuracy": 0.12776002287864685,
      "num_tokens": 5926920.0,
      "step": 2695
    },
    {
      "entropy": 6.002683544158936,
      "epoch": 0.26796347757046446,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499677960169364,
      "loss": 5.9145,
      "mean_token_accuracy": 0.12489431127905845,
      "num_tokens": 5940098.0,
      "step": 2700
    },
    {
      "entropy": 5.963934755325317,
      "epoch": 0.268459706232632,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996760623689711,
      "loss": 5.7692,
      "mean_token_accuracy": 0.13152637630701064,
      "num_tokens": 5949687.0,
      "step": 2705
    },
    {
      "entropy": 5.962456846237183,
      "epoch": 0.2689559348947995,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000499674158997096,
      "loss": 5.8761,
      "mean_token_accuracy": 0.12971905767917633,
      "num_tokens": 5960777.0,
      "step": 2710
    },
    {
      "entropy": 6.018963384628296,
      "epoch": 0.26945216355696705,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996722500537859,
      "loss": 5.7802,
      "mean_token_accuracy": 0.1312619134783745,
      "num_tokens": 5972298.0,
      "step": 2715
    },
    {
      "entropy": 5.9298717975616455,
      "epoch": 0.2699483922191346,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996703355390881,
      "loss": 5.8988,
      "mean_token_accuracy": 0.12656683698296547,
      "num_tokens": 5985613.0,
      "step": 2720
    },
    {
      "entropy": 6.017775201797486,
      "epoch": 0.2704446208813021,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00049966841545305,
      "loss": 5.954,
      "mean_token_accuracy": 0.12761470898985863,
      "num_tokens": 5997530.0,
      "step": 2725
    },
    {
      "entropy": 6.011219215393067,
      "epoch": 0.27094084954346964,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996664897957195,
      "loss": 5.8209,
      "mean_token_accuracy": 0.12617397606372832,
      "num_tokens": 6008123.0,
      "step": 2730
    },
    {
      "entropy": 5.960656690597534,
      "epoch": 0.2714370782056372,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996645585671441,
      "loss": 5.8355,
      "mean_token_accuracy": 0.13165733143687247,
      "num_tokens": 6020578.0,
      "step": 2735
    },
    {
      "entropy": 5.944126558303833,
      "epoch": 0.2719333068678047,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996626217673717,
      "loss": 5.7613,
      "mean_token_accuracy": 0.13792742118239404,
      "num_tokens": 6031625.0,
      "step": 2740
    },
    {
      "entropy": 5.88871202468872,
      "epoch": 0.27242953552997223,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996606793964504,
      "loss": 5.8008,
      "mean_token_accuracy": 0.13623525723814964,
      "num_tokens": 6041320.0,
      "step": 2745
    },
    {
      "entropy": 6.01530237197876,
      "epoch": 0.27292576419213976,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996587314544284,
      "loss": 5.8187,
      "mean_token_accuracy": 0.1311524450778961,
      "num_tokens": 6052627.0,
      "step": 2750
    },
    {
      "entropy": 5.916905069351197,
      "epoch": 0.27342199285430724,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996567779413539,
      "loss": 5.7701,
      "mean_token_accuracy": 0.13783232718706132,
      "num_tokens": 6063468.0,
      "step": 2755
    },
    {
      "entropy": 5.931315326690674,
      "epoch": 0.27391822151647477,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996548188572755,
      "loss": 5.7497,
      "mean_token_accuracy": 0.1367427311837673,
      "num_tokens": 6074208.0,
      "step": 2760
    },
    {
      "entropy": 5.987172603607178,
      "epoch": 0.2744144501786423,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996528542022416,
      "loss": 5.7967,
      "mean_token_accuracy": 0.13130253031849862,
      "num_tokens": 6085068.0,
      "step": 2765
    },
    {
      "entropy": 5.929187154769897,
      "epoch": 0.27491067884080983,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499650883976301,
      "loss": 5.8003,
      "mean_token_accuracy": 0.1321723371744156,
      "num_tokens": 6096445.0,
      "step": 2770
    },
    {
      "entropy": 5.849560213088989,
      "epoch": 0.27540690750297736,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996489081795026,
      "loss": 5.7075,
      "mean_token_accuracy": 0.13662762194871902,
      "num_tokens": 6106527.0,
      "step": 2775
    },
    {
      "entropy": 5.87761869430542,
      "epoch": 0.2759031361651449,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996469268118953,
      "loss": 5.7365,
      "mean_token_accuracy": 0.13877005353569985,
      "num_tokens": 6115923.0,
      "step": 2780
    },
    {
      "entropy": 5.892449617385864,
      "epoch": 0.2763993648273124,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996449398735283,
      "loss": 5.6192,
      "mean_token_accuracy": 0.14092544242739677,
      "num_tokens": 6125861.0,
      "step": 2785
    },
    {
      "entropy": 5.953143739700318,
      "epoch": 0.27689559348947995,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996429473644509,
      "loss": 5.8347,
      "mean_token_accuracy": 0.12965964078903197,
      "num_tokens": 6137189.0,
      "step": 2790
    },
    {
      "entropy": 5.9353728771209715,
      "epoch": 0.2773918221516475,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996409492847124,
      "loss": 5.7542,
      "mean_token_accuracy": 0.13289039209485054,
      "num_tokens": 6147920.0,
      "step": 2795
    },
    {
      "entropy": 5.957181549072265,
      "epoch": 0.277888050813815,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996389456343624,
      "loss": 5.7504,
      "mean_token_accuracy": 0.13782199770212172,
      "num_tokens": 6159371.0,
      "step": 2800
    },
    {
      "entropy": 6.024093103408814,
      "epoch": 0.27838427947598254,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004996369364134507,
      "loss": 5.9497,
      "mean_token_accuracy": 0.12122881636023522,
      "num_tokens": 6172679.0,
      "step": 2805
    },
    {
      "entropy": 5.923732948303223,
      "epoch": 0.2788805081381501,
      "grad_norm": 0.875,
      "learning_rate": 0.000499634921622027,
      "loss": 5.7933,
      "mean_token_accuracy": 0.12945920675992967,
      "num_tokens": 6184244.0,
      "step": 2810
    },
    {
      "entropy": 5.903921222686767,
      "epoch": 0.2793767368003176,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996329012601413,
      "loss": 5.681,
      "mean_token_accuracy": 0.13114461600780486,
      "num_tokens": 6194826.0,
      "step": 2815
    },
    {
      "entropy": 5.968857002258301,
      "epoch": 0.27987296546248513,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996308753278437,
      "loss": 5.8212,
      "mean_token_accuracy": 0.13206653594970702,
      "num_tokens": 6206808.0,
      "step": 2820
    },
    {
      "entropy": 5.952521276473999,
      "epoch": 0.28036919412465267,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996288438251843,
      "loss": 5.8217,
      "mean_token_accuracy": 0.12705112993717194,
      "num_tokens": 6217303.0,
      "step": 2825
    },
    {
      "entropy": 5.962418413162231,
      "epoch": 0.28086542278682014,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996268067522137,
      "loss": 5.8215,
      "mean_token_accuracy": 0.12550831884145736,
      "num_tokens": 6227965.0,
      "step": 2830
    },
    {
      "entropy": 6.0486168384552,
      "epoch": 0.28136165144898767,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996247641089823,
      "loss": 5.8799,
      "mean_token_accuracy": 0.12784553319215775,
      "num_tokens": 6239011.0,
      "step": 2835
    },
    {
      "entropy": 5.873768758773804,
      "epoch": 0.2818578801111552,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996227158955409,
      "loss": 5.7375,
      "mean_token_accuracy": 0.13439323455095292,
      "num_tokens": 6249084.0,
      "step": 2840
    },
    {
      "entropy": 5.913141870498658,
      "epoch": 0.28235410877332273,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049962066211194,
      "loss": 5.8041,
      "mean_token_accuracy": 0.12883702740073205,
      "num_tokens": 6260706.0,
      "step": 2845
    },
    {
      "entropy": 5.906771898269653,
      "epoch": 0.28285033743549026,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996186027582308,
      "loss": 5.7267,
      "mean_token_accuracy": 0.13956455439329146,
      "num_tokens": 6272848.0,
      "step": 2850
    },
    {
      "entropy": 5.91385931968689,
      "epoch": 0.2833465660976578,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996165378344642,
      "loss": 5.741,
      "mean_token_accuracy": 0.13406383767724037,
      "num_tokens": 6283820.0,
      "step": 2855
    },
    {
      "entropy": 5.939226245880127,
      "epoch": 0.2838427947598253,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996144673406914,
      "loss": 5.915,
      "mean_token_accuracy": 0.12336270064115525,
      "num_tokens": 6294193.0,
      "step": 2860
    },
    {
      "entropy": 5.937975311279297,
      "epoch": 0.28433902342199285,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996123912769639,
      "loss": 5.7329,
      "mean_token_accuracy": 0.14111651629209518,
      "num_tokens": 6304407.0,
      "step": 2865
    },
    {
      "entropy": 5.87665205001831,
      "epoch": 0.2848352520841604,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996103096433332,
      "loss": 5.8212,
      "mean_token_accuracy": 0.13326509818434715,
      "num_tokens": 6316311.0,
      "step": 2870
    },
    {
      "entropy": 5.942044496536255,
      "epoch": 0.2853314807463279,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996082224398506,
      "loss": 5.7734,
      "mean_token_accuracy": 0.13422515317797662,
      "num_tokens": 6327248.0,
      "step": 2875
    },
    {
      "entropy": 5.964761114120483,
      "epoch": 0.28582770940849545,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004996061296665682,
      "loss": 5.7933,
      "mean_token_accuracy": 0.131377661973238,
      "num_tokens": 6338494.0,
      "step": 2880
    },
    {
      "entropy": 5.899775981903076,
      "epoch": 0.286323938070663,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004996040313235377,
      "loss": 5.6797,
      "mean_token_accuracy": 0.1386905901134014,
      "num_tokens": 6349038.0,
      "step": 2885
    },
    {
      "entropy": 5.949228382110595,
      "epoch": 0.2868201667328305,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996019274108112,
      "loss": 5.8067,
      "mean_token_accuracy": 0.12893664464354515,
      "num_tokens": 6360180.0,
      "step": 2890
    },
    {
      "entropy": 5.813185453414917,
      "epoch": 0.28731639539499804,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499599817928441,
      "loss": 5.7393,
      "mean_token_accuracy": 0.13658856898546218,
      "num_tokens": 6370931.0,
      "step": 2895
    },
    {
      "entropy": 5.841475248336792,
      "epoch": 0.28781262405716557,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995977028764792,
      "loss": 5.7214,
      "mean_token_accuracy": 0.13749603554606438,
      "num_tokens": 6381810.0,
      "step": 2900
    },
    {
      "entropy": 5.949096298217773,
      "epoch": 0.28830885271933304,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995955822549783,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1378832995891571,
      "num_tokens": 6392268.0,
      "step": 2905
    },
    {
      "entropy": 5.916205501556396,
      "epoch": 0.2888050813815006,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995934560639909,
      "loss": 5.7899,
      "mean_token_accuracy": 0.13368103802204132,
      "num_tokens": 6403650.0,
      "step": 2910
    },
    {
      "entropy": 5.920439100265503,
      "epoch": 0.2893013100436681,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995913243035699,
      "loss": 5.7577,
      "mean_token_accuracy": 0.1311557449400425,
      "num_tokens": 6414672.0,
      "step": 2915
    },
    {
      "entropy": 5.954520988464355,
      "epoch": 0.28979753870583563,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995891869737678,
      "loss": 5.8282,
      "mean_token_accuracy": 0.13221259862184526,
      "num_tokens": 6424916.0,
      "step": 2920
    },
    {
      "entropy": 5.9455653667449955,
      "epoch": 0.29029376736800316,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995870440746379,
      "loss": 5.7778,
      "mean_token_accuracy": 0.13875048831105233,
      "num_tokens": 6435950.0,
      "step": 2925
    },
    {
      "entropy": 5.8589989185333256,
      "epoch": 0.2907899960301707,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004995848956062333,
      "loss": 5.7475,
      "mean_token_accuracy": 0.13245701491832734,
      "num_tokens": 6446336.0,
      "step": 2930
    },
    {
      "entropy": 5.905734395980835,
      "epoch": 0.2912862246923382,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995827415686071,
      "loss": 5.8066,
      "mean_token_accuracy": 0.12863307669758797,
      "num_tokens": 6457655.0,
      "step": 2935
    },
    {
      "entropy": 5.950538587570191,
      "epoch": 0.29178245335450576,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499580581961813,
      "loss": 5.6973,
      "mean_token_accuracy": 0.13707054033875465,
      "num_tokens": 6469627.0,
      "step": 2940
    },
    {
      "entropy": 5.882530784606933,
      "epoch": 0.2922786820166733,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004995784167859043,
      "loss": 5.7915,
      "mean_token_accuracy": 0.1303873836994171,
      "num_tokens": 6480351.0,
      "step": 2945
    },
    {
      "entropy": 5.980832910537719,
      "epoch": 0.2927749106788408,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995762460409349,
      "loss": 5.8253,
      "mean_token_accuracy": 0.1353992909193039,
      "num_tokens": 6491651.0,
      "step": 2950
    },
    {
      "entropy": 5.9034021377563475,
      "epoch": 0.29327113934100835,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004995740697269584,
      "loss": 5.6734,
      "mean_token_accuracy": 0.13603452891111373,
      "num_tokens": 6502531.0,
      "step": 2955
    },
    {
      "entropy": 5.818338775634766,
      "epoch": 0.2937673680031759,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995718878440289,
      "loss": 5.6721,
      "mean_token_accuracy": 0.13839882612228394,
      "num_tokens": 6511938.0,
      "step": 2960
    },
    {
      "entropy": 5.965565156936646,
      "epoch": 0.2942635966653434,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995697003922006,
      "loss": 5.8283,
      "mean_token_accuracy": 0.12819456234574317,
      "num_tokens": 6522775.0,
      "step": 2965
    },
    {
      "entropy": 5.922029304504394,
      "epoch": 0.29475982532751094,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995675073715274,
      "loss": 5.6627,
      "mean_token_accuracy": 0.1319277822971344,
      "num_tokens": 6534660.0,
      "step": 2970
    },
    {
      "entropy": 5.8830774307250975,
      "epoch": 0.29525605398967847,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995653087820641,
      "loss": 5.7927,
      "mean_token_accuracy": 0.13659714311361312,
      "num_tokens": 6545673.0,
      "step": 2975
    },
    {
      "entropy": 5.924934196472168,
      "epoch": 0.29575228265184594,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499563104623865,
      "loss": 5.7547,
      "mean_token_accuracy": 0.13512978926301003,
      "num_tokens": 6556912.0,
      "step": 2980
    },
    {
      "entropy": 5.935703706741333,
      "epoch": 0.2962485113140135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995608948969849,
      "loss": 5.8517,
      "mean_token_accuracy": 0.12844418361783028,
      "num_tokens": 6567532.0,
      "step": 2985
    },
    {
      "entropy": 5.927107429504394,
      "epoch": 0.296744739976181,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995586796014784,
      "loss": 5.8246,
      "mean_token_accuracy": 0.13212826922535897,
      "num_tokens": 6577830.0,
      "step": 2990
    },
    {
      "entropy": 5.860514736175537,
      "epoch": 0.29724096863834853,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995564587374005,
      "loss": 5.7143,
      "mean_token_accuracy": 0.13365908712148666,
      "num_tokens": 6587254.0,
      "step": 2995
    },
    {
      "entropy": 5.91803126335144,
      "epoch": 0.29773719730051607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995542323048063,
      "loss": 5.9097,
      "mean_token_accuracy": 0.1270786426961422,
      "num_tokens": 6598294.0,
      "step": 3000
    },
    {
      "epoch": 0.29773719730051607,
      "eval_entropy": 5.736310448491338,
      "eval_loss": 5.833684921264648,
      "eval_mean_token_accuracy": 0.1343776344935248,
      "eval_num_tokens": 6598294.0,
      "eval_runtime": 20.4598,
      "eval_samples_per_second": 1645.621,
      "eval_steps_per_second": 205.721,
      "step": 3000
    },
    {
      "entropy": 5.965045642852783,
      "epoch": 0.2982334259626836,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499552000303751,
      "loss": 5.7748,
      "mean_token_accuracy": 0.13394526094198228,
      "num_tokens": 6608772.0,
      "step": 3005
    },
    {
      "entropy": 5.890723085403442,
      "epoch": 0.2987296546248511,
      "grad_norm": 0.875,
      "learning_rate": 0.00049954976273429,
      "loss": 5.7834,
      "mean_token_accuracy": 0.1345479592680931,
      "num_tokens": 6619973.0,
      "step": 3010
    },
    {
      "entropy": 5.9124915599823,
      "epoch": 0.29922588328701866,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995475195964787,
      "loss": 5.8041,
      "mean_token_accuracy": 0.13232846856117247,
      "num_tokens": 6631172.0,
      "step": 3015
    },
    {
      "entropy": 5.898634815216065,
      "epoch": 0.2997221119491862,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995452708903727,
      "loss": 5.7921,
      "mean_token_accuracy": 0.13078083023428916,
      "num_tokens": 6641343.0,
      "step": 3020
    },
    {
      "entropy": 5.881944179534912,
      "epoch": 0.3002183406113537,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499543016616028,
      "loss": 5.7477,
      "mean_token_accuracy": 0.13408440947532654,
      "num_tokens": 6652892.0,
      "step": 3025
    },
    {
      "entropy": 5.919469356536865,
      "epoch": 0.30071456927352125,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004995407567735002,
      "loss": 5.7607,
      "mean_token_accuracy": 0.13404800221323968,
      "num_tokens": 6664177.0,
      "step": 3030
    },
    {
      "entropy": 5.88273811340332,
      "epoch": 0.3012107979356888,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995384913628456,
      "loss": 5.7202,
      "mean_token_accuracy": 0.13407384157180785,
      "num_tokens": 6674402.0,
      "step": 3035
    },
    {
      "entropy": 5.927703380584717,
      "epoch": 0.3017070265978563,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995362203841202,
      "loss": 5.8649,
      "mean_token_accuracy": 0.1244761310517788,
      "num_tokens": 6684886.0,
      "step": 3040
    },
    {
      "entropy": 5.933945798873902,
      "epoch": 0.30220325526002384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995339438373804,
      "loss": 5.6885,
      "mean_token_accuracy": 0.13560489416122437,
      "num_tokens": 6694774.0,
      "step": 3045
    },
    {
      "entropy": 5.8841780662536625,
      "epoch": 0.30269948392219137,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995316617226825,
      "loss": 5.7557,
      "mean_token_accuracy": 0.13320276215672494,
      "num_tokens": 6705346.0,
      "step": 3050
    },
    {
      "entropy": 5.866413688659668,
      "epoch": 0.30319571258435885,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995293740400833,
      "loss": 5.6994,
      "mean_token_accuracy": 0.13981354087591172,
      "num_tokens": 6715579.0,
      "step": 3055
    },
    {
      "entropy": 5.8568097114562985,
      "epoch": 0.3036919412465264,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995270807896395,
      "loss": 5.6647,
      "mean_token_accuracy": 0.13911256343126296,
      "num_tokens": 6726382.0,
      "step": 3060
    },
    {
      "entropy": 5.868578052520752,
      "epoch": 0.3041881699086939,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995247819714079,
      "loss": 5.7412,
      "mean_token_accuracy": 0.13409880474209784,
      "num_tokens": 6737245.0,
      "step": 3065
    },
    {
      "entropy": 5.922122812271118,
      "epoch": 0.30468439857086144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995224775854454,
      "loss": 5.9378,
      "mean_token_accuracy": 0.1288504920899868,
      "num_tokens": 6749113.0,
      "step": 3070
    },
    {
      "entropy": 6.008074235916138,
      "epoch": 0.30518062723302897,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004995201676318093,
      "loss": 5.7968,
      "mean_token_accuracy": 0.1340916134417057,
      "num_tokens": 6761987.0,
      "step": 3075
    },
    {
      "entropy": 5.899579286575317,
      "epoch": 0.3056768558951965,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995178521105568,
      "loss": 5.728,
      "mean_token_accuracy": 0.13068147823214532,
      "num_tokens": 6771279.0,
      "step": 3080
    },
    {
      "entropy": 5.836708116531372,
      "epoch": 0.30617308455736403,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995155310217454,
      "loss": 5.6775,
      "mean_token_accuracy": 0.13470116928219794,
      "num_tokens": 6781038.0,
      "step": 3085
    },
    {
      "entropy": 5.8924072265625,
      "epoch": 0.30666931321953156,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995132043654327,
      "loss": 5.8087,
      "mean_token_accuracy": 0.13182928264141083,
      "num_tokens": 6792686.0,
      "step": 3090
    },
    {
      "entropy": 5.900499677658081,
      "epoch": 0.3071655418816991,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004995108721416763,
      "loss": 5.7806,
      "mean_token_accuracy": 0.1379054993391037,
      "num_tokens": 6804414.0,
      "step": 3095
    },
    {
      "entropy": 5.904523944854736,
      "epoch": 0.3076617705438666,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995085343505339,
      "loss": 5.7469,
      "mean_token_accuracy": 0.13811479806900023,
      "num_tokens": 6816180.0,
      "step": 3100
    },
    {
      "entropy": 5.8829244613647464,
      "epoch": 0.30815799920603415,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995061909920638,
      "loss": 5.7457,
      "mean_token_accuracy": 0.13014977425336838,
      "num_tokens": 6827114.0,
      "step": 3105
    },
    {
      "entropy": 5.922984552383423,
      "epoch": 0.3086542278682017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995038420663238,
      "loss": 5.7995,
      "mean_token_accuracy": 0.13707774877548218,
      "num_tokens": 6836953.0,
      "step": 3110
    },
    {
      "entropy": 5.955396318435669,
      "epoch": 0.3091504565303692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995014875733724,
      "loss": 5.8367,
      "mean_token_accuracy": 0.13107877224683762,
      "num_tokens": 6847963.0,
      "step": 3115
    },
    {
      "entropy": 5.940256834030151,
      "epoch": 0.30964668519253674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994991275132678,
      "loss": 5.8617,
      "mean_token_accuracy": 0.12713079154491425,
      "num_tokens": 6858264.0,
      "step": 3120
    },
    {
      "entropy": 5.935183382034301,
      "epoch": 0.31014291385470427,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994967618860685,
      "loss": 5.7433,
      "mean_token_accuracy": 0.13888981193304062,
      "num_tokens": 6868861.0,
      "step": 3125
    },
    {
      "entropy": 5.904120779037475,
      "epoch": 0.31063914251687175,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994943906918333,
      "loss": 5.7738,
      "mean_token_accuracy": 0.13864820897579194,
      "num_tokens": 6879797.0,
      "step": 3130
    },
    {
      "entropy": 5.863951730728149,
      "epoch": 0.3111353711790393,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994920139306211,
      "loss": 5.7112,
      "mean_token_accuracy": 0.13419859036803244,
      "num_tokens": 6891323.0,
      "step": 3135
    },
    {
      "entropy": 5.864938402175904,
      "epoch": 0.3116315998412068,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994896316024906,
      "loss": 5.7972,
      "mean_token_accuracy": 0.13606368377804756,
      "num_tokens": 6903345.0,
      "step": 3140
    },
    {
      "entropy": 5.899130201339721,
      "epoch": 0.31212782850337434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499487243707501,
      "loss": 5.7245,
      "mean_token_accuracy": 0.1321530058979988,
      "num_tokens": 6914342.0,
      "step": 3145
    },
    {
      "entropy": 5.908827257156372,
      "epoch": 0.31262405716554187,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994848502457114,
      "loss": 5.791,
      "mean_token_accuracy": 0.13517412841320037,
      "num_tokens": 6926691.0,
      "step": 3150
    },
    {
      "entropy": 5.828939294815063,
      "epoch": 0.3131202858277094,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994824512171812,
      "loss": 5.6651,
      "mean_token_accuracy": 0.13288380950689316,
      "num_tokens": 6937841.0,
      "step": 3155
    },
    {
      "entropy": 5.834991693496704,
      "epoch": 0.31361651448987693,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994800466219701,
      "loss": 5.7545,
      "mean_token_accuracy": 0.13247875422239302,
      "num_tokens": 6948601.0,
      "step": 3160
    },
    {
      "entropy": 5.919907331466675,
      "epoch": 0.31411274315204446,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994776364601374,
      "loss": 5.6854,
      "mean_token_accuracy": 0.1365569844841957,
      "num_tokens": 6959775.0,
      "step": 3165
    },
    {
      "entropy": 5.804076242446899,
      "epoch": 0.314608971814212,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994752207317431,
      "loss": 5.7132,
      "mean_token_accuracy": 0.13514667749404907,
      "num_tokens": 6969202.0,
      "step": 3170
    },
    {
      "entropy": 5.841969680786133,
      "epoch": 0.3151052004763795,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499472799436847,
      "loss": 5.6923,
      "mean_token_accuracy": 0.13741817697882652,
      "num_tokens": 6978633.0,
      "step": 3175
    },
    {
      "entropy": 5.966694450378418,
      "epoch": 0.31560142913854705,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994703725755092,
      "loss": 5.7608,
      "mean_token_accuracy": 0.1306213304400444,
      "num_tokens": 6989638.0,
      "step": 3180
    },
    {
      "entropy": 5.878809642791748,
      "epoch": 0.3160976578007146,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994679401477899,
      "loss": 5.7345,
      "mean_token_accuracy": 0.13373329117894173,
      "num_tokens": 7001606.0,
      "step": 3185
    },
    {
      "entropy": 5.825911903381348,
      "epoch": 0.3165938864628821,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994655021537492,
      "loss": 5.6948,
      "mean_token_accuracy": 0.13437199518084525,
      "num_tokens": 7012915.0,
      "step": 3190
    },
    {
      "entropy": 5.852247428894043,
      "epoch": 0.31709011512504964,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994630585934478,
      "loss": 5.7659,
      "mean_token_accuracy": 0.1293450452387333,
      "num_tokens": 7023191.0,
      "step": 3195
    },
    {
      "entropy": 5.800911855697632,
      "epoch": 0.3175863437872172,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994606094669463,
      "loss": 5.6613,
      "mean_token_accuracy": 0.14065127149224282,
      "num_tokens": 7032506.0,
      "step": 3200
    },
    {
      "entropy": 5.886978054046631,
      "epoch": 0.3180825724493847,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994581547743053,
      "loss": 5.7915,
      "mean_token_accuracy": 0.12982771173119545,
      "num_tokens": 7046098.0,
      "step": 3205
    },
    {
      "entropy": 5.971642398834229,
      "epoch": 0.3185788011115522,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994556945155857,
      "loss": 5.8062,
      "mean_token_accuracy": 0.13670874536037445,
      "num_tokens": 7057434.0,
      "step": 3210
    },
    {
      "entropy": 5.807867908477784,
      "epoch": 0.3190750297737197,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994532286908486,
      "loss": 5.7201,
      "mean_token_accuracy": 0.13328924551606178,
      "num_tokens": 7067674.0,
      "step": 3215
    },
    {
      "entropy": 5.913128900527954,
      "epoch": 0.31957125843588724,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000499450757300155,
      "loss": 5.7909,
      "mean_token_accuracy": 0.1317487232387066,
      "num_tokens": 7077898.0,
      "step": 3220
    },
    {
      "entropy": 5.948103380203247,
      "epoch": 0.32006748709805477,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994482803435664,
      "loss": 5.7544,
      "mean_token_accuracy": 0.12962503507733344,
      "num_tokens": 7088566.0,
      "step": 3225
    },
    {
      "entropy": 5.848378849029541,
      "epoch": 0.3205637157602223,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499445797821144,
      "loss": 5.7586,
      "mean_token_accuracy": 0.13164688199758529,
      "num_tokens": 7099533.0,
      "step": 3230
    },
    {
      "entropy": 5.902796840667724,
      "epoch": 0.32105994442238983,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004994433097329494,
      "loss": 5.748,
      "mean_token_accuracy": 0.13455025032162665,
      "num_tokens": 7112254.0,
      "step": 3235
    },
    {
      "entropy": 5.976428604125976,
      "epoch": 0.32155617308455736,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004994408160790444,
      "loss": 5.7776,
      "mean_token_accuracy": 0.13113492578268052,
      "num_tokens": 7124704.0,
      "step": 3240
    },
    {
      "entropy": 6.01985354423523,
      "epoch": 0.3220524017467249,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004994383168594907,
      "loss": 5.9524,
      "mean_token_accuracy": 0.12315566837787628,
      "num_tokens": 7137741.0,
      "step": 3245
    },
    {
      "entropy": 5.8972578048706055,
      "epoch": 0.3225486304088924,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994358120743506,
      "loss": 5.7248,
      "mean_token_accuracy": 0.13049985989928245,
      "num_tokens": 7147888.0,
      "step": 3250
    },
    {
      "entropy": 5.7696130752563475,
      "epoch": 0.32304485907105995,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994333017236858,
      "loss": 5.6808,
      "mean_token_accuracy": 0.14042748287320136,
      "num_tokens": 7158950.0,
      "step": 3255
    },
    {
      "entropy": 5.893270206451416,
      "epoch": 0.3235410877332275,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994307858075587,
      "loss": 5.82,
      "mean_token_accuracy": 0.12969032377004625,
      "num_tokens": 7170156.0,
      "step": 3260
    },
    {
      "entropy": 5.88913311958313,
      "epoch": 0.324037316395395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994282643260319,
      "loss": 5.6639,
      "mean_token_accuracy": 0.1393490567803383,
      "num_tokens": 7180576.0,
      "step": 3265
    },
    {
      "entropy": 5.778321409225464,
      "epoch": 0.32453354505756254,
      "grad_norm": 0.875,
      "learning_rate": 0.0004994257372791676,
      "loss": 5.7328,
      "mean_token_accuracy": 0.13770043551921846,
      "num_tokens": 7192380.0,
      "step": 3270
    },
    {
      "entropy": 5.9430975914001465,
      "epoch": 0.3250297737197301,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004994232046670286,
      "loss": 5.785,
      "mean_token_accuracy": 0.13338666409254074,
      "num_tokens": 7203898.0,
      "step": 3275
    },
    {
      "entropy": 5.927338647842407,
      "epoch": 0.3255260023818976,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994206664896778,
      "loss": 5.7436,
      "mean_token_accuracy": 0.13501385301351548,
      "num_tokens": 7215241.0,
      "step": 3280
    },
    {
      "entropy": 5.829563617706299,
      "epoch": 0.3260222310440651,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994181227471779,
      "loss": 5.7053,
      "mean_token_accuracy": 0.13508176431059837,
      "num_tokens": 7226039.0,
      "step": 3285
    },
    {
      "entropy": 5.848139429092408,
      "epoch": 0.3265184597062326,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994155734395923,
      "loss": 5.7402,
      "mean_token_accuracy": 0.1360563226044178,
      "num_tokens": 7237148.0,
      "step": 3290
    },
    {
      "entropy": 5.884560918807983,
      "epoch": 0.32701468836840014,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499413018566984,
      "loss": 5.7319,
      "mean_token_accuracy": 0.13255968913435937,
      "num_tokens": 7249284.0,
      "step": 3295
    },
    {
      "entropy": 5.84690670967102,
      "epoch": 0.32751091703056767,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004994104581294163,
      "loss": 5.7352,
      "mean_token_accuracy": 0.13312482759356498,
      "num_tokens": 7259967.0,
      "step": 3300
    },
    {
      "entropy": 5.812117671966552,
      "epoch": 0.3280071456927352,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994078921269529,
      "loss": 5.6932,
      "mean_token_accuracy": 0.14062133729457854,
      "num_tokens": 7271145.0,
      "step": 3305
    },
    {
      "entropy": 5.945493221282959,
      "epoch": 0.32850337435490273,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994053205596573,
      "loss": 5.7065,
      "mean_token_accuracy": 0.14030177742242814,
      "num_tokens": 7281872.0,
      "step": 3310
    },
    {
      "entropy": 5.81722240447998,
      "epoch": 0.32899960301707026,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994027434275934,
      "loss": 5.7316,
      "mean_token_accuracy": 0.13382603749632835,
      "num_tokens": 7293330.0,
      "step": 3315
    },
    {
      "entropy": 5.945823049545288,
      "epoch": 0.3294958316792378,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004994001607308249,
      "loss": 5.7912,
      "mean_token_accuracy": 0.12995256781578063,
      "num_tokens": 7304187.0,
      "step": 3320
    },
    {
      "entropy": 5.922194385528565,
      "epoch": 0.3299920603414053,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499397572469416,
      "loss": 5.6883,
      "mean_token_accuracy": 0.13331346735358238,
      "num_tokens": 7314756.0,
      "step": 3325
    },
    {
      "entropy": 5.777895164489746,
      "epoch": 0.33048828900357285,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004993949786434308,
      "loss": 5.7265,
      "mean_token_accuracy": 0.13707108423113823,
      "num_tokens": 7325711.0,
      "step": 3330
    },
    {
      "entropy": 5.816650772094727,
      "epoch": 0.3309845176657404,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993923792529337,
      "loss": 5.6539,
      "mean_token_accuracy": 0.1394137054681778,
      "num_tokens": 7335639.0,
      "step": 3335
    },
    {
      "entropy": 5.944006729125976,
      "epoch": 0.3314807463279079,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993897742979892,
      "loss": 5.77,
      "mean_token_accuracy": 0.13584464341402053,
      "num_tokens": 7346621.0,
      "step": 3340
    },
    {
      "entropy": 5.865361833572388,
      "epoch": 0.33197697499007545,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004993871637786616,
      "loss": 5.7419,
      "mean_token_accuracy": 0.12874035984277726,
      "num_tokens": 7358950.0,
      "step": 3345
    },
    {
      "entropy": 5.781374597549439,
      "epoch": 0.332473203652243,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499384547695016,
      "loss": 5.5919,
      "mean_token_accuracy": 0.13987331539392472,
      "num_tokens": 7369629.0,
      "step": 3350
    },
    {
      "entropy": 5.828374862670898,
      "epoch": 0.3329694323144105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499381926047117,
      "loss": 5.6931,
      "mean_token_accuracy": 0.13623791709542274,
      "num_tokens": 7379853.0,
      "step": 3355
    },
    {
      "entropy": 5.859361219406128,
      "epoch": 0.333465660976578,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049937929883503,
      "loss": 5.7286,
      "mean_token_accuracy": 0.134597609937191,
      "num_tokens": 7391323.0,
      "step": 3360
    },
    {
      "entropy": 5.8687690734863285,
      "epoch": 0.3339618896387455,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004993766660588196,
      "loss": 5.6956,
      "mean_token_accuracy": 0.13231184631586074,
      "num_tokens": 7402201.0,
      "step": 3365
    },
    {
      "entropy": 5.923811769485473,
      "epoch": 0.33445811830091304,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993740277185515,
      "loss": 5.9406,
      "mean_token_accuracy": 0.12415642142295838,
      "num_tokens": 7413783.0,
      "step": 3370
    },
    {
      "entropy": 5.868533515930176,
      "epoch": 0.3349543469630806,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499371383814291,
      "loss": 5.7056,
      "mean_token_accuracy": 0.14041850715875626,
      "num_tokens": 7425909.0,
      "step": 3375
    },
    {
      "entropy": 5.7861114025115965,
      "epoch": 0.3354505756252481,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993687343461036,
      "loss": 5.6019,
      "mean_token_accuracy": 0.14431009963154792,
      "num_tokens": 7436346.0,
      "step": 3380
    },
    {
      "entropy": 5.825959920883179,
      "epoch": 0.33594680428741563,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993660793140551,
      "loss": 5.6633,
      "mean_token_accuracy": 0.13939688578248025,
      "num_tokens": 7446596.0,
      "step": 3385
    },
    {
      "entropy": 5.820888185501099,
      "epoch": 0.33644303294958317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993634187182113,
      "loss": 5.7521,
      "mean_token_accuracy": 0.13098914325237274,
      "num_tokens": 7457697.0,
      "step": 3390
    },
    {
      "entropy": 5.968964767456055,
      "epoch": 0.3369392616117507,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993607525586381,
      "loss": 5.8557,
      "mean_token_accuracy": 0.1338909812271595,
      "num_tokens": 7468541.0,
      "step": 3395
    },
    {
      "entropy": 5.908163928985596,
      "epoch": 0.3374354902739182,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993580808354018,
      "loss": 5.7496,
      "mean_token_accuracy": 0.13644585236907006,
      "num_tokens": 7479781.0,
      "step": 3400
    },
    {
      "entropy": 5.845385932922364,
      "epoch": 0.33793171893608576,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993554035485685,
      "loss": 5.6505,
      "mean_token_accuracy": 0.13407665640115737,
      "num_tokens": 7490722.0,
      "step": 3405
    },
    {
      "entropy": 5.858500814437866,
      "epoch": 0.3384279475982533,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993527206982047,
      "loss": 5.7627,
      "mean_token_accuracy": 0.13317100778222085,
      "num_tokens": 7502293.0,
      "step": 3410
    },
    {
      "entropy": 5.861693000793457,
      "epoch": 0.3389241762604208,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004993500322843767,
      "loss": 5.6908,
      "mean_token_accuracy": 0.1373884968459606,
      "num_tokens": 7513846.0,
      "step": 3415
    },
    {
      "entropy": 5.818862819671631,
      "epoch": 0.33942040492258835,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004993473383071514,
      "loss": 5.7087,
      "mean_token_accuracy": 0.1370914064347744,
      "num_tokens": 7524062.0,
      "step": 3420
    },
    {
      "entropy": 5.8636046886444095,
      "epoch": 0.3399166335847559,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004993446387665956,
      "loss": 5.7809,
      "mean_token_accuracy": 0.13834589645266532,
      "num_tokens": 7535533.0,
      "step": 3425
    },
    {
      "entropy": 5.872986125946045,
      "epoch": 0.3404128622469234,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993419336627761,
      "loss": 5.674,
      "mean_token_accuracy": 0.13601678013801574,
      "num_tokens": 7545951.0,
      "step": 3430
    },
    {
      "entropy": 5.807206678390503,
      "epoch": 0.3409090909090909,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993392229957602,
      "loss": 5.6137,
      "mean_token_accuracy": 0.13744947388768197,
      "num_tokens": 7557320.0,
      "step": 3435
    },
    {
      "entropy": 5.772375631332397,
      "epoch": 0.3414053195712584,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993365067656148,
      "loss": 5.7249,
      "mean_token_accuracy": 0.13479891270399094,
      "num_tokens": 7568078.0,
      "step": 3440
    },
    {
      "entropy": 5.857366847991943,
      "epoch": 0.34190154823342594,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004993337849724076,
      "loss": 5.715,
      "mean_token_accuracy": 0.1360277846455574,
      "num_tokens": 7578940.0,
      "step": 3445
    },
    {
      "entropy": 5.8517683982849125,
      "epoch": 0.3423977768955935,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993310576162059,
      "loss": 5.7208,
      "mean_token_accuracy": 0.1363377019762993,
      "num_tokens": 7589864.0,
      "step": 3450
    },
    {
      "entropy": 5.871683740615845,
      "epoch": 0.342894005557761,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993283246970772,
      "loss": 5.6698,
      "mean_token_accuracy": 0.13472990989685057,
      "num_tokens": 7599795.0,
      "step": 3455
    },
    {
      "entropy": 5.839810752868653,
      "epoch": 0.34339023421992854,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004993255862150896,
      "loss": 5.6511,
      "mean_token_accuracy": 0.13812209963798522,
      "num_tokens": 7610056.0,
      "step": 3460
    },
    {
      "entropy": 5.797512435913086,
      "epoch": 0.34388646288209607,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993228421703108,
      "loss": 5.6532,
      "mean_token_accuracy": 0.1349962443113327,
      "num_tokens": 7620218.0,
      "step": 3465
    },
    {
      "entropy": 5.824069023132324,
      "epoch": 0.3443826915442636,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993200925628088,
      "loss": 5.7585,
      "mean_token_accuracy": 0.13310205042362214,
      "num_tokens": 7632781.0,
      "step": 3470
    },
    {
      "entropy": 5.892023372650146,
      "epoch": 0.3448789202064311,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499317337392652,
      "loss": 5.7235,
      "mean_token_accuracy": 0.1346145860850811,
      "num_tokens": 7643845.0,
      "step": 3475
    },
    {
      "entropy": 5.82875280380249,
      "epoch": 0.34537514886859866,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993145766599085,
      "loss": 5.7389,
      "mean_token_accuracy": 0.13387636840343475,
      "num_tokens": 7654583.0,
      "step": 3480
    },
    {
      "entropy": 5.771491527557373,
      "epoch": 0.3458713775307662,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993118103646469,
      "loss": 5.6007,
      "mean_token_accuracy": 0.14086305797100068,
      "num_tokens": 7665714.0,
      "step": 3485
    },
    {
      "entropy": 5.85463547706604,
      "epoch": 0.3463676061929337,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993090385069357,
      "loss": 5.6571,
      "mean_token_accuracy": 0.13845691010355948,
      "num_tokens": 7676011.0,
      "step": 3490
    },
    {
      "entropy": 5.796469354629517,
      "epoch": 0.34686383485510125,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993062610868437,
      "loss": 5.7147,
      "mean_token_accuracy": 0.13811295256018638,
      "num_tokens": 7686927.0,
      "step": 3495
    },
    {
      "entropy": 5.791467809677124,
      "epoch": 0.3473600635172688,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993034781044398,
      "loss": 5.6567,
      "mean_token_accuracy": 0.13712866380810737,
      "num_tokens": 7698419.0,
      "step": 3500
    },
    {
      "entropy": 5.894871520996094,
      "epoch": 0.3478562921794363,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499300689559793,
      "loss": 5.6472,
      "mean_token_accuracy": 0.14020137637853622,
      "num_tokens": 7709695.0,
      "step": 3505
    },
    {
      "entropy": 5.8230983257293705,
      "epoch": 0.3483525208416038,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004992978954529724,
      "loss": 5.6153,
      "mean_token_accuracy": 0.139559406042099,
      "num_tokens": 7721309.0,
      "step": 3510
    },
    {
      "entropy": 5.824827814102173,
      "epoch": 0.3488487495037713,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992950957840473,
      "loss": 5.6925,
      "mean_token_accuracy": 0.1407937191426754,
      "num_tokens": 7732074.0,
      "step": 3515
    },
    {
      "entropy": 5.877378273010254,
      "epoch": 0.34934497816593885,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499292290553087,
      "loss": 5.7629,
      "mean_token_accuracy": 0.13451908826828002,
      "num_tokens": 7742869.0,
      "step": 3520
    },
    {
      "entropy": 5.786126756668091,
      "epoch": 0.3498412068281064,
      "grad_norm": 0.875,
      "learning_rate": 0.0004992894797601614,
      "loss": 5.6633,
      "mean_token_accuracy": 0.13667808771133422,
      "num_tokens": 7754895.0,
      "step": 3525
    },
    {
      "entropy": 5.858791208267212,
      "epoch": 0.3503374354902739,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049928666340534,
      "loss": 5.6375,
      "mean_token_accuracy": 0.142080157995224,
      "num_tokens": 7765262.0,
      "step": 3530
    },
    {
      "entropy": 5.75749773979187,
      "epoch": 0.35083366415244144,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992838414886925,
      "loss": 5.7192,
      "mean_token_accuracy": 0.13321268782019616,
      "num_tokens": 7777348.0,
      "step": 3535
    },
    {
      "entropy": 5.875128316879272,
      "epoch": 0.35132989281460897,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992810140102891,
      "loss": 5.704,
      "mean_token_accuracy": 0.13238653540611267,
      "num_tokens": 7788161.0,
      "step": 3540
    },
    {
      "entropy": 5.822514963150025,
      "epoch": 0.3518261214767765,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992781809701999,
      "loss": 5.6876,
      "mean_token_accuracy": 0.13483116030693054,
      "num_tokens": 7798629.0,
      "step": 3545
    },
    {
      "entropy": 5.76167573928833,
      "epoch": 0.35232235013894403,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004992753423684951,
      "loss": 5.6964,
      "mean_token_accuracy": 0.1355031356215477,
      "num_tokens": 7810148.0,
      "step": 3550
    },
    {
      "entropy": 5.86360993385315,
      "epoch": 0.35281857880111156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992724982052449,
      "loss": 5.679,
      "mean_token_accuracy": 0.13520049154758454,
      "num_tokens": 7821667.0,
      "step": 3555
    },
    {
      "entropy": 5.813494110107422,
      "epoch": 0.3533148074632791,
      "grad_norm": 0.8125,
      "learning_rate": 0.00049926964848052,
      "loss": 5.6628,
      "mean_token_accuracy": 0.14084112048149108,
      "num_tokens": 7832184.0,
      "step": 3560
    },
    {
      "entropy": 5.829951238632202,
      "epoch": 0.3538110361254466,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992667931943913,
      "loss": 5.6773,
      "mean_token_accuracy": 0.1355358488857746,
      "num_tokens": 7842983.0,
      "step": 3565
    },
    {
      "entropy": 5.801300001144409,
      "epoch": 0.35430726478761415,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004992639323469293,
      "loss": 5.6128,
      "mean_token_accuracy": 0.14425275325775147,
      "num_tokens": 7854783.0,
      "step": 3570
    },
    {
      "entropy": 5.775907325744629,
      "epoch": 0.3548034934497817,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499261065938205,
      "loss": 5.6231,
      "mean_token_accuracy": 0.14949834495782852,
      "num_tokens": 7864684.0,
      "step": 3575
    },
    {
      "entropy": 5.849696445465088,
      "epoch": 0.3552997221119492,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992581939682895,
      "loss": 5.7852,
      "mean_token_accuracy": 0.12427379041910172,
      "num_tokens": 7876846.0,
      "step": 3580
    },
    {
      "entropy": 5.9438694477081295,
      "epoch": 0.3557959507741167,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499255316437254,
      "loss": 5.6506,
      "mean_token_accuracy": 0.13649129793047904,
      "num_tokens": 7886826.0,
      "step": 3585
    },
    {
      "entropy": 5.746095418930054,
      "epoch": 0.3562921794362842,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992524333451699,
      "loss": 5.7279,
      "mean_token_accuracy": 0.13332464173436165,
      "num_tokens": 7898941.0,
      "step": 3590
    },
    {
      "entropy": 5.789261960983277,
      "epoch": 0.35678840809845175,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992495446921087,
      "loss": 5.6356,
      "mean_token_accuracy": 0.1385039545595646,
      "num_tokens": 7909916.0,
      "step": 3595
    },
    {
      "entropy": 5.799201631546021,
      "epoch": 0.3572846367606193,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004992466504781419,
      "loss": 5.6639,
      "mean_token_accuracy": 0.1378471478819847,
      "num_tokens": 7922137.0,
      "step": 3600
    },
    {
      "entropy": 5.73127179145813,
      "epoch": 0.3577808654227868,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992437507033415,
      "loss": 5.5756,
      "mean_token_accuracy": 0.14868025481700897,
      "num_tokens": 7932940.0,
      "step": 3605
    },
    {
      "entropy": 5.8899187564849855,
      "epoch": 0.35827709408495434,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992408453677792,
      "loss": 5.7552,
      "mean_token_accuracy": 0.12681759595870973,
      "num_tokens": 7943735.0,
      "step": 3610
    },
    {
      "entropy": 5.776127576828003,
      "epoch": 0.35877332274712187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992379344715273,
      "loss": 5.633,
      "mean_token_accuracy": 0.14180014356970788,
      "num_tokens": 7955323.0,
      "step": 3615
    },
    {
      "entropy": 5.764155054092408,
      "epoch": 0.3592695514092894,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992350180146577,
      "loss": 5.6128,
      "mean_token_accuracy": 0.1401304893195629,
      "num_tokens": 7966073.0,
      "step": 3620
    },
    {
      "entropy": 5.823654460906982,
      "epoch": 0.35976578007145693,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992320959972428,
      "loss": 5.7631,
      "mean_token_accuracy": 0.13467661589384078,
      "num_tokens": 7977027.0,
      "step": 3625
    },
    {
      "entropy": 5.789421558380127,
      "epoch": 0.36026200873362446,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004992291684193551,
      "loss": 5.6321,
      "mean_token_accuracy": 0.1438021257519722,
      "num_tokens": 7987368.0,
      "step": 3630
    },
    {
      "entropy": 5.857334232330322,
      "epoch": 0.360758237395792,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004992262352810672,
      "loss": 5.7018,
      "mean_token_accuracy": 0.13492405116558076,
      "num_tokens": 8000060.0,
      "step": 3635
    },
    {
      "entropy": 5.858621263504029,
      "epoch": 0.3612544660579595,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992232965824518,
      "loss": 5.7432,
      "mean_token_accuracy": 0.13379527181386947,
      "num_tokens": 8011388.0,
      "step": 3640
    },
    {
      "entropy": 5.75277156829834,
      "epoch": 0.36175069472012705,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499220352323582,
      "loss": 5.6126,
      "mean_token_accuracy": 0.14420773088932037,
      "num_tokens": 8021682.0,
      "step": 3645
    },
    {
      "entropy": 5.801869106292725,
      "epoch": 0.3622469233822946,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992174025045304,
      "loss": 5.6393,
      "mean_token_accuracy": 0.14032537564635278,
      "num_tokens": 8032799.0,
      "step": 3650
    },
    {
      "entropy": 5.752224397659302,
      "epoch": 0.3627431520444621,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992144471253704,
      "loss": 5.5225,
      "mean_token_accuracy": 0.14599559903144838,
      "num_tokens": 8042742.0,
      "step": 3655
    },
    {
      "entropy": 5.81301326751709,
      "epoch": 0.3632393807066296,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004992114861861753,
      "loss": 5.6986,
      "mean_token_accuracy": 0.13800043836236,
      "num_tokens": 8054068.0,
      "step": 3660
    },
    {
      "entropy": 5.849695634841919,
      "epoch": 0.3637356093687971,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004992085196870184,
      "loss": 5.6822,
      "mean_token_accuracy": 0.13372902050614358,
      "num_tokens": 8065286.0,
      "step": 3665
    },
    {
      "entropy": 5.7867927074432375,
      "epoch": 0.36423183803096465,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992055476279734,
      "loss": 5.7418,
      "mean_token_accuracy": 0.12366731837391853,
      "num_tokens": 8077122.0,
      "step": 3670
    },
    {
      "entropy": 5.8696094989776615,
      "epoch": 0.3647280666931322,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004992025700091139,
      "loss": 5.6782,
      "mean_token_accuracy": 0.14306551367044448,
      "num_tokens": 8087036.0,
      "step": 3675
    },
    {
      "entropy": 5.827376890182495,
      "epoch": 0.3652242953552997,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991995868305138,
      "loss": 5.7159,
      "mean_token_accuracy": 0.13650970607995988,
      "num_tokens": 8098517.0,
      "step": 3680
    },
    {
      "entropy": 5.739241123199463,
      "epoch": 0.36572052401746724,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991965980922471,
      "loss": 5.5848,
      "mean_token_accuracy": 0.1435575470328331,
      "num_tokens": 8107475.0,
      "step": 3685
    },
    {
      "entropy": 5.760336208343506,
      "epoch": 0.36621675267963477,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004991936037943878,
      "loss": 5.6944,
      "mean_token_accuracy": 0.13551692590117453,
      "num_tokens": 8118529.0,
      "step": 3690
    },
    {
      "entropy": 5.863595008850098,
      "epoch": 0.3667129813418023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991906039370104,
      "loss": 5.6001,
      "mean_token_accuracy": 0.14009833559393883,
      "num_tokens": 8128951.0,
      "step": 3695
    },
    {
      "entropy": 5.708594083786011,
      "epoch": 0.36720921000396983,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991875985201889,
      "loss": 5.6132,
      "mean_token_accuracy": 0.14244235828518867,
      "num_tokens": 8139385.0,
      "step": 3700
    },
    {
      "entropy": 5.8222747325897215,
      "epoch": 0.36770543866613736,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499184587543998,
      "loss": 5.6545,
      "mean_token_accuracy": 0.14134933277964593,
      "num_tokens": 8149376.0,
      "step": 3705
    },
    {
      "entropy": 5.854859828948975,
      "epoch": 0.3682016673283049,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991815710085126,
      "loss": 5.6847,
      "mean_token_accuracy": 0.13634160533547401,
      "num_tokens": 8160687.0,
      "step": 3710
    },
    {
      "entropy": 5.803466701507569,
      "epoch": 0.3686978959904724,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004991785489138072,
      "loss": 5.6818,
      "mean_token_accuracy": 0.1403013363480568,
      "num_tokens": 8171762.0,
      "step": 3715
    },
    {
      "entropy": 5.721362352371216,
      "epoch": 0.36919412465263995,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991755212599569,
      "loss": 5.4577,
      "mean_token_accuracy": 0.14683803170919418,
      "num_tokens": 8181643.0,
      "step": 3720
    },
    {
      "entropy": 5.672873020172119,
      "epoch": 0.3696903533148075,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991724880470366,
      "loss": 5.7064,
      "mean_token_accuracy": 0.1410878300666809,
      "num_tokens": 8192692.0,
      "step": 3725
    },
    {
      "entropy": 5.838234233856201,
      "epoch": 0.370186581976975,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991694492751217,
      "loss": 5.6497,
      "mean_token_accuracy": 0.13540541008114815,
      "num_tokens": 8202941.0,
      "step": 3730
    },
    {
      "entropy": 5.778851127624511,
      "epoch": 0.3706828106391425,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004991664049442876,
      "loss": 5.6432,
      "mean_token_accuracy": 0.1387119397521019,
      "num_tokens": 8214206.0,
      "step": 3735
    },
    {
      "entropy": 5.730884981155396,
      "epoch": 0.37117903930131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991633550546097,
      "loss": 5.4762,
      "mean_token_accuracy": 0.15001685842871665,
      "num_tokens": 8223672.0,
      "step": 3740
    },
    {
      "entropy": 5.768949842453003,
      "epoch": 0.37167526796347755,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991602996061635,
      "loss": 5.6247,
      "mean_token_accuracy": 0.1420579731464386,
      "num_tokens": 8234189.0,
      "step": 3745
    },
    {
      "entropy": 5.854026174545288,
      "epoch": 0.3721714966256451,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991572385990249,
      "loss": 5.6554,
      "mean_token_accuracy": 0.1406680442392826,
      "num_tokens": 8245382.0,
      "step": 3750
    },
    {
      "entropy": 5.804176998138428,
      "epoch": 0.3726677252878126,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991541720332698,
      "loss": 5.6099,
      "mean_token_accuracy": 0.1439274363219738,
      "num_tokens": 8254530.0,
      "step": 3755
    },
    {
      "entropy": 5.7249938488006595,
      "epoch": 0.37316395394998014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991510999089743,
      "loss": 5.6537,
      "mean_token_accuracy": 0.14627458304166793,
      "num_tokens": 8264030.0,
      "step": 3760
    },
    {
      "entropy": 5.750170564651489,
      "epoch": 0.3736601826121477,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991480222262145,
      "loss": 5.6295,
      "mean_token_accuracy": 0.14486483335494996,
      "num_tokens": 8273537.0,
      "step": 3765
    },
    {
      "entropy": 5.831386995315552,
      "epoch": 0.3741564112743152,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004991449389850668,
      "loss": 5.6716,
      "mean_token_accuracy": 0.13380085602402686,
      "num_tokens": 8286123.0,
      "step": 3770
    },
    {
      "entropy": 5.782286834716797,
      "epoch": 0.37465263993648273,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991418501856076,
      "loss": 5.6734,
      "mean_token_accuracy": 0.13890725746750832,
      "num_tokens": 8298115.0,
      "step": 3775
    },
    {
      "entropy": 5.843253946304321,
      "epoch": 0.37514886859865026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991387558279134,
      "loss": 5.688,
      "mean_token_accuracy": 0.13645929098129272,
      "num_tokens": 8309376.0,
      "step": 3780
    },
    {
      "entropy": 5.7840159893035885,
      "epoch": 0.3756450972608178,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991356559120612,
      "loss": 5.6475,
      "mean_token_accuracy": 0.14034804850816726,
      "num_tokens": 8319272.0,
      "step": 3785
    },
    {
      "entropy": 5.77095308303833,
      "epoch": 0.3761413259229853,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991325504381275,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1374574065208435,
      "num_tokens": 8330042.0,
      "step": 3790
    },
    {
      "entropy": 5.779422807693481,
      "epoch": 0.37663755458515286,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991294394061896,
      "loss": 5.6021,
      "mean_token_accuracy": 0.13860200792551042,
      "num_tokens": 8340394.0,
      "step": 3795
    },
    {
      "entropy": 5.723526668548584,
      "epoch": 0.3771337832473204,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004991263228163246,
      "loss": 5.642,
      "mean_token_accuracy": 0.13743859454989432,
      "num_tokens": 8351103.0,
      "step": 3800
    },
    {
      "entropy": 5.808193016052246,
      "epoch": 0.3776300119094879,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004991232006686096,
      "loss": 5.6811,
      "mean_token_accuracy": 0.1397139772772789,
      "num_tokens": 8363215.0,
      "step": 3805
    },
    {
      "entropy": 5.881165933609009,
      "epoch": 0.3781262405716554,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004991200729631223,
      "loss": 5.7125,
      "mean_token_accuracy": 0.12982109040021897,
      "num_tokens": 8374602.0,
      "step": 3810
    },
    {
      "entropy": 5.758335256576538,
      "epoch": 0.3786224692338229,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991169396999401,
      "loss": 5.6376,
      "mean_token_accuracy": 0.13426343947649003,
      "num_tokens": 8386470.0,
      "step": 3815
    },
    {
      "entropy": 5.800685262680053,
      "epoch": 0.37911869789599045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991138008791408,
      "loss": 5.6486,
      "mean_token_accuracy": 0.14015363454818724,
      "num_tokens": 8396284.0,
      "step": 3820
    },
    {
      "entropy": 5.8608063697814945,
      "epoch": 0.379614926558158,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991106565008021,
      "loss": 5.6744,
      "mean_token_accuracy": 0.13724475055933,
      "num_tokens": 8407856.0,
      "step": 3825
    },
    {
      "entropy": 5.797486114501953,
      "epoch": 0.3801111552203255,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004991075065650019,
      "loss": 5.5876,
      "mean_token_accuracy": 0.14327726662158966,
      "num_tokens": 8419560.0,
      "step": 3830
    },
    {
      "entropy": 5.7997008800506595,
      "epoch": 0.38060738388249304,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991043510718186,
      "loss": 5.5831,
      "mean_token_accuracy": 0.14024922102689744,
      "num_tokens": 8430594.0,
      "step": 3835
    },
    {
      "entropy": 5.714972496032715,
      "epoch": 0.3811036125446606,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004991011900213303,
      "loss": 5.5476,
      "mean_token_accuracy": 0.1445930302143097,
      "num_tokens": 8440549.0,
      "step": 3840
    },
    {
      "entropy": 5.737625694274902,
      "epoch": 0.3815998412068281,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990980234136152,
      "loss": 5.6006,
      "mean_token_accuracy": 0.14232855290174484,
      "num_tokens": 8451553.0,
      "step": 3845
    },
    {
      "entropy": 5.765847969055176,
      "epoch": 0.38209606986899564,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990948512487522,
      "loss": 5.6025,
      "mean_token_accuracy": 0.14687582924962045,
      "num_tokens": 8462899.0,
      "step": 3850
    },
    {
      "entropy": 5.786074161529541,
      "epoch": 0.38259229853116317,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990916735268198,
      "loss": 5.6439,
      "mean_token_accuracy": 0.14027897045016288,
      "num_tokens": 8474566.0,
      "step": 3855
    },
    {
      "entropy": 5.82865571975708,
      "epoch": 0.3830885271933307,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004990884902478966,
      "loss": 5.6542,
      "mean_token_accuracy": 0.14067775905132293,
      "num_tokens": 8484786.0,
      "step": 3860
    },
    {
      "entropy": 5.737159204483032,
      "epoch": 0.3835847558554982,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004990853014120619,
      "loss": 5.584,
      "mean_token_accuracy": 0.14371388256549836,
      "num_tokens": 8496416.0,
      "step": 3865
    },
    {
      "entropy": 5.7627674579620365,
      "epoch": 0.38408098451766576,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990821070193944,
      "loss": 5.618,
      "mean_token_accuracy": 0.13939571902155876,
      "num_tokens": 8507287.0,
      "step": 3870
    },
    {
      "entropy": 5.808990955352783,
      "epoch": 0.3845772131798333,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990789070699736,
      "loss": 5.6644,
      "mean_token_accuracy": 0.1379352405667305,
      "num_tokens": 8517922.0,
      "step": 3875
    },
    {
      "entropy": 5.746287488937378,
      "epoch": 0.3850734418420008,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990757015638787,
      "loss": 5.5966,
      "mean_token_accuracy": 0.14340450689196588,
      "num_tokens": 8529532.0,
      "step": 3880
    },
    {
      "entropy": 5.7676066875457765,
      "epoch": 0.3855696705041683,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990724905011894,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1429295666515827,
      "num_tokens": 8540207.0,
      "step": 3885
    },
    {
      "entropy": 5.785931348800659,
      "epoch": 0.3860658991663358,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499069273881985,
      "loss": 5.6022,
      "mean_token_accuracy": 0.1414462760090828,
      "num_tokens": 8551634.0,
      "step": 3890
    },
    {
      "entropy": 5.780277299880981,
      "epoch": 0.38656212782850335,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004990660517063455,
      "loss": 5.6715,
      "mean_token_accuracy": 0.13274993374943733,
      "num_tokens": 8563256.0,
      "step": 3895
    },
    {
      "entropy": 5.78124361038208,
      "epoch": 0.3870583564906709,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990628239743508,
      "loss": 5.6826,
      "mean_token_accuracy": 0.1389344312250614,
      "num_tokens": 8573833.0,
      "step": 3900
    },
    {
      "entropy": 5.764538526535034,
      "epoch": 0.3875545851528384,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990595906860807,
      "loss": 5.5955,
      "mean_token_accuracy": 0.14482991695404052,
      "num_tokens": 8584283.0,
      "step": 3905
    },
    {
      "entropy": 5.716322326660157,
      "epoch": 0.38805081381500595,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990563518416157,
      "loss": 5.6519,
      "mean_token_accuracy": 0.1405915029346943,
      "num_tokens": 8595344.0,
      "step": 3910
    },
    {
      "entropy": 5.803377866744995,
      "epoch": 0.3885470424771735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990531074410358,
      "loss": 5.5732,
      "mean_token_accuracy": 0.13988425508141517,
      "num_tokens": 8605625.0,
      "step": 3915
    },
    {
      "entropy": 5.761973237991333,
      "epoch": 0.389043271139341,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990498574844218,
      "loss": 5.5584,
      "mean_token_accuracy": 0.14352284371852875,
      "num_tokens": 8616259.0,
      "step": 3920
    },
    {
      "entropy": 5.779136419296265,
      "epoch": 0.38953949980150854,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990466019718542,
      "loss": 5.5826,
      "mean_token_accuracy": 0.14573069661855698,
      "num_tokens": 8627179.0,
      "step": 3925
    },
    {
      "entropy": 5.787642526626587,
      "epoch": 0.39003572846367607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990433409034134,
      "loss": 5.6789,
      "mean_token_accuracy": 0.1468874804675579,
      "num_tokens": 8637448.0,
      "step": 3930
    },
    {
      "entropy": 5.69730954170227,
      "epoch": 0.3905319571258436,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990400742791806,
      "loss": 5.6063,
      "mean_token_accuracy": 0.14139263927936555,
      "num_tokens": 8649168.0,
      "step": 3935
    },
    {
      "entropy": 5.844192600250244,
      "epoch": 0.39102818578801113,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004990368020992368,
      "loss": 5.703,
      "mean_token_accuracy": 0.13631367385387422,
      "num_tokens": 8660653.0,
      "step": 3940
    },
    {
      "entropy": 5.797797536849975,
      "epoch": 0.39152441445017866,
      "grad_norm": 0.875,
      "learning_rate": 0.000499033524363663,
      "loss": 5.6704,
      "mean_token_accuracy": 0.144366754591465,
      "num_tokens": 8672000.0,
      "step": 3945
    },
    {
      "entropy": 5.731224822998047,
      "epoch": 0.3920206431123462,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990302410725405,
      "loss": 5.604,
      "mean_token_accuracy": 0.1445721998810768,
      "num_tokens": 8682455.0,
      "step": 3950
    },
    {
      "entropy": 5.75863299369812,
      "epoch": 0.3925168717745137,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990269522259509,
      "loss": 5.5723,
      "mean_token_accuracy": 0.14195302575826646,
      "num_tokens": 8692975.0,
      "step": 3955
    },
    {
      "entropy": 5.718239879608154,
      "epoch": 0.3930131004366812,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990236578239755,
      "loss": 5.5478,
      "mean_token_accuracy": 0.14248109981417656,
      "num_tokens": 8703569.0,
      "step": 3960
    },
    {
      "entropy": 5.799000358581543,
      "epoch": 0.3935093290988487,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004990203578666962,
      "loss": 5.6646,
      "mean_token_accuracy": 0.13780656680464745,
      "num_tokens": 8713734.0,
      "step": 3965
    },
    {
      "entropy": 5.760756349563598,
      "epoch": 0.39400555776101626,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990170523541948,
      "loss": 5.5569,
      "mean_token_accuracy": 0.14608304128050803,
      "num_tokens": 8723852.0,
      "step": 3970
    },
    {
      "entropy": 5.771544027328491,
      "epoch": 0.3945017864231838,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004990137412865531,
      "loss": 5.71,
      "mean_token_accuracy": 0.14007417038083075,
      "num_tokens": 8734300.0,
      "step": 3975
    },
    {
      "entropy": 5.848857593536377,
      "epoch": 0.3949980150853513,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990104246638533,
      "loss": 5.7202,
      "mean_token_accuracy": 0.13323673605918884,
      "num_tokens": 8745022.0,
      "step": 3980
    },
    {
      "entropy": 5.809627294540405,
      "epoch": 0.39549424374751885,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990071024861778,
      "loss": 5.5961,
      "mean_token_accuracy": 0.14009736552834512,
      "num_tokens": 8754870.0,
      "step": 3985
    },
    {
      "entropy": 5.842597007751465,
      "epoch": 0.3959904724096864,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004990037747536087,
      "loss": 5.6527,
      "mean_token_accuracy": 0.14414928182959558,
      "num_tokens": 8765536.0,
      "step": 3990
    },
    {
      "entropy": 5.751135635375976,
      "epoch": 0.3964867010718539,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990004414662288,
      "loss": 5.6229,
      "mean_token_accuracy": 0.1373239904642105,
      "num_tokens": 8776898.0,
      "step": 3995
    },
    {
      "entropy": 5.7811463356018065,
      "epoch": 0.39698292973402144,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989971026241205,
      "loss": 5.6857,
      "mean_token_accuracy": 0.13665838539600372,
      "num_tokens": 8788176.0,
      "step": 4000
    },
    {
      "entropy": 5.726622819900513,
      "epoch": 0.39747915839618897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989937582273668,
      "loss": 5.6103,
      "mean_token_accuracy": 0.13971234038472174,
      "num_tokens": 8798939.0,
      "step": 4005
    },
    {
      "entropy": 5.74769024848938,
      "epoch": 0.3979753870583565,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989904082760506,
      "loss": 5.6462,
      "mean_token_accuracy": 0.14307962059974672,
      "num_tokens": 8809092.0,
      "step": 4010
    },
    {
      "entropy": 5.775501155853272,
      "epoch": 0.39847161572052403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000498987052770255,
      "loss": 5.5599,
      "mean_token_accuracy": 0.1454466737806797,
      "num_tokens": 8819352.0,
      "step": 4015
    },
    {
      "entropy": 5.769831609725952,
      "epoch": 0.39896784438269156,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498983691710063,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1373866803944111,
      "num_tokens": 8829878.0,
      "step": 4020
    },
    {
      "entropy": 5.749059391021729,
      "epoch": 0.3994640730448591,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498980325095558,
      "loss": 5.5516,
      "mean_token_accuracy": 0.14222921505570413,
      "num_tokens": 8840364.0,
      "step": 4025
    },
    {
      "entropy": 5.767441511154175,
      "epoch": 0.3999603017070266,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989769529268238,
      "loss": 5.6677,
      "mean_token_accuracy": 0.13673522099852561,
      "num_tokens": 8851340.0,
      "step": 4030
    },
    {
      "entropy": 5.749474573135376,
      "epoch": 0.4004565303691941,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989735752039436,
      "loss": 5.6345,
      "mean_token_accuracy": 0.1377050243318081,
      "num_tokens": 8862571.0,
      "step": 4035
    },
    {
      "entropy": 5.774168300628662,
      "epoch": 0.4009527590313616,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989701919270015,
      "loss": 5.6335,
      "mean_token_accuracy": 0.14452187493443489,
      "num_tokens": 8873927.0,
      "step": 4040
    },
    {
      "entropy": 5.838969755172729,
      "epoch": 0.40144898769352916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989668030960812,
      "loss": 5.831,
      "mean_token_accuracy": 0.1299641877412796,
      "num_tokens": 8886542.0,
      "step": 4045
    },
    {
      "entropy": 5.8212137699127195,
      "epoch": 0.4019452163556967,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004989634087112667,
      "loss": 5.5686,
      "mean_token_accuracy": 0.13974880948662757,
      "num_tokens": 8896114.0,
      "step": 4050
    },
    {
      "entropy": 5.77008090019226,
      "epoch": 0.4024414450178642,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989600087726425,
      "loss": 5.6514,
      "mean_token_accuracy": 0.13421038687229156,
      "num_tokens": 8906303.0,
      "step": 4055
    },
    {
      "entropy": 5.800097465515137,
      "epoch": 0.40293767368003175,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004989566032802924,
      "loss": 5.6815,
      "mean_token_accuracy": 0.13664470240473747,
      "num_tokens": 8916961.0,
      "step": 4060
    },
    {
      "entropy": 5.839651823043823,
      "epoch": 0.4034339023421993,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989531922343013,
      "loss": 5.7699,
      "mean_token_accuracy": 0.13481825292110444,
      "num_tokens": 8927700.0,
      "step": 4065
    },
    {
      "entropy": 5.787808132171631,
      "epoch": 0.4039301310043668,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004989497756347534,
      "loss": 5.6065,
      "mean_token_accuracy": 0.14093791767954827,
      "num_tokens": 8939655.0,
      "step": 4070
    },
    {
      "entropy": 5.725460624694824,
      "epoch": 0.40442635966653434,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004989463534817339,
      "loss": 5.5419,
      "mean_token_accuracy": 0.14649698138237,
      "num_tokens": 8951238.0,
      "step": 4075
    },
    {
      "entropy": 5.797673511505127,
      "epoch": 0.40492258832870187,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989429257753273,
      "loss": 5.6068,
      "mean_token_accuracy": 0.1479934424161911,
      "num_tokens": 8962019.0,
      "step": 4080
    },
    {
      "entropy": 5.749550437927246,
      "epoch": 0.4054188169908694,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004989394925156187,
      "loss": 5.6114,
      "mean_token_accuracy": 0.1434945769608021,
      "num_tokens": 8973861.0,
      "step": 4085
    },
    {
      "entropy": 5.706510066986084,
      "epoch": 0.40591504565303693,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004989360537026932,
      "loss": 5.6103,
      "mean_token_accuracy": 0.1431720182299614,
      "num_tokens": 8985981.0,
      "step": 4090
    },
    {
      "entropy": 5.760543870925903,
      "epoch": 0.40641127431520446,
      "grad_norm": 0.90625,
      "learning_rate": 0.000498932609336636,
      "loss": 5.4961,
      "mean_token_accuracy": 0.14572610855102539,
      "num_tokens": 8995318.0,
      "step": 4095
    },
    {
      "entropy": 5.7298054695129395,
      "epoch": 0.406907502977372,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989291594175327,
      "loss": 5.5462,
      "mean_token_accuracy": 0.14044902995228767,
      "num_tokens": 9004679.0,
      "step": 4100
    },
    {
      "entropy": 5.880003118515015,
      "epoch": 0.4074037316395395,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989257039454689,
      "loss": 5.6814,
      "mean_token_accuracy": 0.14026053249835968,
      "num_tokens": 9016422.0,
      "step": 4105
    },
    {
      "entropy": 5.727775716781617,
      "epoch": 0.40789996030170705,
      "grad_norm": 0.859375,
      "learning_rate": 0.00049892224292053,
      "loss": 5.5937,
      "mean_token_accuracy": 0.14182559922337531,
      "num_tokens": 9027914.0,
      "step": 4110
    },
    {
      "entropy": 5.742984390258789,
      "epoch": 0.40839618896387453,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004989187763428021,
      "loss": 5.5636,
      "mean_token_accuracy": 0.1467101253569126,
      "num_tokens": 9038701.0,
      "step": 4115
    },
    {
      "entropy": 5.701486301422119,
      "epoch": 0.40889241762604206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498915304212371,
      "loss": 5.6182,
      "mean_token_accuracy": 0.13621484637260436,
      "num_tokens": 9050620.0,
      "step": 4120
    },
    {
      "entropy": 5.775696706771851,
      "epoch": 0.4093886462882096,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004989118265293228,
      "loss": 5.655,
      "mean_token_accuracy": 0.14236382097005845,
      "num_tokens": 9060900.0,
      "step": 4125
    },
    {
      "entropy": 5.7544715881347654,
      "epoch": 0.4098848749503771,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004989083432937439,
      "loss": 5.5758,
      "mean_token_accuracy": 0.1405758798122406,
      "num_tokens": 9072122.0,
      "step": 4130
    },
    {
      "entropy": 5.742540979385376,
      "epoch": 0.41038110361254465,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004989048545057204,
      "loss": 5.5826,
      "mean_token_accuracy": 0.1416861116886139,
      "num_tokens": 9082780.0,
      "step": 4135
    },
    {
      "entropy": 5.752718877792359,
      "epoch": 0.4108773322747122,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989013601653391,
      "loss": 5.6289,
      "mean_token_accuracy": 0.13903876021504402,
      "num_tokens": 9093518.0,
      "step": 4140
    },
    {
      "entropy": 5.7312267303466795,
      "epoch": 0.4113735609368797,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988978602726866,
      "loss": 5.6576,
      "mean_token_accuracy": 0.1419159837067127,
      "num_tokens": 9105035.0,
      "step": 4145
    },
    {
      "entropy": 5.741264677047729,
      "epoch": 0.41186978959904724,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004988943548278494,
      "loss": 5.5582,
      "mean_token_accuracy": 0.1427288256585598,
      "num_tokens": 9117154.0,
      "step": 4150
    },
    {
      "entropy": 5.755410146713257,
      "epoch": 0.4123660182612148,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004988908438309149,
      "loss": 5.6359,
      "mean_token_accuracy": 0.14003294780850412,
      "num_tokens": 9128818.0,
      "step": 4155
    },
    {
      "entropy": 5.80935230255127,
      "epoch": 0.4128622469233823,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004988873272819697,
      "loss": 5.6466,
      "mean_token_accuracy": 0.14533814191818237,
      "num_tokens": 9140188.0,
      "step": 4160
    },
    {
      "entropy": 5.893560743331909,
      "epoch": 0.41335847558554983,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988838051811014,
      "loss": 5.7204,
      "mean_token_accuracy": 0.13877229392528534,
      "num_tokens": 9150565.0,
      "step": 4165
    },
    {
      "entropy": 5.745661354064941,
      "epoch": 0.41385470424771736,
      "grad_norm": 0.9375,
      "learning_rate": 0.000498880277528397,
      "loss": 5.5804,
      "mean_token_accuracy": 0.14375061690807342,
      "num_tokens": 9160130.0,
      "step": 4170
    },
    {
      "entropy": 5.754231119155884,
      "epoch": 0.4143509329098849,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988767443239443,
      "loss": 5.571,
      "mean_token_accuracy": 0.14534730836749077,
      "num_tokens": 9171740.0,
      "step": 4175
    },
    {
      "entropy": 5.717862701416015,
      "epoch": 0.4148471615720524,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004988732055678307,
      "loss": 5.5199,
      "mean_token_accuracy": 0.15634380429983138,
      "num_tokens": 9182459.0,
      "step": 4180
    },
    {
      "entropy": 5.786018323898316,
      "epoch": 0.41534339023421996,
      "grad_norm": 0.890625,
      "learning_rate": 0.000498869661260144,
      "loss": 5.5884,
      "mean_token_accuracy": 0.14363244622945787,
      "num_tokens": 9192284.0,
      "step": 4185
    },
    {
      "entropy": 5.728206729888916,
      "epoch": 0.41583961889638743,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498866111400972,
      "loss": 5.5836,
      "mean_token_accuracy": 0.14470189064741135,
      "num_tokens": 9202700.0,
      "step": 4190
    },
    {
      "entropy": 5.72704291343689,
      "epoch": 0.41633584755855496,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988625559904029,
      "loss": 5.6021,
      "mean_token_accuracy": 0.13874267488718034,
      "num_tokens": 9213013.0,
      "step": 4195
    },
    {
      "entropy": 5.800160312652588,
      "epoch": 0.4168320762207225,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004988589950285248,
      "loss": 5.6744,
      "mean_token_accuracy": 0.139788468927145,
      "num_tokens": 9224568.0,
      "step": 4200
    },
    {
      "entropy": 5.701272773742676,
      "epoch": 0.41732830488289,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988554285154259,
      "loss": 5.5374,
      "mean_token_accuracy": 0.14081604033708572,
      "num_tokens": 9233791.0,
      "step": 4205
    },
    {
      "entropy": 5.765557289123535,
      "epoch": 0.41782453354505755,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988518564511948,
      "loss": 5.612,
      "mean_token_accuracy": 0.13587745726108552,
      "num_tokens": 9245048.0,
      "step": 4210
    },
    {
      "entropy": 5.812022590637207,
      "epoch": 0.4183207622072251,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988482788359202,
      "loss": 5.609,
      "mean_token_accuracy": 0.1474679045379162,
      "num_tokens": 9256508.0,
      "step": 4215
    },
    {
      "entropy": 5.726398229598999,
      "epoch": 0.4188169908693926,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004988446956696904,
      "loss": 5.6384,
      "mean_token_accuracy": 0.14110739529132843,
      "num_tokens": 9267871.0,
      "step": 4220
    },
    {
      "entropy": 5.8059971809387205,
      "epoch": 0.41931321953156014,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988411069525947,
      "loss": 5.7039,
      "mean_token_accuracy": 0.13741991817951202,
      "num_tokens": 9279287.0,
      "step": 4225
    },
    {
      "entropy": 5.858490753173828,
      "epoch": 0.4198094481937277,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988375126847216,
      "loss": 5.7035,
      "mean_token_accuracy": 0.13470567613840104,
      "num_tokens": 9291257.0,
      "step": 4230
    },
    {
      "entropy": 5.822464179992676,
      "epoch": 0.4203056768558952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988339128661607,
      "loss": 5.6692,
      "mean_token_accuracy": 0.13757289052009583,
      "num_tokens": 9302590.0,
      "step": 4235
    },
    {
      "entropy": 5.662470579147339,
      "epoch": 0.42080190551806274,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000498830307497001,
      "loss": 5.5275,
      "mean_token_accuracy": 0.14331217855215073,
      "num_tokens": 9313174.0,
      "step": 4240
    },
    {
      "entropy": 5.739015722274781,
      "epoch": 0.42129813418023027,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498826696577332,
      "loss": 5.6218,
      "mean_token_accuracy": 0.14237825125455855,
      "num_tokens": 9325034.0,
      "step": 4245
    },
    {
      "entropy": 5.795171403884888,
      "epoch": 0.4217943628423978,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988230801072432,
      "loss": 5.5942,
      "mean_token_accuracy": 0.1420249953866005,
      "num_tokens": 9336834.0,
      "step": 4250
    },
    {
      "entropy": 5.717333030700684,
      "epoch": 0.4222905915045653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988194580868244,
      "loss": 5.6078,
      "mean_token_accuracy": 0.14135593175888062,
      "num_tokens": 9347747.0,
      "step": 4255
    },
    {
      "entropy": 5.720476293563843,
      "epoch": 0.42278682016673286,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004988158305161654,
      "loss": 5.6047,
      "mean_token_accuracy": 0.14483303427696229,
      "num_tokens": 9359186.0,
      "step": 4260
    },
    {
      "entropy": 5.750612545013428,
      "epoch": 0.42328304882890033,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000498812197395356,
      "loss": 5.5335,
      "mean_token_accuracy": 0.14376334547996522,
      "num_tokens": 9369485.0,
      "step": 4265
    },
    {
      "entropy": 5.791775417327881,
      "epoch": 0.42377927749106786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988085587244863,
      "loss": 5.6396,
      "mean_token_accuracy": 0.13861620277166367,
      "num_tokens": 9382658.0,
      "step": 4270
    },
    {
      "entropy": 5.711907196044922,
      "epoch": 0.4242755061532354,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004988049145036466,
      "loss": 5.5891,
      "mean_token_accuracy": 0.13826264888048173,
      "num_tokens": 9394151.0,
      "step": 4275
    },
    {
      "entropy": 5.8195891857147215,
      "epoch": 0.4247717348154029,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988012647329273,
      "loss": 5.6811,
      "mean_token_accuracy": 0.1363990142941475,
      "num_tokens": 9405956.0,
      "step": 4280
    },
    {
      "entropy": 5.767521047592163,
      "epoch": 0.42526796347757045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498797609412419,
      "loss": 5.568,
      "mean_token_accuracy": 0.13977680504322051,
      "num_tokens": 9416889.0,
      "step": 4285
    },
    {
      "entropy": 5.776399183273315,
      "epoch": 0.425764192139738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987939485422122,
      "loss": 5.676,
      "mean_token_accuracy": 0.13941217437386513,
      "num_tokens": 9427602.0,
      "step": 4290
    },
    {
      "entropy": 5.7858906269073485,
      "epoch": 0.4262604208019055,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004987902821223976,
      "loss": 5.5936,
      "mean_token_accuracy": 0.14471960067749023,
      "num_tokens": 9438557.0,
      "step": 4295
    },
    {
      "entropy": 5.720623922348023,
      "epoch": 0.42675664946407305,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004987866101530663,
      "loss": 5.6526,
      "mean_token_accuracy": 0.13875658437609673,
      "num_tokens": 9449598.0,
      "step": 4300
    },
    {
      "entropy": 5.74213547706604,
      "epoch": 0.4272528781262406,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004987829326343092,
      "loss": 5.5739,
      "mean_token_accuracy": 0.14770149141550065,
      "num_tokens": 9459740.0,
      "step": 4305
    },
    {
      "entropy": 5.751595497131348,
      "epoch": 0.4277491067884081,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987792495662177,
      "loss": 5.5875,
      "mean_token_accuracy": 0.1395449496805668,
      "num_tokens": 9469835.0,
      "step": 4310
    },
    {
      "entropy": 5.778836297988891,
      "epoch": 0.42824533545057564,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498775560948883,
      "loss": 5.6679,
      "mean_token_accuracy": 0.14196610748767852,
      "num_tokens": 9480735.0,
      "step": 4315
    },
    {
      "entropy": 5.751088190078735,
      "epoch": 0.42874156411274317,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987718667823965,
      "loss": 5.553,
      "mean_token_accuracy": 0.14639294296503066,
      "num_tokens": 9491321.0,
      "step": 4320
    },
    {
      "entropy": 5.78062949180603,
      "epoch": 0.4292377927749107,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00049876816706685,
      "loss": 5.6959,
      "mean_token_accuracy": 0.13560934662818908,
      "num_tokens": 9503507.0,
      "step": 4325
    },
    {
      "entropy": 5.804309272766114,
      "epoch": 0.42973402143707823,
      "grad_norm": 1.7734375,
      "learning_rate": 0.000498764461802335,
      "loss": 5.6855,
      "mean_token_accuracy": 0.14237746670842172,
      "num_tokens": 9514404.0,
      "step": 4330
    },
    {
      "entropy": 5.7196800231933596,
      "epoch": 0.43023025009924576,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004987607509889437,
      "loss": 5.5732,
      "mean_token_accuracy": 0.14124854281544685,
      "num_tokens": 9525555.0,
      "step": 4335
    },
    {
      "entropy": 5.751436376571656,
      "epoch": 0.43072647876141323,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004987570346267677,
      "loss": 5.5479,
      "mean_token_accuracy": 0.14677441567182542,
      "num_tokens": 9535084.0,
      "step": 4340
    },
    {
      "entropy": 5.703538227081299,
      "epoch": 0.43122270742358076,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987533127158996,
      "loss": 5.5771,
      "mean_token_accuracy": 0.1481661930680275,
      "num_tokens": 9546170.0,
      "step": 4345
    },
    {
      "entropy": 5.749481678009033,
      "epoch": 0.4317189360857483,
      "grad_norm": 0.875,
      "learning_rate": 0.0004987495852564314,
      "loss": 5.5798,
      "mean_token_accuracy": 0.14525138661265374,
      "num_tokens": 9556793.0,
      "step": 4350
    },
    {
      "entropy": 5.748826742172241,
      "epoch": 0.4322151647479158,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987458522484556,
      "loss": 5.538,
      "mean_token_accuracy": 0.14604015201330184,
      "num_tokens": 9568114.0,
      "step": 4355
    },
    {
      "entropy": 5.746814250946045,
      "epoch": 0.43271139341008336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987421136920648,
      "loss": 5.5481,
      "mean_token_accuracy": 0.13940781205892563,
      "num_tokens": 9579119.0,
      "step": 4360
    },
    {
      "entropy": 5.661113643646241,
      "epoch": 0.4332076220722509,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987383695873517,
      "loss": 5.5826,
      "mean_token_accuracy": 0.14255948811769487,
      "num_tokens": 9590259.0,
      "step": 4365
    },
    {
      "entropy": 5.782752943038941,
      "epoch": 0.4337038507344184,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987346199344092,
      "loss": 5.6206,
      "mean_token_accuracy": 0.13845986723899842,
      "num_tokens": 9600431.0,
      "step": 4370
    },
    {
      "entropy": 5.7111358642578125,
      "epoch": 0.43420007939658595,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004987308647333303,
      "loss": 5.5995,
      "mean_token_accuracy": 0.142484463006258,
      "num_tokens": 9610286.0,
      "step": 4375
    },
    {
      "entropy": 5.817176818847656,
      "epoch": 0.4346963080587535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987271039842079,
      "loss": 5.6225,
      "mean_token_accuracy": 0.1391633003950119,
      "num_tokens": 9621148.0,
      "step": 4380
    },
    {
      "entropy": 5.740184688568116,
      "epoch": 0.435192536720921,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004987233376871355,
      "loss": 5.5045,
      "mean_token_accuracy": 0.14313893243670464,
      "num_tokens": 9631300.0,
      "step": 4385
    },
    {
      "entropy": 5.72639217376709,
      "epoch": 0.43568876538308854,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987195658422063,
      "loss": 5.6027,
      "mean_token_accuracy": 0.13765663132071496,
      "num_tokens": 9641986.0,
      "step": 4390
    },
    {
      "entropy": 5.783410167694091,
      "epoch": 0.43618499404525607,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004987157884495141,
      "loss": 5.618,
      "mean_token_accuracy": 0.14522770568728446,
      "num_tokens": 9652962.0,
      "step": 4395
    },
    {
      "entropy": 5.733834934234619,
      "epoch": 0.4366812227074236,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987120055091523,
      "loss": 5.6467,
      "mean_token_accuracy": 0.13846302777528763,
      "num_tokens": 9664084.0,
      "step": 4400
    },
    {
      "entropy": 5.686210298538208,
      "epoch": 0.43717745136959113,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004987082170212149,
      "loss": 5.5498,
      "mean_token_accuracy": 0.13982029631733894,
      "num_tokens": 9675821.0,
      "step": 4405
    },
    {
      "entropy": 5.777454996109009,
      "epoch": 0.43767368003175866,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987044229857957,
      "loss": 5.6793,
      "mean_token_accuracy": 0.13991234824061394,
      "num_tokens": 9687964.0,
      "step": 4410
    },
    {
      "entropy": 5.76395092010498,
      "epoch": 0.43816990869392614,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004987006234029889,
      "loss": 5.6423,
      "mean_token_accuracy": 0.14243705421686173,
      "num_tokens": 9699653.0,
      "step": 4415
    },
    {
      "entropy": 5.776408624649048,
      "epoch": 0.43866613735609367,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004986968182728885,
      "loss": 5.6439,
      "mean_token_accuracy": 0.14064784720540047,
      "num_tokens": 9709872.0,
      "step": 4420
    },
    {
      "entropy": 5.771645736694336,
      "epoch": 0.4391623660182612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986930075955892,
      "loss": 5.5691,
      "mean_token_accuracy": 0.13983753249049186,
      "num_tokens": 9720589.0,
      "step": 4425
    },
    {
      "entropy": 5.705972909927368,
      "epoch": 0.4396585946804287,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986891913711853,
      "loss": 5.5097,
      "mean_token_accuracy": 0.14566828012466432,
      "num_tokens": 9731862.0,
      "step": 4430
    },
    {
      "entropy": 5.70891170501709,
      "epoch": 0.44015482334259626,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986853695997715,
      "loss": 5.6521,
      "mean_token_accuracy": 0.1459285154938698,
      "num_tokens": 9743710.0,
      "step": 4435
    },
    {
      "entropy": 5.741820478439331,
      "epoch": 0.4406510520047638,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004986815422814425,
      "loss": 5.6018,
      "mean_token_accuracy": 0.13335008248686792,
      "num_tokens": 9753855.0,
      "step": 4440
    },
    {
      "entropy": 5.869410133361816,
      "epoch": 0.4411472806669313,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986777094162932,
      "loss": 5.6164,
      "mean_token_accuracy": 0.14792418032884597,
      "num_tokens": 9763883.0,
      "step": 4445
    },
    {
      "entropy": 5.728934240341187,
      "epoch": 0.44164350932909885,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986738710044187,
      "loss": 5.5744,
      "mean_token_accuracy": 0.1445605419576168,
      "num_tokens": 9774890.0,
      "step": 4450
    },
    {
      "entropy": 5.669674301147461,
      "epoch": 0.4421397379912664,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004986700270459142,
      "loss": 5.559,
      "mean_token_accuracy": 0.14981079697608948,
      "num_tokens": 9785249.0,
      "step": 4455
    },
    {
      "entropy": 5.749480390548706,
      "epoch": 0.4426359666534339,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004986661775408749,
      "loss": 5.5614,
      "mean_token_accuracy": 0.14750658124685287,
      "num_tokens": 9795723.0,
      "step": 4460
    },
    {
      "entropy": 5.719463205337524,
      "epoch": 0.44313219531560144,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004986623224893965,
      "loss": 5.5139,
      "mean_token_accuracy": 0.1400698557496071,
      "num_tokens": 9807637.0,
      "step": 4465
    },
    {
      "entropy": 5.7714917182922365,
      "epoch": 0.44362842397776897,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986584618915744,
      "loss": 5.6266,
      "mean_token_accuracy": 0.14391328766942024,
      "num_tokens": 9820298.0,
      "step": 4470
    },
    {
      "entropy": 5.688858079910278,
      "epoch": 0.4441246526399365,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986545957475045,
      "loss": 5.5109,
      "mean_token_accuracy": 0.14995549917221068,
      "num_tokens": 9831251.0,
      "step": 4475
    },
    {
      "entropy": 5.659614849090576,
      "epoch": 0.44462088130210403,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004986507240572825,
      "loss": 5.5749,
      "mean_token_accuracy": 0.14492021575570108,
      "num_tokens": 9842884.0,
      "step": 4480
    },
    {
      "entropy": 5.724112462997437,
      "epoch": 0.44511710996427156,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004986468468210042,
      "loss": 5.4791,
      "mean_token_accuracy": 0.14623685032129288,
      "num_tokens": 9854413.0,
      "step": 4485
    },
    {
      "entropy": 5.7491716861724855,
      "epoch": 0.44561333862643904,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004986429640387663,
      "loss": 5.6268,
      "mean_token_accuracy": 0.14398006424307824,
      "num_tokens": 9865771.0,
      "step": 4490
    },
    {
      "entropy": 5.749975156784058,
      "epoch": 0.44610956728860657,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004986390757106648,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1488683357834816,
      "num_tokens": 9876717.0,
      "step": 4495
    },
    {
      "entropy": 5.688867235183716,
      "epoch": 0.4466057959507741,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986351818367959,
      "loss": 5.6228,
      "mean_token_accuracy": 0.1436481349170208,
      "num_tokens": 9887616.0,
      "step": 4500
    },
    {
      "entropy": 5.749194478988647,
      "epoch": 0.44710202461294163,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986312824172566,
      "loss": 5.5664,
      "mean_token_accuracy": 0.14499585777521135,
      "num_tokens": 9898979.0,
      "step": 4505
    },
    {
      "entropy": 5.762835121154785,
      "epoch": 0.44759825327510916,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004986273774521433,
      "loss": 5.6138,
      "mean_token_accuracy": 0.13522646948695183,
      "num_tokens": 9912158.0,
      "step": 4510
    },
    {
      "entropy": 5.741280889511108,
      "epoch": 0.4480944819372767,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004986234669415529,
      "loss": 5.6189,
      "mean_token_accuracy": 0.13702357113361358,
      "num_tokens": 9923886.0,
      "step": 4515
    },
    {
      "entropy": 5.7603922367095945,
      "epoch": 0.4485907105994442,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004986195508855824,
      "loss": 5.5889,
      "mean_token_accuracy": 0.1485447883605957,
      "num_tokens": 9934979.0,
      "step": 4520
    },
    {
      "entropy": 5.760014390945434,
      "epoch": 0.44908693926161175,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498615629284329,
      "loss": 5.6866,
      "mean_token_accuracy": 0.13860371634364127,
      "num_tokens": 9944715.0,
      "step": 4525
    },
    {
      "entropy": 5.734985065460205,
      "epoch": 0.4495831679237793,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004986117021378896,
      "loss": 5.5541,
      "mean_token_accuracy": 0.1397101826965809,
      "num_tokens": 9954736.0,
      "step": 4530
    },
    {
      "entropy": 5.766929054260254,
      "epoch": 0.4500793965859468,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000498607769446362,
      "loss": 5.6342,
      "mean_token_accuracy": 0.14335151463747026,
      "num_tokens": 9965952.0,
      "step": 4535
    },
    {
      "entropy": 5.743745565414429,
      "epoch": 0.45057562524811434,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986038312098435,
      "loss": 5.5904,
      "mean_token_accuracy": 0.14454033672809602,
      "num_tokens": 9976428.0,
      "step": 4540
    },
    {
      "entropy": 5.73631854057312,
      "epoch": 0.4510718539102819,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985998874284319,
      "loss": 5.5196,
      "mean_token_accuracy": 0.15243304073810576,
      "num_tokens": 9986100.0,
      "step": 4545
    },
    {
      "entropy": 5.637545347213745,
      "epoch": 0.4515680825724494,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985959381022248,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1455839142203331,
      "num_tokens": 9996584.0,
      "step": 4550
    },
    {
      "entropy": 5.702834224700927,
      "epoch": 0.45206431123461693,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985919832313203,
      "loss": 5.5327,
      "mean_token_accuracy": 0.1477558955550194,
      "num_tokens": 10007792.0,
      "step": 4555
    },
    {
      "entropy": 5.85061206817627,
      "epoch": 0.45256053989678446,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985880228158164,
      "loss": 5.736,
      "mean_token_accuracy": 0.13626319989562036,
      "num_tokens": 10019615.0,
      "step": 4560
    },
    {
      "entropy": 5.681415987014771,
      "epoch": 0.45305676855895194,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985840568558112,
      "loss": 5.5471,
      "mean_token_accuracy": 0.1462045878171921,
      "num_tokens": 10030223.0,
      "step": 4565
    },
    {
      "entropy": 5.722938919067383,
      "epoch": 0.45355299722111947,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004985800853514033,
      "loss": 5.5653,
      "mean_token_accuracy": 0.14124826863408088,
      "num_tokens": 10041522.0,
      "step": 4570
    },
    {
      "entropy": 5.779463434219361,
      "epoch": 0.454049225883287,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985761083026911,
      "loss": 5.494,
      "mean_token_accuracy": 0.15177883207798004,
      "num_tokens": 10052847.0,
      "step": 4575
    },
    {
      "entropy": 5.673524332046509,
      "epoch": 0.45454545454545453,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985721257097731,
      "loss": 5.5208,
      "mean_token_accuracy": 0.14756261557340622,
      "num_tokens": 10062259.0,
      "step": 4580
    },
    {
      "entropy": 5.774816703796387,
      "epoch": 0.45504168320762206,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004985681375727482,
      "loss": 5.673,
      "mean_token_accuracy": 0.14000654220581055,
      "num_tokens": 10073341.0,
      "step": 4585
    },
    {
      "entropy": 5.774615240097046,
      "epoch": 0.4555379118697896,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004985641438917151,
      "loss": 5.5377,
      "mean_token_accuracy": 0.14914253503084182,
      "num_tokens": 10083992.0,
      "step": 4590
    },
    {
      "entropy": 5.756216049194336,
      "epoch": 0.4560341405319571,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985601446667731,
      "loss": 5.5195,
      "mean_token_accuracy": 0.14849689081311226,
      "num_tokens": 10094245.0,
      "step": 4595
    },
    {
      "entropy": 5.731303691864014,
      "epoch": 0.45653036919412465,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004985561398980213,
      "loss": 5.5228,
      "mean_token_accuracy": 0.14894152730703353,
      "num_tokens": 10105976.0,
      "step": 4600
    },
    {
      "entropy": 5.680754852294922,
      "epoch": 0.4570265978562922,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985521295855588,
      "loss": 5.598,
      "mean_token_accuracy": 0.1440993420779705,
      "num_tokens": 10117373.0,
      "step": 4605
    },
    {
      "entropy": 5.6981371402740475,
      "epoch": 0.4575228265184597,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004985481137294853,
      "loss": 5.5345,
      "mean_token_accuracy": 0.1376381404697895,
      "num_tokens": 10128470.0,
      "step": 4610
    },
    {
      "entropy": 5.678988218307495,
      "epoch": 0.45801905518062724,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004985440923299002,
      "loss": 5.5288,
      "mean_token_accuracy": 0.14953379333019257,
      "num_tokens": 10140217.0,
      "step": 4615
    },
    {
      "entropy": 5.707973003387451,
      "epoch": 0.4585152838427948,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985400653869034,
      "loss": 5.5948,
      "mean_token_accuracy": 0.1433454155921936,
      "num_tokens": 10152308.0,
      "step": 4620
    },
    {
      "entropy": 5.798585367202759,
      "epoch": 0.4590115125049623,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004985360329005945,
      "loss": 5.6299,
      "mean_token_accuracy": 0.1438091740012169,
      "num_tokens": 10164234.0,
      "step": 4625
    },
    {
      "entropy": 5.854622650146484,
      "epoch": 0.45950774116712984,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985319948710738,
      "loss": 5.6894,
      "mean_token_accuracy": 0.13872022926807404,
      "num_tokens": 10175287.0,
      "step": 4630
    },
    {
      "entropy": 5.773370361328125,
      "epoch": 0.46000396982929737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985279512984413,
      "loss": 5.5585,
      "mean_token_accuracy": 0.1476829767227173,
      "num_tokens": 10184658.0,
      "step": 4635
    },
    {
      "entropy": 5.721188688278199,
      "epoch": 0.46050019849146484,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985239021827972,
      "loss": 5.5495,
      "mean_token_accuracy": 0.14203211516141892,
      "num_tokens": 10194593.0,
      "step": 4640
    },
    {
      "entropy": 5.689546966552735,
      "epoch": 0.46099642715363237,
      "grad_norm": 0.8125,
      "learning_rate": 0.000498519847524242,
      "loss": 5.4787,
      "mean_token_accuracy": 0.14570856094360352,
      "num_tokens": 10205200.0,
      "step": 4645
    },
    {
      "entropy": 5.67290678024292,
      "epoch": 0.4614926558157999,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004985157873228762,
      "loss": 5.5025,
      "mean_token_accuracy": 0.14620592445135117,
      "num_tokens": 10216848.0,
      "step": 4650
    },
    {
      "entropy": 5.735358572006225,
      "epoch": 0.46198888447796743,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004985117215788005,
      "loss": 5.5506,
      "mean_token_accuracy": 0.14470544680953026,
      "num_tokens": 10228188.0,
      "step": 4655
    },
    {
      "entropy": 5.7273204803466795,
      "epoch": 0.46248511314013496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004985076502921158,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1441972553730011,
      "num_tokens": 10238253.0,
      "step": 4660
    },
    {
      "entropy": 5.752149772644043,
      "epoch": 0.4629813418023025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498503573462923,
      "loss": 5.5725,
      "mean_token_accuracy": 0.14394041150808334,
      "num_tokens": 10248169.0,
      "step": 4665
    },
    {
      "entropy": 5.658182430267334,
      "epoch": 0.46347757046447,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004984994910913232,
      "loss": 5.451,
      "mean_token_accuracy": 0.14721139520406723,
      "num_tokens": 10258962.0,
      "step": 4670
    },
    {
      "entropy": 5.762846517562866,
      "epoch": 0.46397379912663755,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004984954031774175,
      "loss": 5.5725,
      "mean_token_accuracy": 0.14608847945928574,
      "num_tokens": 10270303.0,
      "step": 4675
    },
    {
      "entropy": 5.633331775665283,
      "epoch": 0.4644700277888051,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004984913097213075,
      "loss": 5.4871,
      "mean_token_accuracy": 0.1489083856344223,
      "num_tokens": 10281782.0,
      "step": 4680
    },
    {
      "entropy": 5.703283452987671,
      "epoch": 0.4649662564509726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004984872107230945,
      "loss": 5.5247,
      "mean_token_accuracy": 0.14655748009681702,
      "num_tokens": 10292937.0,
      "step": 4685
    },
    {
      "entropy": 5.714847564697266,
      "epoch": 0.46546248511314015,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004984831061828803,
      "loss": 5.5931,
      "mean_token_accuracy": 0.13885734528303145,
      "num_tokens": 10305443.0,
      "step": 4690
    },
    {
      "entropy": 5.7107819557189945,
      "epoch": 0.4659587137753077,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984789961007667,
      "loss": 5.5171,
      "mean_token_accuracy": 0.14956869482994078,
      "num_tokens": 10316489.0,
      "step": 4695
    },
    {
      "entropy": 5.687655687332153,
      "epoch": 0.4664549424374752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984748804768554,
      "loss": 5.5752,
      "mean_token_accuracy": 0.14270322620868683,
      "num_tokens": 10326690.0,
      "step": 4700
    },
    {
      "entropy": 5.755337333679199,
      "epoch": 0.46695117109964274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984707593112487,
      "loss": 5.5544,
      "mean_token_accuracy": 0.1444431722164154,
      "num_tokens": 10337059.0,
      "step": 4705
    },
    {
      "entropy": 5.693855142593383,
      "epoch": 0.46744739976181027,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004984666326040487,
      "loss": 5.449,
      "mean_token_accuracy": 0.1503336876630783,
      "num_tokens": 10347035.0,
      "step": 4710
    },
    {
      "entropy": 5.661470556259156,
      "epoch": 0.46794362842397774,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004984625003553578,
      "loss": 5.573,
      "mean_token_accuracy": 0.14357958883047103,
      "num_tokens": 10359492.0,
      "step": 4715
    },
    {
      "entropy": 5.760824918746948,
      "epoch": 0.4684398570861453,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984583625652783,
      "loss": 5.545,
      "mean_token_accuracy": 0.1473846912384033,
      "num_tokens": 10370128.0,
      "step": 4720
    },
    {
      "entropy": 5.720354604721069,
      "epoch": 0.4689360857483128,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498454219233913,
      "loss": 5.5537,
      "mean_token_accuracy": 0.14348298013210298,
      "num_tokens": 10380276.0,
      "step": 4725
    },
    {
      "entropy": 5.634743022918701,
      "epoch": 0.46943231441048033,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984500703613645,
      "loss": 5.542,
      "mean_token_accuracy": 0.14880713149905206,
      "num_tokens": 10392489.0,
      "step": 4730
    },
    {
      "entropy": 5.787004089355468,
      "epoch": 0.46992854307264786,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004984459159477358,
      "loss": 5.6264,
      "mean_token_accuracy": 0.13868094533681868,
      "num_tokens": 10404259.0,
      "step": 4735
    },
    {
      "entropy": 5.6543690204620365,
      "epoch": 0.4704247717348154,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004984417559931298,
      "loss": 5.4786,
      "mean_token_accuracy": 0.14785647839307786,
      "num_tokens": 10415000.0,
      "step": 4740
    },
    {
      "entropy": 5.750206661224365,
      "epoch": 0.4709210003969829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984375904976498,
      "loss": 5.6124,
      "mean_token_accuracy": 0.14082250148057937,
      "num_tokens": 10426078.0,
      "step": 4745
    },
    {
      "entropy": 5.733554029464722,
      "epoch": 0.47141722905915046,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004984334194613989,
      "loss": 5.5392,
      "mean_token_accuracy": 0.14412411376833917,
      "num_tokens": 10436838.0,
      "step": 4750
    },
    {
      "entropy": 5.690441417694092,
      "epoch": 0.471913457721318,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984292428844809,
      "loss": 5.5247,
      "mean_token_accuracy": 0.1464602991938591,
      "num_tokens": 10448544.0,
      "step": 4755
    },
    {
      "entropy": 5.755918741226196,
      "epoch": 0.4724096863834855,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004984250607669989,
      "loss": 5.5996,
      "mean_token_accuracy": 0.13803669214248657,
      "num_tokens": 10459730.0,
      "step": 4760
    },
    {
      "entropy": 5.781705665588379,
      "epoch": 0.47290591504565305,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004984208731090569,
      "loss": 5.6513,
      "mean_token_accuracy": 0.1432105079293251,
      "num_tokens": 10470257.0,
      "step": 4765
    },
    {
      "entropy": 5.780958747863769,
      "epoch": 0.4734021437078206,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004984166799107587,
      "loss": 5.5937,
      "mean_token_accuracy": 0.14317760914564132,
      "num_tokens": 10480835.0,
      "step": 4770
    },
    {
      "entropy": 5.692502403259278,
      "epoch": 0.4738983723699881,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984124811722081,
      "loss": 5.6298,
      "mean_token_accuracy": 0.14655901938676835,
      "num_tokens": 10491720.0,
      "step": 4775
    },
    {
      "entropy": 5.710153818130493,
      "epoch": 0.47439460103215564,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004984082768935096,
      "loss": 5.6483,
      "mean_token_accuracy": 0.14328090101480484,
      "num_tokens": 10502313.0,
      "step": 4780
    },
    {
      "entropy": 5.764196062088013,
      "epoch": 0.47489082969432317,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984040670747672,
      "loss": 5.5147,
      "mean_token_accuracy": 0.1483422562479973,
      "num_tokens": 10513333.0,
      "step": 4785
    },
    {
      "entropy": 5.677767276763916,
      "epoch": 0.47538705835649064,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983998517160852,
      "loss": 5.4912,
      "mean_token_accuracy": 0.1518000140786171,
      "num_tokens": 10522766.0,
      "step": 4790
    },
    {
      "entropy": 5.683748245239258,
      "epoch": 0.4758832870186582,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983956308175684,
      "loss": 5.5425,
      "mean_token_accuracy": 0.14448487013578415,
      "num_tokens": 10534362.0,
      "step": 4795
    },
    {
      "entropy": 5.7460737228393555,
      "epoch": 0.4763795156808257,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004983914043793212,
      "loss": 5.5566,
      "mean_token_accuracy": 0.14349017888307572,
      "num_tokens": 10545072.0,
      "step": 4800
    },
    {
      "entropy": 5.669580745697021,
      "epoch": 0.47687574434299324,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983871724014487,
      "loss": 5.4952,
      "mean_token_accuracy": 0.14265795350074767,
      "num_tokens": 10555953.0,
      "step": 4805
    },
    {
      "entropy": 5.748552846908569,
      "epoch": 0.47737197300516077,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004983829348840556,
      "loss": 5.5455,
      "mean_token_accuracy": 0.14659083187580108,
      "num_tokens": 10567757.0,
      "step": 4810
    },
    {
      "entropy": 5.620566129684448,
      "epoch": 0.4778682016673283,
      "grad_norm": 0.875,
      "learning_rate": 0.000498378691827247,
      "loss": 5.4415,
      "mean_token_accuracy": 0.14495614916086197,
      "num_tokens": 10579390.0,
      "step": 4815
    },
    {
      "entropy": 5.726225280761719,
      "epoch": 0.4783644303294958,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983744432311282,
      "loss": 5.6348,
      "mean_token_accuracy": 0.14205840975046158,
      "num_tokens": 10589624.0,
      "step": 4820
    },
    {
      "entropy": 5.801154088973999,
      "epoch": 0.47886065899166336,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004983701890958046,
      "loss": 5.6162,
      "mean_token_accuracy": 0.13612727597355842,
      "num_tokens": 10601927.0,
      "step": 4825
    },
    {
      "entropy": 5.644892024993896,
      "epoch": 0.4793568876538309,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983659294213815,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1519262485206127,
      "num_tokens": 10614073.0,
      "step": 4830
    },
    {
      "entropy": 5.691881275177002,
      "epoch": 0.4798531163159984,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983616642079649,
      "loss": 5.5339,
      "mean_token_accuracy": 0.14844077527523042,
      "num_tokens": 10625535.0,
      "step": 4835
    },
    {
      "entropy": 5.790453100204468,
      "epoch": 0.48034934497816595,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983573934556601,
      "loss": 5.5024,
      "mean_token_accuracy": 0.14909157603979112,
      "num_tokens": 10637160.0,
      "step": 4840
    },
    {
      "entropy": 5.7284924507141115,
      "epoch": 0.4808455736403335,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004983531171645734,
      "loss": 5.6049,
      "mean_token_accuracy": 0.1403811626136303,
      "num_tokens": 10649061.0,
      "step": 4845
    },
    {
      "entropy": 5.594764184951782,
      "epoch": 0.481341802302501,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983488353348106,
      "loss": 5.4582,
      "mean_token_accuracy": 0.1496405154466629,
      "num_tokens": 10661130.0,
      "step": 4850
    },
    {
      "entropy": 5.695295190811157,
      "epoch": 0.48183803096466854,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004983445479664779,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1511105142533779,
      "num_tokens": 10673017.0,
      "step": 4855
    },
    {
      "entropy": 5.777050113677978,
      "epoch": 0.48233425962683607,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983402550596817,
      "loss": 5.5856,
      "mean_token_accuracy": 0.1438658468425274,
      "num_tokens": 10683268.0,
      "step": 4860
    },
    {
      "entropy": 5.7851903438568115,
      "epoch": 0.48283048828900355,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983359566145285,
      "loss": 5.6021,
      "mean_token_accuracy": 0.14283096343278884,
      "num_tokens": 10694220.0,
      "step": 4865
    },
    {
      "entropy": 5.711629629135132,
      "epoch": 0.4833267169511711,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004983316526311248,
      "loss": 5.5067,
      "mean_token_accuracy": 0.14021326303482057,
      "num_tokens": 10704829.0,
      "step": 4870
    },
    {
      "entropy": 5.698852491378784,
      "epoch": 0.4838229456133386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983273431095773,
      "loss": 5.6997,
      "mean_token_accuracy": 0.13279356062412262,
      "num_tokens": 10717160.0,
      "step": 4875
    },
    {
      "entropy": 5.75347752571106,
      "epoch": 0.48431917427550614,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983230280499929,
      "loss": 5.5064,
      "mean_token_accuracy": 0.1471844308078289,
      "num_tokens": 10728638.0,
      "step": 4880
    },
    {
      "entropy": 5.696688318252564,
      "epoch": 0.48481540293767367,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004983187074524787,
      "loss": 5.5652,
      "mean_token_accuracy": 0.1435106612741947,
      "num_tokens": 10740122.0,
      "step": 4885
    },
    {
      "entropy": 5.691134309768676,
      "epoch": 0.4853116315998412,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004983143813171418,
      "loss": 5.5214,
      "mean_token_accuracy": 0.15215135216712952,
      "num_tokens": 10751103.0,
      "step": 4890
    },
    {
      "entropy": 5.685958290100098,
      "epoch": 0.48580786026200873,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983100496440895,
      "loss": 5.4871,
      "mean_token_accuracy": 0.148745596408844,
      "num_tokens": 10761937.0,
      "step": 4895
    },
    {
      "entropy": 5.778647232055664,
      "epoch": 0.48630408892417626,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498305712433429,
      "loss": 5.7061,
      "mean_token_accuracy": 0.13792461901903152,
      "num_tokens": 10773527.0,
      "step": 4900
    },
    {
      "entropy": 5.718804550170899,
      "epoch": 0.4868003175863438,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000498301369685268,
      "loss": 5.5456,
      "mean_token_accuracy": 0.14536294639110564,
      "num_tokens": 10783873.0,
      "step": 4905
    },
    {
      "entropy": 5.64908618927002,
      "epoch": 0.4872965462485113,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004982970213997142,
      "loss": 5.3937,
      "mean_token_accuracy": 0.15310315489768983,
      "num_tokens": 10796124.0,
      "step": 4910
    },
    {
      "entropy": 5.615315866470337,
      "epoch": 0.48779277491067885,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004982926675768756,
      "loss": 5.5009,
      "mean_token_accuracy": 0.1461801566183567,
      "num_tokens": 10806468.0,
      "step": 4915
    },
    {
      "entropy": 5.687642908096313,
      "epoch": 0.4882890035728464,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004982883082168599,
      "loss": 5.5506,
      "mean_token_accuracy": 0.1436970427632332,
      "num_tokens": 10817126.0,
      "step": 4920
    },
    {
      "entropy": 5.748031282424927,
      "epoch": 0.4887852322350139,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982839433197752,
      "loss": 5.5764,
      "mean_token_accuracy": 0.13784480467438698,
      "num_tokens": 10828985.0,
      "step": 4925
    },
    {
      "entropy": 5.684624195098877,
      "epoch": 0.48928146089718144,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00049827957288573,
      "loss": 5.4938,
      "mean_token_accuracy": 0.1424009010195732,
      "num_tokens": 10840528.0,
      "step": 4930
    },
    {
      "entropy": 5.703600263595581,
      "epoch": 0.48977768955934897,
      "grad_norm": 0.875,
      "learning_rate": 0.0004982751969148324,
      "loss": 5.4988,
      "mean_token_accuracy": 0.14564865082502365,
      "num_tokens": 10851312.0,
      "step": 4935
    },
    {
      "entropy": 5.6754545211792,
      "epoch": 0.49027391822151645,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982708154071912,
      "loss": 5.4635,
      "mean_token_accuracy": 0.14783769845962524,
      "num_tokens": 10861882.0,
      "step": 4940
    },
    {
      "entropy": 5.6775837421417235,
      "epoch": 0.490770146883684,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982664283629147,
      "loss": 5.6191,
      "mean_token_accuracy": 0.13809697479009628,
      "num_tokens": 10873187.0,
      "step": 4945
    },
    {
      "entropy": 5.710150241851807,
      "epoch": 0.4912663755458515,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004982620357821118,
      "loss": 5.5006,
      "mean_token_accuracy": 0.1505657970905304,
      "num_tokens": 10884254.0,
      "step": 4950
    },
    {
      "entropy": 5.636944675445557,
      "epoch": 0.49176260420801904,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982576376648918,
      "loss": 5.4854,
      "mean_token_accuracy": 0.1480466514825821,
      "num_tokens": 10895770.0,
      "step": 4955
    },
    {
      "entropy": 5.6990667343139645,
      "epoch": 0.49225883287018657,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004982532340113633,
      "loss": 5.5043,
      "mean_token_accuracy": 0.14898262172937393,
      "num_tokens": 10906436.0,
      "step": 4960
    },
    {
      "entropy": 5.721513271331787,
      "epoch": 0.4927550615323541,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004982488248216357,
      "loss": 5.472,
      "mean_token_accuracy": 0.1475836418569088,
      "num_tokens": 10917655.0,
      "step": 4965
    },
    {
      "entropy": 5.664124584197998,
      "epoch": 0.49325129019452163,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004982444100958182,
      "loss": 5.4821,
      "mean_token_accuracy": 0.1533268451690674,
      "num_tokens": 10928359.0,
      "step": 4970
    },
    {
      "entropy": 5.724602460861206,
      "epoch": 0.49374751885668916,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004982399898340205,
      "loss": 5.6448,
      "mean_token_accuracy": 0.13842156752943993,
      "num_tokens": 10940494.0,
      "step": 4975
    },
    {
      "entropy": 5.756507015228271,
      "epoch": 0.4942437475188567,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000498235564036352,
      "loss": 5.6435,
      "mean_token_accuracy": 0.13922389969229698,
      "num_tokens": 10951298.0,
      "step": 4980
    },
    {
      "entropy": 5.708109712600708,
      "epoch": 0.4947399761810242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982311327029226,
      "loss": 5.5474,
      "mean_token_accuracy": 0.1470474436879158,
      "num_tokens": 10961680.0,
      "step": 4985
    },
    {
      "entropy": 5.7936280250549315,
      "epoch": 0.49523620484319175,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004982266958338421,
      "loss": 5.6652,
      "mean_token_accuracy": 0.14016711637377738,
      "num_tokens": 10974968.0,
      "step": 4990
    },
    {
      "entropy": 5.678466844558716,
      "epoch": 0.4957324335053593,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982222534292205,
      "loss": 5.583,
      "mean_token_accuracy": 0.13718260005116462,
      "num_tokens": 10986383.0,
      "step": 4995
    },
    {
      "entropy": 5.776921606063842,
      "epoch": 0.4962286621675268,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982178054891681,
      "loss": 5.6299,
      "mean_token_accuracy": 0.14285891950130464,
      "num_tokens": 10998384.0,
      "step": 5000
    },
    {
      "entropy": 5.70658597946167,
      "epoch": 0.49672489082969434,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982133520137951,
      "loss": 5.4022,
      "mean_token_accuracy": 0.153546941280365,
      "num_tokens": 11009420.0,
      "step": 5005
    },
    {
      "entropy": 5.716489982604981,
      "epoch": 0.4972211194918619,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004982088930032119,
      "loss": 5.5618,
      "mean_token_accuracy": 0.1433442324399948,
      "num_tokens": 11019601.0,
      "step": 5010
    },
    {
      "entropy": 5.6606697082519535,
      "epoch": 0.4977173481540294,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982044284575293,
      "loss": 5.4955,
      "mean_token_accuracy": 0.15104733407497406,
      "num_tokens": 11031138.0,
      "step": 5015
    },
    {
      "entropy": 5.685225868225098,
      "epoch": 0.4982135768161969,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004981999583768576,
      "loss": 5.5122,
      "mean_token_accuracy": 0.14943165183067322,
      "num_tokens": 11041643.0,
      "step": 5020
    },
    {
      "entropy": 5.689207077026367,
      "epoch": 0.4987098054783644,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498195482761308,
      "loss": 5.5301,
      "mean_token_accuracy": 0.1403835617005825,
      "num_tokens": 11052116.0,
      "step": 5025
    },
    {
      "entropy": 5.7984169006347654,
      "epoch": 0.49920603414053194,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004981910016109914,
      "loss": 5.717,
      "mean_token_accuracy": 0.13269980624318123,
      "num_tokens": 11064796.0,
      "step": 5030
    },
    {
      "entropy": 5.802556705474854,
      "epoch": 0.49970226280269947,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981865149260189,
      "loss": 5.5571,
      "mean_token_accuracy": 0.1468698725104332,
      "num_tokens": 11075029.0,
      "step": 5035
    },
    {
      "entropy": 5.733177566528321,
      "epoch": 0.5001984914648671,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981820227065016,
      "loss": 5.5683,
      "mean_token_accuracy": 0.14491811394691467,
      "num_tokens": 11086702.0,
      "step": 5040
    },
    {
      "entropy": 5.671647214889527,
      "epoch": 0.5006947201270345,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004981775249525513,
      "loss": 5.5705,
      "mean_token_accuracy": 0.14098249450325967,
      "num_tokens": 11097417.0,
      "step": 5045
    },
    {
      "entropy": 5.646207284927368,
      "epoch": 0.5011909487892021,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004981730216642791,
      "loss": 5.4522,
      "mean_token_accuracy": 0.15363001972436904,
      "num_tokens": 11108458.0,
      "step": 5050
    },
    {
      "entropy": 5.73930606842041,
      "epoch": 0.5016871774513696,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004981685128417969,
      "loss": 5.6282,
      "mean_token_accuracy": 0.1416224032640457,
      "num_tokens": 11121466.0,
      "step": 5055
    },
    {
      "entropy": 5.693162202835083,
      "epoch": 0.5021834061135371,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981639984852165,
      "loss": 5.5909,
      "mean_token_accuracy": 0.13639208003878595,
      "num_tokens": 11132489.0,
      "step": 5060
    },
    {
      "entropy": 5.683169937133789,
      "epoch": 0.5026796347757047,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981594785946498,
      "loss": 5.5223,
      "mean_token_accuracy": 0.14929886907339096,
      "num_tokens": 11143087.0,
      "step": 5065
    },
    {
      "entropy": 5.77903037071228,
      "epoch": 0.5031758634378721,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004981549531702089,
      "loss": 5.5633,
      "mean_token_accuracy": 0.1406758099794388,
      "num_tokens": 11154229.0,
      "step": 5070
    },
    {
      "entropy": 5.728841781616211,
      "epoch": 0.5036720921000397,
      "grad_norm": 0.81640625,
      "learning_rate": 0.000498150422212006,
      "loss": 5.5295,
      "mean_token_accuracy": 0.1439318038523197,
      "num_tokens": 11165698.0,
      "step": 5075
    },
    {
      "entropy": 5.718142414093018,
      "epoch": 0.5041683207622072,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004981458857201534,
      "loss": 5.5242,
      "mean_token_accuracy": 0.14167542159557342,
      "num_tokens": 11177025.0,
      "step": 5080
    },
    {
      "entropy": 5.772231388092041,
      "epoch": 0.5046645494243748,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981413436947639,
      "loss": 5.5914,
      "mean_token_accuracy": 0.14499209746718406,
      "num_tokens": 11188766.0,
      "step": 5085
    },
    {
      "entropy": 5.687726974487305,
      "epoch": 0.5051607780865423,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004981367961359496,
      "loss": 5.6424,
      "mean_token_accuracy": 0.134550778567791,
      "num_tokens": 11201626.0,
      "step": 5090
    },
    {
      "entropy": 5.6953493595123295,
      "epoch": 0.5056570067487098,
      "grad_norm": 0.875,
      "learning_rate": 0.0004981322430438239,
      "loss": 5.5333,
      "mean_token_accuracy": 0.13979216814041137,
      "num_tokens": 11213975.0,
      "step": 5095
    },
    {
      "entropy": 5.698822259902954,
      "epoch": 0.5061532354108773,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981276844184991,
      "loss": 5.4967,
      "mean_token_accuracy": 0.15024842023849488,
      "num_tokens": 11223636.0,
      "step": 5100
    },
    {
      "entropy": 5.673787069320679,
      "epoch": 0.5066494640730449,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004981231202600886,
      "loss": 5.4646,
      "mean_token_accuracy": 0.15102217048406602,
      "num_tokens": 11233549.0,
      "step": 5105
    },
    {
      "entropy": 5.5892860889434814,
      "epoch": 0.5071456927352124,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981185505687056,
      "loss": 5.4613,
      "mean_token_accuracy": 0.15288246273994446,
      "num_tokens": 11242915.0,
      "step": 5110
    },
    {
      "entropy": 5.766301584243775,
      "epoch": 0.50764192139738,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981139753444632,
      "loss": 5.5965,
      "mean_token_accuracy": 0.14191838204860688,
      "num_tokens": 11254370.0,
      "step": 5115
    },
    {
      "entropy": 5.747890424728394,
      "epoch": 0.5081381500595474,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981093945874752,
      "loss": 5.5289,
      "mean_token_accuracy": 0.1528825044631958,
      "num_tokens": 11264226.0,
      "step": 5120
    },
    {
      "entropy": 5.665480470657348,
      "epoch": 0.508634378721715,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981048082978547,
      "loss": 5.4259,
      "mean_token_accuracy": 0.1531237244606018,
      "num_tokens": 11275304.0,
      "step": 5125
    },
    {
      "entropy": 5.6558972835540775,
      "epoch": 0.5091306073838825,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004981002164757158,
      "loss": 5.4928,
      "mean_token_accuracy": 0.14390195682644844,
      "num_tokens": 11286179.0,
      "step": 5130
    },
    {
      "entropy": 5.741279888153076,
      "epoch": 0.50962683604605,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980956191211723,
      "loss": 5.5171,
      "mean_token_accuracy": 0.148561180382967,
      "num_tokens": 11297244.0,
      "step": 5135
    },
    {
      "entropy": 5.711801767349243,
      "epoch": 0.5101230647082176,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004980910162343382,
      "loss": 5.5288,
      "mean_token_accuracy": 0.14664965122938156,
      "num_tokens": 11307251.0,
      "step": 5140
    },
    {
      "entropy": 5.699382686614991,
      "epoch": 0.510619293370385,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980864078153275,
      "loss": 5.5037,
      "mean_token_accuracy": 0.14559799134731294,
      "num_tokens": 11318441.0,
      "step": 5145
    },
    {
      "entropy": 5.614227199554444,
      "epoch": 0.5111155220325526,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004980817938642548,
      "loss": 5.4707,
      "mean_token_accuracy": 0.14990834891796112,
      "num_tokens": 11328743.0,
      "step": 5150
    },
    {
      "entropy": 5.629807090759277,
      "epoch": 0.5116117506947201,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000498077174381234,
      "loss": 5.4739,
      "mean_token_accuracy": 0.14851214960217476,
      "num_tokens": 11340745.0,
      "step": 5155
    },
    {
      "entropy": 5.737399864196777,
      "epoch": 0.5121079793568877,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980725493663802,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1490337945520878,
      "num_tokens": 11352060.0,
      "step": 5160
    },
    {
      "entropy": 5.671787452697754,
      "epoch": 0.5126042080190552,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980679188198078,
      "loss": 5.5496,
      "mean_token_accuracy": 0.14169612154364586,
      "num_tokens": 11363097.0,
      "step": 5165
    },
    {
      "entropy": 5.7175041198730465,
      "epoch": 0.5131004366812227,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004980632827416317,
      "loss": 5.5562,
      "mean_token_accuracy": 0.15641497671604157,
      "num_tokens": 11373410.0,
      "step": 5170
    },
    {
      "entropy": 5.740688133239746,
      "epoch": 0.5135966653433902,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980586411319669,
      "loss": 5.6724,
      "mean_token_accuracy": 0.14097218215465546,
      "num_tokens": 11385615.0,
      "step": 5175
    },
    {
      "entropy": 5.627192544937134,
      "epoch": 0.5140928940055578,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004980539939909283,
      "loss": 5.4421,
      "mean_token_accuracy": 0.14328499883413315,
      "num_tokens": 11396492.0,
      "step": 5180
    },
    {
      "entropy": 5.721903657913208,
      "epoch": 0.5145891226677253,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980493413186314,
      "loss": 5.5636,
      "mean_token_accuracy": 0.13742735236883163,
      "num_tokens": 11406377.0,
      "step": 5185
    },
    {
      "entropy": 5.716456031799316,
      "epoch": 0.5150853513298929,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004980446831151914,
      "loss": 5.5313,
      "mean_token_accuracy": 0.14166169092059136,
      "num_tokens": 11416862.0,
      "step": 5190
    },
    {
      "entropy": 5.645942306518554,
      "epoch": 0.5155815799920603,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980400193807239,
      "loss": 5.473,
      "mean_token_accuracy": 0.15332718044519425,
      "num_tokens": 11428246.0,
      "step": 5195
    },
    {
      "entropy": 5.6899733543396,
      "epoch": 0.5160778086542279,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004980353501153446,
      "loss": 5.5189,
      "mean_token_accuracy": 0.14616449400782586,
      "num_tokens": 11438147.0,
      "step": 5200
    },
    {
      "entropy": 5.755989599227905,
      "epoch": 0.5165740373163954,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004980306753191691,
      "loss": 5.6215,
      "mean_token_accuracy": 0.13850996643304825,
      "num_tokens": 11449154.0,
      "step": 5205
    },
    {
      "entropy": 5.735669231414795,
      "epoch": 0.5170702659785629,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980259949923135,
      "loss": 5.546,
      "mean_token_accuracy": 0.14134582728147507,
      "num_tokens": 11459262.0,
      "step": 5210
    },
    {
      "entropy": 5.733550071716309,
      "epoch": 0.5175664946407305,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004980213091348938,
      "loss": 5.5271,
      "mean_token_accuracy": 0.14430459290742875,
      "num_tokens": 11471991.0,
      "step": 5215
    },
    {
      "entropy": 5.647618341445923,
      "epoch": 0.5180627233028979,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980166177470262,
      "loss": 5.4634,
      "mean_token_accuracy": 0.14960172548890113,
      "num_tokens": 11483176.0,
      "step": 5220
    },
    {
      "entropy": 5.7239342212677,
      "epoch": 0.5185589519650655,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980119208288271,
      "loss": 5.5633,
      "mean_token_accuracy": 0.14328117445111274,
      "num_tokens": 11494015.0,
      "step": 5225
    },
    {
      "entropy": 5.63838210105896,
      "epoch": 0.519055180627233,
      "grad_norm": 0.75,
      "learning_rate": 0.0004980072183804128,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1470133349299431,
      "num_tokens": 11506365.0,
      "step": 5230
    },
    {
      "entropy": 5.654445743560791,
      "epoch": 0.5195514092894006,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004980025104019,
      "loss": 5.5617,
      "mean_token_accuracy": 0.15119016021490098,
      "num_tokens": 11517199.0,
      "step": 5235
    },
    {
      "entropy": 5.672596406936646,
      "epoch": 0.520047637951568,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979977968934056,
      "loss": 5.4085,
      "mean_token_accuracy": 0.15655839592218398,
      "num_tokens": 11527591.0,
      "step": 5240
    },
    {
      "entropy": 5.704338216781617,
      "epoch": 0.5205438666137356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004979930778550463,
      "loss": 5.5863,
      "mean_token_accuracy": 0.14597027748823166,
      "num_tokens": 11538657.0,
      "step": 5245
    },
    {
      "entropy": 5.647687292098999,
      "epoch": 0.5210400952759031,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979883532869392,
      "loss": 5.4042,
      "mean_token_accuracy": 0.16328341513872147,
      "num_tokens": 11549513.0,
      "step": 5250
    },
    {
      "entropy": 5.702465009689331,
      "epoch": 0.5215363239380707,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004979836231892013,
      "loss": 5.5045,
      "mean_token_accuracy": 0.1428775258362293,
      "num_tokens": 11560376.0,
      "step": 5255
    },
    {
      "entropy": 5.598912048339844,
      "epoch": 0.5220325526002382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004979788875619501,
      "loss": 5.4769,
      "mean_token_accuracy": 0.144529527425766,
      "num_tokens": 11571218.0,
      "step": 5260
    },
    {
      "entropy": 5.695925712585449,
      "epoch": 0.5225287812624058,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000497974146405303,
      "loss": 5.5399,
      "mean_token_accuracy": 0.14513622969388962,
      "num_tokens": 11581981.0,
      "step": 5265
    },
    {
      "entropy": 5.692674779891968,
      "epoch": 0.5230250099245732,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004979693997193776,
      "loss": 5.4934,
      "mean_token_accuracy": 0.1450166270136833,
      "num_tokens": 11592561.0,
      "step": 5270
    },
    {
      "entropy": 5.696996021270752,
      "epoch": 0.5235212385867408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979646475042914,
      "loss": 5.4746,
      "mean_token_accuracy": 0.1476963773369789,
      "num_tokens": 11601933.0,
      "step": 5275
    },
    {
      "entropy": 5.623072099685669,
      "epoch": 0.5240174672489083,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979598897601625,
      "loss": 5.3984,
      "mean_token_accuracy": 0.15914093554019929,
      "num_tokens": 11613938.0,
      "step": 5280
    },
    {
      "entropy": 5.741873025894165,
      "epoch": 0.5245136959110758,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979551264871088,
      "loss": 5.5782,
      "mean_token_accuracy": 0.14074024707078933,
      "num_tokens": 11624890.0,
      "step": 5285
    },
    {
      "entropy": 5.699547481536865,
      "epoch": 0.5250099245732434,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979503576852482,
      "loss": 5.5742,
      "mean_token_accuracy": 0.1491569310426712,
      "num_tokens": 11635080.0,
      "step": 5290
    },
    {
      "entropy": 5.596605730056763,
      "epoch": 0.5255061532354108,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004979455833546994,
      "loss": 5.47,
      "mean_token_accuracy": 0.14806455075740815,
      "num_tokens": 11645180.0,
      "step": 5295
    },
    {
      "entropy": 5.72585654258728,
      "epoch": 0.5260023818975784,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004979408034955804,
      "loss": 5.5278,
      "mean_token_accuracy": 0.14772395491600038,
      "num_tokens": 11657599.0,
      "step": 5300
    },
    {
      "entropy": 5.728936290740966,
      "epoch": 0.5264986105597459,
      "grad_norm": 0.875,
      "learning_rate": 0.0004979360181080099,
      "loss": 5.5593,
      "mean_token_accuracy": 0.13599722683429719,
      "num_tokens": 11669177.0,
      "step": 5305
    },
    {
      "entropy": 5.638133859634399,
      "epoch": 0.5269948392219135,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004979312271921065,
      "loss": 5.4153,
      "mean_token_accuracy": 0.1493235096335411,
      "num_tokens": 11679756.0,
      "step": 5310
    },
    {
      "entropy": 5.640149021148682,
      "epoch": 0.527491067884081,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497926430747989,
      "loss": 5.4883,
      "mean_token_accuracy": 0.14818147718906402,
      "num_tokens": 11689805.0,
      "step": 5315
    },
    {
      "entropy": 5.669553899765015,
      "epoch": 0.5279872965462485,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004979216287757764,
      "loss": 5.5018,
      "mean_token_accuracy": 0.1502943143248558,
      "num_tokens": 11701150.0,
      "step": 5320
    },
    {
      "entropy": 5.623200416564941,
      "epoch": 0.528483525208416,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497916821275588,
      "loss": 5.4294,
      "mean_token_accuracy": 0.15086933523416518,
      "num_tokens": 11712256.0,
      "step": 5325
    },
    {
      "entropy": 5.702307796478271,
      "epoch": 0.5289797538705836,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004979120082475424,
      "loss": 5.4842,
      "mean_token_accuracy": 0.15134317278862,
      "num_tokens": 11722547.0,
      "step": 5330
    },
    {
      "entropy": 5.702078723907471,
      "epoch": 0.5294759825327511,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004979071896917595,
      "loss": 5.4992,
      "mean_token_accuracy": 0.15180247128009797,
      "num_tokens": 11732309.0,
      "step": 5335
    },
    {
      "entropy": 5.69677963256836,
      "epoch": 0.5299722111949187,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004979023656083586,
      "loss": 5.5597,
      "mean_token_accuracy": 0.14569039940834044,
      "num_tokens": 11743982.0,
      "step": 5340
    },
    {
      "entropy": 5.63079833984375,
      "epoch": 0.5304684398570861,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004978975359974594,
      "loss": 5.4775,
      "mean_token_accuracy": 0.15016728341579438,
      "num_tokens": 11755506.0,
      "step": 5345
    },
    {
      "entropy": 5.711451768875122,
      "epoch": 0.5309646685192537,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978927008591816,
      "loss": 5.5136,
      "mean_token_accuracy": 0.14680191427469252,
      "num_tokens": 11767024.0,
      "step": 5350
    },
    {
      "entropy": 5.7230995178222654,
      "epoch": 0.5314608971814212,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978878601936451,
      "loss": 5.5021,
      "mean_token_accuracy": 0.14281241297721864,
      "num_tokens": 11777625.0,
      "step": 5355
    },
    {
      "entropy": 5.599544477462769,
      "epoch": 0.5319571258435887,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978830140009698,
      "loss": 5.4763,
      "mean_token_accuracy": 0.15325788259506226,
      "num_tokens": 11788610.0,
      "step": 5360
    },
    {
      "entropy": 5.683329153060913,
      "epoch": 0.5324533545057563,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004978781622812762,
      "loss": 5.5901,
      "mean_token_accuracy": 0.14808462858200072,
      "num_tokens": 11800828.0,
      "step": 5365
    },
    {
      "entropy": 5.7085408687591555,
      "epoch": 0.5329495831679237,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004978733050346844,
      "loss": 5.5759,
      "mean_token_accuracy": 0.1497088834643364,
      "num_tokens": 11811729.0,
      "step": 5370
    },
    {
      "entropy": 5.729819393157959,
      "epoch": 0.5334458118300913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978684422613149,
      "loss": 5.6052,
      "mean_token_accuracy": 0.1488908611238003,
      "num_tokens": 11822081.0,
      "step": 5375
    },
    {
      "entropy": 5.682017040252686,
      "epoch": 0.5339420404922588,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004978635739612882,
      "loss": 5.4742,
      "mean_token_accuracy": 0.14923994690179826,
      "num_tokens": 11832984.0,
      "step": 5380
    },
    {
      "entropy": 5.643860673904419,
      "epoch": 0.5344382691544264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004978587001347252,
      "loss": 5.4421,
      "mean_token_accuracy": 0.1456388145685196,
      "num_tokens": 11842918.0,
      "step": 5385
    },
    {
      "entropy": 5.611401128768921,
      "epoch": 0.5349344978165939,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004978538207817465,
      "loss": 5.4753,
      "mean_token_accuracy": 0.15014837682247162,
      "num_tokens": 11854364.0,
      "step": 5390
    },
    {
      "entropy": 5.665836906433105,
      "epoch": 0.5354307264787614,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978489359024734,
      "loss": 5.5602,
      "mean_token_accuracy": 0.1467744901776314,
      "num_tokens": 11866400.0,
      "step": 5395
    },
    {
      "entropy": 5.746421003341675,
      "epoch": 0.5359269551409289,
      "grad_norm": 0.875,
      "learning_rate": 0.0004978440454970268,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1467426374554634,
      "num_tokens": 11877094.0,
      "step": 5400
    },
    {
      "entropy": 5.692235851287842,
      "epoch": 0.5364231838030965,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978391495655281,
      "loss": 5.5029,
      "mean_token_accuracy": 0.14505623951554297,
      "num_tokens": 11887289.0,
      "step": 5405
    },
    {
      "entropy": 5.696217060089111,
      "epoch": 0.536919412465264,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978342481080987,
      "loss": 5.4836,
      "mean_token_accuracy": 0.1445485085248947,
      "num_tokens": 11897700.0,
      "step": 5410
    },
    {
      "entropy": 5.578903913497925,
      "epoch": 0.5374156411274316,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978293411248601,
      "loss": 5.402,
      "mean_token_accuracy": 0.15335436314344406,
      "num_tokens": 11908623.0,
      "step": 5415
    },
    {
      "entropy": 5.6434561252594,
      "epoch": 0.537911869789599,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978244286159339,
      "loss": 5.4282,
      "mean_token_accuracy": 0.1539301946759224,
      "num_tokens": 11917883.0,
      "step": 5420
    },
    {
      "entropy": 5.6117928981781,
      "epoch": 0.5384080984517666,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0004978195105814421,
      "loss": 5.4679,
      "mean_token_accuracy": 0.15135658234357835,
      "num_tokens": 11929145.0,
      "step": 5425
    },
    {
      "entropy": 5.65803861618042,
      "epoch": 0.5389043271139341,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004978145870215065,
      "loss": 5.4603,
      "mean_token_accuracy": 0.14786406978964806,
      "num_tokens": 11940073.0,
      "step": 5430
    },
    {
      "entropy": 5.69571967124939,
      "epoch": 0.5394005557761016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004978096579362493,
      "loss": 5.584,
      "mean_token_accuracy": 0.14479052647948265,
      "num_tokens": 11953140.0,
      "step": 5435
    },
    {
      "entropy": 5.694457769393921,
      "epoch": 0.5398967844382692,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978047233257927,
      "loss": 5.5123,
      "mean_token_accuracy": 0.15029323399066924,
      "num_tokens": 11963853.0,
      "step": 5440
    },
    {
      "entropy": 5.65546989440918,
      "epoch": 0.5403930131004366,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004977997831902591,
      "loss": 5.4588,
      "mean_token_accuracy": 0.15378749817609788,
      "num_tokens": 11975469.0,
      "step": 5445
    },
    {
      "entropy": 5.66071457862854,
      "epoch": 0.5408892417626042,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977948375297709,
      "loss": 5.4669,
      "mean_token_accuracy": 0.15611819177865982,
      "num_tokens": 11985227.0,
      "step": 5450
    },
    {
      "entropy": 5.542568397521973,
      "epoch": 0.5413854704247717,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977898863444509,
      "loss": 5.4603,
      "mean_token_accuracy": 0.15204665064811707,
      "num_tokens": 11995459.0,
      "step": 5455
    },
    {
      "entropy": 5.653804159164428,
      "epoch": 0.5418816990869393,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977849296344217,
      "loss": 5.5202,
      "mean_token_accuracy": 0.14583294242620468,
      "num_tokens": 12007705.0,
      "step": 5460
    },
    {
      "entropy": 5.644483661651611,
      "epoch": 0.5423779277491068,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004977799673998063,
      "loss": 5.4183,
      "mean_token_accuracy": 0.1514454334974289,
      "num_tokens": 12019729.0,
      "step": 5465
    },
    {
      "entropy": 5.667261981964112,
      "epoch": 0.5428741564112743,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004977749996407278,
      "loss": 5.4897,
      "mean_token_accuracy": 0.14822353720664977,
      "num_tokens": 12031009.0,
      "step": 5470
    },
    {
      "entropy": 5.725725555419922,
      "epoch": 0.5433703850734418,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977700263573094,
      "loss": 5.4485,
      "mean_token_accuracy": 0.1510414496064186,
      "num_tokens": 12041396.0,
      "step": 5475
    },
    {
      "entropy": 5.632419443130493,
      "epoch": 0.5438666137356094,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004977650475496743,
      "loss": 5.6064,
      "mean_token_accuracy": 0.14299786016345023,
      "num_tokens": 12053112.0,
      "step": 5480
    },
    {
      "entropy": 5.669924259185791,
      "epoch": 0.5443628423977769,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004977600632179459,
      "loss": 5.4854,
      "mean_token_accuracy": 0.1433200530707836,
      "num_tokens": 12064823.0,
      "step": 5485
    },
    {
      "entropy": 5.714823436737061,
      "epoch": 0.5448590710599445,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977550733622482,
      "loss": 5.6022,
      "mean_token_accuracy": 0.14357881546020507,
      "num_tokens": 12076656.0,
      "step": 5490
    },
    {
      "entropy": 5.673926258087159,
      "epoch": 0.5453552997221119,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004977500779827045,
      "loss": 5.3926,
      "mean_token_accuracy": 0.14801742434501647,
      "num_tokens": 12086324.0,
      "step": 5495
    },
    {
      "entropy": 5.579193353652954,
      "epoch": 0.5458515283842795,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004977450770794389,
      "loss": 5.43,
      "mean_token_accuracy": 0.15020886734127997,
      "num_tokens": 12098076.0,
      "step": 5500
    },
    {
      "entropy": 5.623298311233521,
      "epoch": 0.546347757046447,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004977400706525755,
      "loss": 5.4999,
      "mean_token_accuracy": 0.14142108112573623,
      "num_tokens": 12111282.0,
      "step": 5505
    },
    {
      "entropy": 5.665348815917969,
      "epoch": 0.5468439857086145,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004977350587022381,
      "loss": 5.4268,
      "mean_token_accuracy": 0.15393007397651673,
      "num_tokens": 12121932.0,
      "step": 5510
    },
    {
      "entropy": 5.642740154266358,
      "epoch": 0.5473402143707821,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977300412285513,
      "loss": 5.4335,
      "mean_token_accuracy": 0.15426870584487914,
      "num_tokens": 12133040.0,
      "step": 5515
    },
    {
      "entropy": 5.646396446228027,
      "epoch": 0.5478364430329495,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977250182316394,
      "loss": 5.5045,
      "mean_token_accuracy": 0.1467506393790245,
      "num_tokens": 12142577.0,
      "step": 5520
    },
    {
      "entropy": 5.665579605102539,
      "epoch": 0.5483326716951171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000497719989711627,
      "loss": 5.4692,
      "mean_token_accuracy": 0.15208701193332672,
      "num_tokens": 12153941.0,
      "step": 5525
    },
    {
      "entropy": 5.641311073303223,
      "epoch": 0.5488289003572846,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004977149556686387,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1455488011240959,
      "num_tokens": 12165347.0,
      "step": 5530
    },
    {
      "entropy": 5.528408813476562,
      "epoch": 0.5493251290194522,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977099161027996,
      "loss": 5.4083,
      "mean_token_accuracy": 0.14915482550859452,
      "num_tokens": 12175705.0,
      "step": 5535
    },
    {
      "entropy": 5.67162766456604,
      "epoch": 0.5498213576816197,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977048710142342,
      "loss": 5.4617,
      "mean_token_accuracy": 0.14935485273599625,
      "num_tokens": 12186781.0,
      "step": 5540
    },
    {
      "entropy": 5.696006155014038,
      "epoch": 0.5503175863437872,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976998204030681,
      "loss": 5.4922,
      "mean_token_accuracy": 0.1457320123910904,
      "num_tokens": 12197145.0,
      "step": 5545
    },
    {
      "entropy": 5.612071228027344,
      "epoch": 0.5508138150059547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976947642694262,
      "loss": 5.4933,
      "mean_token_accuracy": 0.1500881627202034,
      "num_tokens": 12207637.0,
      "step": 5550
    },
    {
      "entropy": 5.618806934356689,
      "epoch": 0.5513100436681223,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004976897026134341,
      "loss": 5.4132,
      "mean_token_accuracy": 0.1535649225115776,
      "num_tokens": 12218156.0,
      "step": 5555
    },
    {
      "entropy": 5.657208681106567,
      "epoch": 0.5518062723302898,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004976846354352172,
      "loss": 5.5089,
      "mean_token_accuracy": 0.14579661935567856,
      "num_tokens": 12229987.0,
      "step": 5560
    },
    {
      "entropy": 5.652930545806885,
      "epoch": 0.5523025009924574,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976795627349012,
      "loss": 5.4851,
      "mean_token_accuracy": 0.14279719218611717,
      "num_tokens": 12240508.0,
      "step": 5565
    },
    {
      "entropy": 5.621035194396972,
      "epoch": 0.5527987296546248,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976744845126118,
      "loss": 5.4208,
      "mean_token_accuracy": 0.15830605030059813,
      "num_tokens": 12250317.0,
      "step": 5570
    },
    {
      "entropy": 5.580662393569947,
      "epoch": 0.5532949583167924,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004976694007684751,
      "loss": 5.4636,
      "mean_token_accuracy": 0.15025635659694672,
      "num_tokens": 12262104.0,
      "step": 5575
    },
    {
      "entropy": 5.737037944793701,
      "epoch": 0.5537911869789599,
      "grad_norm": 0.875,
      "learning_rate": 0.000497664311502617,
      "loss": 5.4966,
      "mean_token_accuracy": 0.1508025825023651,
      "num_tokens": 12273259.0,
      "step": 5580
    },
    {
      "entropy": 5.720252656936646,
      "epoch": 0.5542874156411274,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004976592167151638,
      "loss": 5.5391,
      "mean_token_accuracy": 0.1400592029094696,
      "num_tokens": 12285504.0,
      "step": 5585
    },
    {
      "entropy": 5.683357810974121,
      "epoch": 0.554783644303295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004976541164062418,
      "loss": 5.4685,
      "mean_token_accuracy": 0.14882799685001374,
      "num_tokens": 12296512.0,
      "step": 5590
    },
    {
      "entropy": 5.541460037231445,
      "epoch": 0.5552798729654624,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004976490105759774,
      "loss": 5.3466,
      "mean_token_accuracy": 0.15359402894973756,
      "num_tokens": 12307948.0,
      "step": 5595
    },
    {
      "entropy": 5.623454666137695,
      "epoch": 0.55577610162763,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976438992244976,
      "loss": 5.5132,
      "mean_token_accuracy": 0.14310732185840608,
      "num_tokens": 12320020.0,
      "step": 5600
    },
    {
      "entropy": 5.672675514221192,
      "epoch": 0.5562723302897975,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004976387823519286,
      "loss": 5.3731,
      "mean_token_accuracy": 0.15214211344718934,
      "num_tokens": 12331091.0,
      "step": 5605
    },
    {
      "entropy": 5.638494157791138,
      "epoch": 0.5567685589519651,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976336599583975,
      "loss": 5.4929,
      "mean_token_accuracy": 0.14731888025999068,
      "num_tokens": 12342621.0,
      "step": 5610
    },
    {
      "entropy": 5.752030944824218,
      "epoch": 0.5572647876141326,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004976285320440315,
      "loss": 5.7177,
      "mean_token_accuracy": 0.13635711893439292,
      "num_tokens": 12353330.0,
      "step": 5615
    },
    {
      "entropy": 5.713182973861694,
      "epoch": 0.5577610162763001,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004976233986089577,
      "loss": 5.4805,
      "mean_token_accuracy": 0.1484640821814537,
      "num_tokens": 12364418.0,
      "step": 5620
    },
    {
      "entropy": 5.621135520935058,
      "epoch": 0.5582572449384676,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004976182596533031,
      "loss": 5.467,
      "mean_token_accuracy": 0.1504691794514656,
      "num_tokens": 12377022.0,
      "step": 5625
    },
    {
      "entropy": 5.677449464797974,
      "epoch": 0.5587534736006352,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976131151771955,
      "loss": 5.5259,
      "mean_token_accuracy": 0.1481981545686722,
      "num_tokens": 12387799.0,
      "step": 5630
    },
    {
      "entropy": 5.613097858428955,
      "epoch": 0.5592497022628027,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004976079651807623,
      "loss": 5.4567,
      "mean_token_accuracy": 0.14033942222595214,
      "num_tokens": 12397387.0,
      "step": 5635
    },
    {
      "entropy": 5.583207321166992,
      "epoch": 0.5597459309249703,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004976028096641313,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1553087905049324,
      "num_tokens": 12407994.0,
      "step": 5640
    },
    {
      "entropy": 5.7475121974945065,
      "epoch": 0.5602421595871377,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004975976486274302,
      "loss": 5.5885,
      "mean_token_accuracy": 0.14509609416127206,
      "num_tokens": 12419124.0,
      "step": 5645
    },
    {
      "entropy": 5.654124402999878,
      "epoch": 0.5607383882493053,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975924820707872,
      "loss": 5.4571,
      "mean_token_accuracy": 0.15169062614440917,
      "num_tokens": 12428699.0,
      "step": 5650
    },
    {
      "entropy": 5.672088861465454,
      "epoch": 0.5612346169114728,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004975873099943302,
      "loss": 5.538,
      "mean_token_accuracy": 0.14438563510775565,
      "num_tokens": 12439025.0,
      "step": 5655
    },
    {
      "entropy": 5.711086797714233,
      "epoch": 0.5617308455736403,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975821323981875,
      "loss": 5.489,
      "mean_token_accuracy": 0.1508912116289139,
      "num_tokens": 12449439.0,
      "step": 5660
    },
    {
      "entropy": 5.681025075912475,
      "epoch": 0.5622270742358079,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004975769492824877,
      "loss": 5.4494,
      "mean_token_accuracy": 0.149598528444767,
      "num_tokens": 12460066.0,
      "step": 5665
    },
    {
      "entropy": 5.626625871658325,
      "epoch": 0.5627233028979753,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975717606473589,
      "loss": 5.4573,
      "mean_token_accuracy": 0.15219570994377135,
      "num_tokens": 12471093.0,
      "step": 5670
    },
    {
      "entropy": 5.690846967697143,
      "epoch": 0.5632195315601429,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975665664929302,
      "loss": 5.4767,
      "mean_token_accuracy": 0.14674275368452072,
      "num_tokens": 12481181.0,
      "step": 5675
    },
    {
      "entropy": 5.674772453308106,
      "epoch": 0.5637157602223104,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975613668193303,
      "loss": 5.5165,
      "mean_token_accuracy": 0.14470747113227844,
      "num_tokens": 12493068.0,
      "step": 5680
    },
    {
      "entropy": 5.626017284393311,
      "epoch": 0.564211988884478,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497556161626688,
      "loss": 5.4572,
      "mean_token_accuracy": 0.15124062970280647,
      "num_tokens": 12504025.0,
      "step": 5685
    },
    {
      "entropy": 5.655818891525269,
      "epoch": 0.5647082175466455,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975509509151325,
      "loss": 5.47,
      "mean_token_accuracy": 0.1495766207575798,
      "num_tokens": 12514567.0,
      "step": 5690
    },
    {
      "entropy": 5.745582246780396,
      "epoch": 0.565204446208813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975457346847929,
      "loss": 5.5109,
      "mean_token_accuracy": 0.1437394253909588,
      "num_tokens": 12525088.0,
      "step": 5695
    },
    {
      "entropy": 5.619305181503296,
      "epoch": 0.5657006748709805,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004975405129357986,
      "loss": 5.4592,
      "mean_token_accuracy": 0.14793002754449844,
      "num_tokens": 12536002.0,
      "step": 5700
    },
    {
      "entropy": 5.66227536201477,
      "epoch": 0.5661969035331481,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004975352856682792,
      "loss": 5.4825,
      "mean_token_accuracy": 0.1485205963253975,
      "num_tokens": 12547468.0,
      "step": 5705
    },
    {
      "entropy": 5.696508073806763,
      "epoch": 0.5666931321953156,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975300528823642,
      "loss": 5.4495,
      "mean_token_accuracy": 0.14709952771663665,
      "num_tokens": 12557670.0,
      "step": 5710
    },
    {
      "entropy": 5.630479097366333,
      "epoch": 0.5671893608574832,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004975248145781834,
      "loss": 5.4912,
      "mean_token_accuracy": 0.15083353519439696,
      "num_tokens": 12568425.0,
      "step": 5715
    },
    {
      "entropy": 5.593966245651245,
      "epoch": 0.5676855895196506,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004975195707558667,
      "loss": 5.4597,
      "mean_token_accuracy": 0.14737943261861802,
      "num_tokens": 12579138.0,
      "step": 5720
    },
    {
      "entropy": 5.632019186019898,
      "epoch": 0.5681818181818182,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975143214155443,
      "loss": 5.4871,
      "mean_token_accuracy": 0.14770486801862717,
      "num_tokens": 12589743.0,
      "step": 5725
    },
    {
      "entropy": 5.631976079940796,
      "epoch": 0.5686780468439857,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975090665573459,
      "loss": 5.4636,
      "mean_token_accuracy": 0.15182324796915053,
      "num_tokens": 12600648.0,
      "step": 5730
    },
    {
      "entropy": 5.6232400894165036,
      "epoch": 0.5691742755061532,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004975038061814023,
      "loss": 5.4726,
      "mean_token_accuracy": 0.15145858526229858,
      "num_tokens": 12611369.0,
      "step": 5735
    },
    {
      "entropy": 5.679605722427368,
      "epoch": 0.5696705041683208,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004974985402878437,
      "loss": 5.4434,
      "mean_token_accuracy": 0.14840132668614386,
      "num_tokens": 12621992.0,
      "step": 5740
    },
    {
      "entropy": 5.661918497085571,
      "epoch": 0.5701667328304882,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004974932688768007,
      "loss": 5.5247,
      "mean_token_accuracy": 0.14979631304740906,
      "num_tokens": 12632463.0,
      "step": 5745
    },
    {
      "entropy": 5.5828722476959225,
      "epoch": 0.5706629614926558,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004974879919484042,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1506527543067932,
      "num_tokens": 12642282.0,
      "step": 5750
    },
    {
      "entropy": 5.623836517333984,
      "epoch": 0.5711591901548233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004974827095027848,
      "loss": 5.4178,
      "mean_token_accuracy": 0.1547690063714981,
      "num_tokens": 12653495.0,
      "step": 5755
    },
    {
      "entropy": 5.632310438156128,
      "epoch": 0.5716554188169909,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004974774215400736,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1610822707414627,
      "num_tokens": 12664322.0,
      "step": 5760
    },
    {
      "entropy": 5.529138660430908,
      "epoch": 0.5721516474791584,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974721280604019,
      "loss": 5.4479,
      "mean_token_accuracy": 0.14687813818454742,
      "num_tokens": 12673981.0,
      "step": 5765
    },
    {
      "entropy": 5.586023998260498,
      "epoch": 0.572647876141326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974668290639007,
      "loss": 5.453,
      "mean_token_accuracy": 0.14951502829790114,
      "num_tokens": 12684778.0,
      "step": 5770
    },
    {
      "entropy": 5.684868001937867,
      "epoch": 0.5731441048034934,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004974615245507014,
      "loss": 5.4192,
      "mean_token_accuracy": 0.15309299677610397,
      "num_tokens": 12695287.0,
      "step": 5775
    },
    {
      "entropy": 5.587672472000122,
      "epoch": 0.573640333465661,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004974562145209359,
      "loss": 5.4742,
      "mean_token_accuracy": 0.1430980756878853,
      "num_tokens": 12705773.0,
      "step": 5780
    },
    {
      "entropy": 5.590578365325928,
      "epoch": 0.5741365621278285,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974508989747354,
      "loss": 5.4093,
      "mean_token_accuracy": 0.15542308688163758,
      "num_tokens": 12716783.0,
      "step": 5785
    },
    {
      "entropy": 5.659099435806274,
      "epoch": 0.5746327907899961,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974455779122321,
      "loss": 5.4672,
      "mean_token_accuracy": 0.14360931888222694,
      "num_tokens": 12728146.0,
      "step": 5790
    },
    {
      "entropy": 5.695368003845215,
      "epoch": 0.5751290194521635,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004974402513335577,
      "loss": 5.5032,
      "mean_token_accuracy": 0.1484452560544014,
      "num_tokens": 12739683.0,
      "step": 5795
    },
    {
      "entropy": 5.710870695114136,
      "epoch": 0.5756252481143311,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974349192388444,
      "loss": 5.462,
      "mean_token_accuracy": 0.1511399433016777,
      "num_tokens": 12750257.0,
      "step": 5800
    },
    {
      "entropy": 5.714806365966797,
      "epoch": 0.5761214767764986,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004974295816282245,
      "loss": 5.5488,
      "mean_token_accuracy": 0.14585890769958496,
      "num_tokens": 12761055.0,
      "step": 5805
    },
    {
      "entropy": 5.653726387023926,
      "epoch": 0.5766177054386661,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974242385018303,
      "loss": 5.4809,
      "mean_token_accuracy": 0.1489928610622883,
      "num_tokens": 12771014.0,
      "step": 5810
    },
    {
      "entropy": 5.612311744689942,
      "epoch": 0.5771139341008337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497418889859794,
      "loss": 5.3926,
      "mean_token_accuracy": 0.15432902723550795,
      "num_tokens": 12782623.0,
      "step": 5815
    },
    {
      "entropy": 5.671354532241821,
      "epoch": 0.5776101627630011,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004974135357022488,
      "loss": 5.5132,
      "mean_token_accuracy": 0.1461852729320526,
      "num_tokens": 12793533.0,
      "step": 5820
    },
    {
      "entropy": 5.575941944122315,
      "epoch": 0.5781063914251687,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974081760293269,
      "loss": 5.3838,
      "mean_token_accuracy": 0.15085983872413636,
      "num_tokens": 12804389.0,
      "step": 5825
    },
    {
      "entropy": 5.647847127914429,
      "epoch": 0.5786026200873362,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004974028108411617,
      "loss": 5.4929,
      "mean_token_accuracy": 0.14978726506233214,
      "num_tokens": 12815672.0,
      "step": 5830
    },
    {
      "entropy": 5.676796579360962,
      "epoch": 0.5790988487495038,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004973974401378859,
      "loss": 5.4833,
      "mean_token_accuracy": 0.15055473297834396,
      "num_tokens": 12826113.0,
      "step": 5835
    },
    {
      "entropy": 5.590520238876342,
      "epoch": 0.5795950774116713,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973920639196329,
      "loss": 5.421,
      "mean_token_accuracy": 0.15158675462007523,
      "num_tokens": 12835941.0,
      "step": 5840
    },
    {
      "entropy": 5.642331504821778,
      "epoch": 0.5800913060738389,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973866821865358,
      "loss": 5.4907,
      "mean_token_accuracy": 0.1440528742969036,
      "num_tokens": 12846186.0,
      "step": 5845
    },
    {
      "entropy": 5.726757955551148,
      "epoch": 0.5805875347360063,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004973812949387283,
      "loss": 5.4691,
      "mean_token_accuracy": 0.1499832808971405,
      "num_tokens": 12856380.0,
      "step": 5850
    },
    {
      "entropy": 5.603196668624878,
      "epoch": 0.5810837633981739,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004973759021763439,
      "loss": 5.507,
      "mean_token_accuracy": 0.15144973993301392,
      "num_tokens": 12865999.0,
      "step": 5855
    },
    {
      "entropy": 5.636685609817505,
      "epoch": 0.5815799920603414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973705038995162,
      "loss": 5.3893,
      "mean_token_accuracy": 0.15000173300504685,
      "num_tokens": 12876388.0,
      "step": 5860
    },
    {
      "entropy": 5.606403160095215,
      "epoch": 0.582076220722509,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973651001083792,
      "loss": 5.3874,
      "mean_token_accuracy": 0.14825770109891892,
      "num_tokens": 12885862.0,
      "step": 5865
    },
    {
      "entropy": 5.584816598892212,
      "epoch": 0.5825724493846764,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973596908030669,
      "loss": 5.4797,
      "mean_token_accuracy": 0.14800314903259276,
      "num_tokens": 12897489.0,
      "step": 5870
    },
    {
      "entropy": 5.6848473072052,
      "epoch": 0.583068678046844,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973542759837135,
      "loss": 5.5115,
      "mean_token_accuracy": 0.14478374123573304,
      "num_tokens": 12907395.0,
      "step": 5875
    },
    {
      "entropy": 5.6072350025177,
      "epoch": 0.5835649067090115,
      "grad_norm": 0.875,
      "learning_rate": 0.000497348855650453,
      "loss": 5.4428,
      "mean_token_accuracy": 0.15394969433546066,
      "num_tokens": 12919439.0,
      "step": 5880
    },
    {
      "entropy": 5.679680824279785,
      "epoch": 0.584061135371179,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004973434298034201,
      "loss": 5.5469,
      "mean_token_accuracy": 0.14551895558834077,
      "num_tokens": 12931892.0,
      "step": 5885
    },
    {
      "entropy": 5.7289111614227295,
      "epoch": 0.5845573640333466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973379984427492,
      "loss": 5.5865,
      "mean_token_accuracy": 0.1434505380690098,
      "num_tokens": 12943189.0,
      "step": 5890
    },
    {
      "entropy": 5.691043043136597,
      "epoch": 0.585053592695514,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000497332561568575,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1459617681801319,
      "num_tokens": 12954662.0,
      "step": 5895
    },
    {
      "entropy": 5.628396463394165,
      "epoch": 0.5855498213576816,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973271191810325,
      "loss": 5.4609,
      "mean_token_accuracy": 0.14732947945594788,
      "num_tokens": 12964924.0,
      "step": 5900
    },
    {
      "entropy": 5.6725269794464115,
      "epoch": 0.5860460500198491,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004973216712802564,
      "loss": 5.5401,
      "mean_token_accuracy": 0.14416725933551788,
      "num_tokens": 12976483.0,
      "step": 5905
    },
    {
      "entropy": 5.625630235671997,
      "epoch": 0.5865422786820167,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000497316217866382,
      "loss": 5.437,
      "mean_token_accuracy": 0.15544704645872115,
      "num_tokens": 12988054.0,
      "step": 5910
    },
    {
      "entropy": 5.686157274246216,
      "epoch": 0.5870385073441842,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004973107589395444,
      "loss": 5.4957,
      "mean_token_accuracy": 0.15475226789712906,
      "num_tokens": 13000596.0,
      "step": 5915
    },
    {
      "entropy": 5.523101806640625,
      "epoch": 0.5875347360063518,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004973052944998791,
      "loss": 5.3173,
      "mean_token_accuracy": 0.1552049234509468,
      "num_tokens": 13012360.0,
      "step": 5920
    },
    {
      "entropy": 5.615753221511841,
      "epoch": 0.5880309646685192,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972998245475214,
      "loss": 5.4847,
      "mean_token_accuracy": 0.1488572522997856,
      "num_tokens": 13023806.0,
      "step": 5925
    },
    {
      "entropy": 5.640971994400024,
      "epoch": 0.5885271933306868,
      "grad_norm": 0.90625,
      "learning_rate": 0.000497294349082607,
      "loss": 5.405,
      "mean_token_accuracy": 0.1497488260269165,
      "num_tokens": 13034049.0,
      "step": 5930
    },
    {
      "entropy": 5.653295993804932,
      "epoch": 0.5890234219928543,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004972888681052719,
      "loss": 5.5142,
      "mean_token_accuracy": 0.14652270674705506,
      "num_tokens": 13045006.0,
      "step": 5935
    },
    {
      "entropy": 5.716931772232056,
      "epoch": 0.5895196506550219,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004972833816156518,
      "loss": 5.4253,
      "mean_token_accuracy": 0.14769369661808013,
      "num_tokens": 13055337.0,
      "step": 5940
    },
    {
      "entropy": 5.6028156757354735,
      "epoch": 0.5900158793171894,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972778896138828,
      "loss": 5.4078,
      "mean_token_accuracy": 0.15486730337142945,
      "num_tokens": 13065267.0,
      "step": 5945
    },
    {
      "entropy": 5.623899745941162,
      "epoch": 0.5905121079793569,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972723921001011,
      "loss": 5.4847,
      "mean_token_accuracy": 0.14757974594831466,
      "num_tokens": 13075173.0,
      "step": 5950
    },
    {
      "entropy": 5.687821960449218,
      "epoch": 0.5910083366415244,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972668890744431,
      "loss": 5.4719,
      "mean_token_accuracy": 0.14572732746601105,
      "num_tokens": 13086848.0,
      "step": 5955
    },
    {
      "entropy": 5.620846462249756,
      "epoch": 0.5915045653036919,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972613805370452,
      "loss": 5.4505,
      "mean_token_accuracy": 0.14893534108996392,
      "num_tokens": 13096865.0,
      "step": 5960
    },
    {
      "entropy": 5.585564231872558,
      "epoch": 0.5920007939658595,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004972558664880439,
      "loss": 5.4326,
      "mean_token_accuracy": 0.15041809976100923,
      "num_tokens": 13107907.0,
      "step": 5965
    },
    {
      "entropy": 5.569582748413086,
      "epoch": 0.592497022628027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000497250346927576,
      "loss": 5.3497,
      "mean_token_accuracy": 0.15736243277788162,
      "num_tokens": 13118171.0,
      "step": 5970
    },
    {
      "entropy": 5.619860219955444,
      "epoch": 0.5929932512901945,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972448218557785,
      "loss": 5.3339,
      "mean_token_accuracy": 0.15721192657947541,
      "num_tokens": 13127042.0,
      "step": 5975
    },
    {
      "entropy": 5.661720275878906,
      "epoch": 0.593489479952362,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004972392912727882,
      "loss": 5.5899,
      "mean_token_accuracy": 0.14976902306079865,
      "num_tokens": 13137026.0,
      "step": 5980
    },
    {
      "entropy": 5.64065465927124,
      "epoch": 0.5939857086145296,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004972337551787425,
      "loss": 5.444,
      "mean_token_accuracy": 0.15167996287345886,
      "num_tokens": 13146947.0,
      "step": 5985
    },
    {
      "entropy": 5.735600185394287,
      "epoch": 0.5944819372766971,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004972282135737785,
      "loss": 5.4972,
      "mean_token_accuracy": 0.14979468882083893,
      "num_tokens": 13158714.0,
      "step": 5990
    },
    {
      "entropy": 5.6617473602294925,
      "epoch": 0.5949781659388647,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972226664580337,
      "loss": 5.5063,
      "mean_token_accuracy": 0.14766890555620193,
      "num_tokens": 13170156.0,
      "step": 5995
    },
    {
      "entropy": 5.558388185501099,
      "epoch": 0.5954743946010321,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972171138316456,
      "loss": 5.4141,
      "mean_token_accuracy": 0.14991588592529298,
      "num_tokens": 13179919.0,
      "step": 6000
    },
    {
      "epoch": 0.5954743946010321,
      "eval_entropy": 5.478772575199873,
      "eval_loss": 5.523029327392578,
      "eval_mean_token_accuracy": 0.15203531923196922,
      "eval_num_tokens": 13179919.0,
      "eval_runtime": 20.5539,
      "eval_samples_per_second": 1638.085,
      "eval_steps_per_second": 204.779,
      "step": 6000
    },
    {
      "entropy": 5.683072948455811,
      "epoch": 0.5959706232631997,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004972115556947518,
      "loss": 5.4889,
      "mean_token_accuracy": 0.14972810372710227,
      "num_tokens": 13190777.0,
      "step": 6005
    },
    {
      "entropy": 5.699641561508178,
      "epoch": 0.5964668519253672,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004972059920474904,
      "loss": 5.4775,
      "mean_token_accuracy": 0.1444459415972233,
      "num_tokens": 13202179.0,
      "step": 6010
    },
    {
      "entropy": 5.610394668579102,
      "epoch": 0.5969630805875348,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004972004228899991,
      "loss": 5.429,
      "mean_token_accuracy": 0.15376247465610504,
      "num_tokens": 13212764.0,
      "step": 6015
    },
    {
      "entropy": 5.605767440795899,
      "epoch": 0.5974593092497023,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971948482224162,
      "loss": 5.4626,
      "mean_token_accuracy": 0.14900390803813934,
      "num_tokens": 13223343.0,
      "step": 6020
    },
    {
      "entropy": 5.676936149597168,
      "epoch": 0.5979555379118698,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971892680448798,
      "loss": 5.442,
      "mean_token_accuracy": 0.15177161395549774,
      "num_tokens": 13233839.0,
      "step": 6025
    },
    {
      "entropy": 5.705409002304077,
      "epoch": 0.5984517665740373,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004971836823575283,
      "loss": 5.5464,
      "mean_token_accuracy": 0.14601044803857804,
      "num_tokens": 13243854.0,
      "step": 6030
    },
    {
      "entropy": 5.690624094009399,
      "epoch": 0.5989479952362048,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004971780911605003,
      "loss": 5.4617,
      "mean_token_accuracy": 0.15180273205041886,
      "num_tokens": 13256073.0,
      "step": 6035
    },
    {
      "entropy": 5.63414511680603,
      "epoch": 0.5994442238983724,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971724944539343,
      "loss": 5.4996,
      "mean_token_accuracy": 0.14705433696508408,
      "num_tokens": 13267568.0,
      "step": 6040
    },
    {
      "entropy": 5.659912729263306,
      "epoch": 0.5999404525605398,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971668922379692,
      "loss": 5.4752,
      "mean_token_accuracy": 0.14323367699980735,
      "num_tokens": 13278829.0,
      "step": 6045
    },
    {
      "entropy": 5.646172332763672,
      "epoch": 0.6004366812227074,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497161284512744,
      "loss": 5.4441,
      "mean_token_accuracy": 0.15281899720430375,
      "num_tokens": 13289156.0,
      "step": 6050
    },
    {
      "entropy": 5.535249948501587,
      "epoch": 0.6009329098848749,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004971556712783976,
      "loss": 5.3175,
      "mean_token_accuracy": 0.15756299942731858,
      "num_tokens": 13299817.0,
      "step": 6055
    },
    {
      "entropy": 5.549290895462036,
      "epoch": 0.6014291385470425,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971500525350692,
      "loss": 5.3981,
      "mean_token_accuracy": 0.1537265032529831,
      "num_tokens": 13310290.0,
      "step": 6060
    },
    {
      "entropy": 5.585313129425049,
      "epoch": 0.60192536720921,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004971444282828983,
      "loss": 5.4953,
      "mean_token_accuracy": 0.14957028925418853,
      "num_tokens": 13321747.0,
      "step": 6065
    },
    {
      "entropy": 5.725924921035767,
      "epoch": 0.6024215958713776,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004971387985220242,
      "loss": 5.4427,
      "mean_token_accuracy": 0.15472995787858962,
      "num_tokens": 13333249.0,
      "step": 6070
    },
    {
      "entropy": 5.5447015285491945,
      "epoch": 0.602917824533545,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971331632525866,
      "loss": 5.4125,
      "mean_token_accuracy": 0.15427030101418496,
      "num_tokens": 13345336.0,
      "step": 6075
    },
    {
      "entropy": 5.704487752914429,
      "epoch": 0.6034140531957126,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497127522474725,
      "loss": 5.5776,
      "mean_token_accuracy": 0.1410766564309597,
      "num_tokens": 13355797.0,
      "step": 6080
    },
    {
      "entropy": 5.722353935241699,
      "epoch": 0.6039102818578801,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971218761885797,
      "loss": 5.5173,
      "mean_token_accuracy": 0.1494237206876278,
      "num_tokens": 13367058.0,
      "step": 6085
    },
    {
      "entropy": 5.626148748397827,
      "epoch": 0.6044065105200477,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004971162243942904,
      "loss": 5.4701,
      "mean_token_accuracy": 0.1484804220497608,
      "num_tokens": 13378579.0,
      "step": 6090
    },
    {
      "entropy": 5.6616168975830075,
      "epoch": 0.6049027391822152,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004971105670919975,
      "loss": 5.4369,
      "mean_token_accuracy": 0.15101820826530457,
      "num_tokens": 13389240.0,
      "step": 6095
    },
    {
      "entropy": 5.687566232681275,
      "epoch": 0.6053989678443827,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497104904281841,
      "loss": 5.5812,
      "mean_token_accuracy": 0.1416316546499729,
      "num_tokens": 13401895.0,
      "step": 6100
    },
    {
      "entropy": 5.703687334060669,
      "epoch": 0.6058951965065502,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970992359639616,
      "loss": 5.5449,
      "mean_token_accuracy": 0.13987237513065337,
      "num_tokens": 13413334.0,
      "step": 6105
    },
    {
      "entropy": 5.668133687973023,
      "epoch": 0.6063914251687177,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970935621384996,
      "loss": 5.4597,
      "mean_token_accuracy": 0.1523440472781658,
      "num_tokens": 13423371.0,
      "step": 6110
    },
    {
      "entropy": 5.682617950439453,
      "epoch": 0.6068876538308853,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004970878828055959,
      "loss": 5.4389,
      "mean_token_accuracy": 0.1553918406367302,
      "num_tokens": 13434782.0,
      "step": 6115
    },
    {
      "entropy": 5.619333124160766,
      "epoch": 0.6073838824930528,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004970821979653912,
      "loss": 5.4173,
      "mean_token_accuracy": 0.14554376155138016,
      "num_tokens": 13444493.0,
      "step": 6120
    },
    {
      "entropy": 5.5923535346984865,
      "epoch": 0.6078801111552203,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970765076180265,
      "loss": 5.419,
      "mean_token_accuracy": 0.1528603345155716,
      "num_tokens": 13455518.0,
      "step": 6125
    },
    {
      "entropy": 5.710022735595703,
      "epoch": 0.6083763398173878,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004970708117636431,
      "loss": 5.4969,
      "mean_token_accuracy": 0.15345498919487,
      "num_tokens": 13467631.0,
      "step": 6130
    },
    {
      "entropy": 5.618014335632324,
      "epoch": 0.6088725684795554,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000497065110402382,
      "loss": 5.4309,
      "mean_token_accuracy": 0.14801218509674072,
      "num_tokens": 13478217.0,
      "step": 6135
    },
    {
      "entropy": 5.620378017425537,
      "epoch": 0.6093687971417229,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970594035343846,
      "loss": 5.4208,
      "mean_token_accuracy": 0.15309711545705795,
      "num_tokens": 13489169.0,
      "step": 6140
    },
    {
      "entropy": 5.629087209701538,
      "epoch": 0.6098650258038905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004970536911597926,
      "loss": 5.436,
      "mean_token_accuracy": 0.14721201807260514,
      "num_tokens": 13499273.0,
      "step": 6145
    },
    {
      "entropy": 5.703438997268677,
      "epoch": 0.6103612544660579,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970479732787476,
      "loss": 5.4705,
      "mean_token_accuracy": 0.15132996588945388,
      "num_tokens": 13509664.0,
      "step": 6150
    },
    {
      "entropy": 5.606472826004028,
      "epoch": 0.6108574831282255,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004970422498913911,
      "loss": 5.3599,
      "mean_token_accuracy": 0.14891189187765122,
      "num_tokens": 13520288.0,
      "step": 6155
    },
    {
      "entropy": 5.583337450027466,
      "epoch": 0.611353711790393,
      "grad_norm": 0.875,
      "learning_rate": 0.0004970365209978655,
      "loss": 5.4508,
      "mean_token_accuracy": 0.1503366321325302,
      "num_tokens": 13531167.0,
      "step": 6160
    },
    {
      "entropy": 5.6312812805175785,
      "epoch": 0.6118499404525606,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004970307865983124,
      "loss": 5.4948,
      "mean_token_accuracy": 0.15321791023015977,
      "num_tokens": 13544109.0,
      "step": 6165
    },
    {
      "entropy": 5.676211786270142,
      "epoch": 0.6123461691147281,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004970250466928744,
      "loss": 5.423,
      "mean_token_accuracy": 0.1523990511894226,
      "num_tokens": 13554935.0,
      "step": 6170
    },
    {
      "entropy": 5.644534635543823,
      "epoch": 0.6128423977768956,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970193012816936,
      "loss": 5.447,
      "mean_token_accuracy": 0.14704892188310623,
      "num_tokens": 13565959.0,
      "step": 6175
    },
    {
      "entropy": 5.563157749176026,
      "epoch": 0.6133386264390631,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004970135503649125,
      "loss": 5.4599,
      "mean_token_accuracy": 0.1467708870768547,
      "num_tokens": 13577883.0,
      "step": 6180
    },
    {
      "entropy": 5.6085638999938965,
      "epoch": 0.6138348551012306,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004970077939426738,
      "loss": 5.4151,
      "mean_token_accuracy": 0.1497310072183609,
      "num_tokens": 13588597.0,
      "step": 6185
    },
    {
      "entropy": 5.6277988910675045,
      "epoch": 0.6143310837633982,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970020320151202,
      "loss": 5.3776,
      "mean_token_accuracy": 0.15045155137777327,
      "num_tokens": 13598772.0,
      "step": 6190
    },
    {
      "entropy": 5.680472135543823,
      "epoch": 0.6148273124255657,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969962645823946,
      "loss": 5.4615,
      "mean_token_accuracy": 0.14864098876714707,
      "num_tokens": 13609696.0,
      "step": 6195
    },
    {
      "entropy": 5.569826126098633,
      "epoch": 0.6153235410877332,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049699049164464,
      "loss": 5.443,
      "mean_token_accuracy": 0.1524896204471588,
      "num_tokens": 13619389.0,
      "step": 6200
    },
    {
      "entropy": 5.589665842056275,
      "epoch": 0.6158197697499007,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004969847132019995,
      "loss": 5.4128,
      "mean_token_accuracy": 0.15218402892351152,
      "num_tokens": 13632133.0,
      "step": 6205
    },
    {
      "entropy": 5.620619630813598,
      "epoch": 0.6163159984120683,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969789292546164,
      "loss": 5.3578,
      "mean_token_accuracy": 0.16294796764850616,
      "num_tokens": 13642715.0,
      "step": 6210
    },
    {
      "entropy": 5.590146780014038,
      "epoch": 0.6168122270742358,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969731398026343,
      "loss": 5.4578,
      "mean_token_accuracy": 0.14966798424720765,
      "num_tokens": 13653223.0,
      "step": 6215
    },
    {
      "entropy": 5.54031400680542,
      "epoch": 0.6173084557364034,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969673448461966,
      "loss": 5.4379,
      "mean_token_accuracy": 0.15165145099163055,
      "num_tokens": 13664206.0,
      "step": 6220
    },
    {
      "entropy": 5.641337442398071,
      "epoch": 0.6178046843985708,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000496961544385447,
      "loss": 5.4486,
      "mean_token_accuracy": 0.15540513396263123,
      "num_tokens": 13673974.0,
      "step": 6225
    },
    {
      "entropy": 5.600081539154052,
      "epoch": 0.6183009130607384,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004969557384205295,
      "loss": 5.3912,
      "mean_token_accuracy": 0.1543283775448799,
      "num_tokens": 13685189.0,
      "step": 6230
    },
    {
      "entropy": 5.643392515182495,
      "epoch": 0.6187971417229059,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004969499269515878,
      "loss": 5.4062,
      "mean_token_accuracy": 0.1569247215986252,
      "num_tokens": 13696181.0,
      "step": 6235
    },
    {
      "entropy": 5.714352369308472,
      "epoch": 0.6192933703850735,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004969441099787663,
      "loss": 5.5049,
      "mean_token_accuracy": 0.14568494707345964,
      "num_tokens": 13706719.0,
      "step": 6240
    },
    {
      "entropy": 5.667921352386474,
      "epoch": 0.619789599047241,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000496938287502209,
      "loss": 5.4679,
      "mean_token_accuracy": 0.14937815815210342,
      "num_tokens": 13716653.0,
      "step": 6245
    },
    {
      "entropy": 5.553180885314942,
      "epoch": 0.6202858277094085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004969324595220605,
      "loss": 5.3832,
      "mean_token_accuracy": 0.15513207465410234,
      "num_tokens": 13728220.0,
      "step": 6250
    },
    {
      "entropy": 5.5249491214752195,
      "epoch": 0.620782056371576,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969266260384652,
      "loss": 5.3487,
      "mean_token_accuracy": 0.15591224431991577,
      "num_tokens": 13738103.0,
      "step": 6255
    },
    {
      "entropy": 5.5973594188690186,
      "epoch": 0.6212782850337435,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004969207870515678,
      "loss": 5.4277,
      "mean_token_accuracy": 0.15172035992145538,
      "num_tokens": 13749845.0,
      "step": 6260
    },
    {
      "entropy": 5.627196502685547,
      "epoch": 0.6217745136959111,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496914942561513,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1480039581656456,
      "num_tokens": 13759983.0,
      "step": 6265
    },
    {
      "entropy": 5.704885482788086,
      "epoch": 0.6222707423580786,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969090925684459,
      "loss": 5.5466,
      "mean_token_accuracy": 0.15355510264635086,
      "num_tokens": 13770045.0,
      "step": 6270
    },
    {
      "entropy": 5.55440673828125,
      "epoch": 0.6227669710202461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004969032370725114,
      "loss": 5.3943,
      "mean_token_accuracy": 0.15380197763442993,
      "num_tokens": 13779823.0,
      "step": 6275
    },
    {
      "entropy": 5.600495672225952,
      "epoch": 0.6232631996824136,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004968973760738548,
      "loss": 5.4012,
      "mean_token_accuracy": 0.15008694380521775,
      "num_tokens": 13791165.0,
      "step": 6280
    },
    {
      "entropy": 5.694874572753906,
      "epoch": 0.6237594283445812,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968915095726216,
      "loss": 5.5017,
      "mean_token_accuracy": 0.14536357000470163,
      "num_tokens": 13802872.0,
      "step": 6285
    },
    {
      "entropy": 5.637458944320679,
      "epoch": 0.6242556570067487,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968856375689568,
      "loss": 5.4329,
      "mean_token_accuracy": 0.1524604231119156,
      "num_tokens": 13813893.0,
      "step": 6290
    },
    {
      "entropy": 5.622157192230224,
      "epoch": 0.6247518856689163,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968797600630065,
      "loss": 5.4951,
      "mean_token_accuracy": 0.15133117213845254,
      "num_tokens": 13823613.0,
      "step": 6295
    },
    {
      "entropy": 5.703844118118286,
      "epoch": 0.6252481143310837,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968738770549162,
      "loss": 5.5131,
      "mean_token_accuracy": 0.15189366936683654,
      "num_tokens": 13835413.0,
      "step": 6300
    },
    {
      "entropy": 5.6580039978027346,
      "epoch": 0.6257443429932513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968679885448318,
      "loss": 5.4351,
      "mean_token_accuracy": 0.14945173934102057,
      "num_tokens": 13845806.0,
      "step": 6305
    },
    {
      "entropy": 5.645914840698242,
      "epoch": 0.6262405716554188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004968620945328995,
      "loss": 5.5287,
      "mean_token_accuracy": 0.149135921895504,
      "num_tokens": 13856540.0,
      "step": 6310
    },
    {
      "entropy": 5.665123462677002,
      "epoch": 0.6267368003175864,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004968561950192652,
      "loss": 5.3848,
      "mean_token_accuracy": 0.15508267730474473,
      "num_tokens": 13867609.0,
      "step": 6315
    },
    {
      "entropy": 5.6689471244812015,
      "epoch": 0.6272330289797539,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968502900040753,
      "loss": 5.4488,
      "mean_token_accuracy": 0.1467331439256668,
      "num_tokens": 13878299.0,
      "step": 6320
    },
    {
      "entropy": 5.646787309646607,
      "epoch": 0.6277292576419214,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004968443794874764,
      "loss": 5.4609,
      "mean_token_accuracy": 0.14990048110485077,
      "num_tokens": 13889625.0,
      "step": 6325
    },
    {
      "entropy": 5.628449296951294,
      "epoch": 0.6282254863040889,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968384634696149,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1459612876176834,
      "num_tokens": 13899680.0,
      "step": 6330
    },
    {
      "entropy": 5.582142543792725,
      "epoch": 0.6287217149662565,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004968325419506375,
      "loss": 5.4055,
      "mean_token_accuracy": 0.15465095788240432,
      "num_tokens": 13910883.0,
      "step": 6335
    },
    {
      "entropy": 5.56435079574585,
      "epoch": 0.629217943628424,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004968266149306909,
      "loss": 5.4181,
      "mean_token_accuracy": 0.15198481529951097,
      "num_tokens": 13921442.0,
      "step": 6340
    },
    {
      "entropy": 5.6125036716461185,
      "epoch": 0.6297141722905915,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004968206824099224,
      "loss": 5.4423,
      "mean_token_accuracy": 0.15273929834365846,
      "num_tokens": 13930309.0,
      "step": 6345
    },
    {
      "entropy": 5.672091722488403,
      "epoch": 0.630210400952759,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004968147443884789,
      "loss": 5.4443,
      "mean_token_accuracy": 0.1561305418610573,
      "num_tokens": 13941346.0,
      "step": 6350
    },
    {
      "entropy": 5.602089071273804,
      "epoch": 0.6307066296149265,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004968088008665078,
      "loss": 5.4168,
      "mean_token_accuracy": 0.15606872290372847,
      "num_tokens": 13951933.0,
      "step": 6355
    },
    {
      "entropy": 5.575623893737793,
      "epoch": 0.6312028582770941,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968028518441563,
      "loss": 5.4675,
      "mean_token_accuracy": 0.1488841511309147,
      "num_tokens": 13963112.0,
      "step": 6360
    },
    {
      "entropy": 5.700068998336792,
      "epoch": 0.6316990869392616,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967968973215721,
      "loss": 5.5033,
      "mean_token_accuracy": 0.15618734508752824,
      "num_tokens": 13974218.0,
      "step": 6365
    },
    {
      "entropy": 5.684205865859985,
      "epoch": 0.6321953156014292,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967909372989027,
      "loss": 5.4283,
      "mean_token_accuracy": 0.15275898426771164,
      "num_tokens": 13984841.0,
      "step": 6370
    },
    {
      "entropy": 5.614372253417969,
      "epoch": 0.6326915442635966,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004967849717762959,
      "loss": 5.445,
      "mean_token_accuracy": 0.15169628262519835,
      "num_tokens": 13997504.0,
      "step": 6375
    },
    {
      "entropy": 5.558665800094604,
      "epoch": 0.6331877729257642,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004967790007538999,
      "loss": 5.3751,
      "mean_token_accuracy": 0.15826536267995833,
      "num_tokens": 14009244.0,
      "step": 6380
    },
    {
      "entropy": 5.5746911525726315,
      "epoch": 0.6336840015879317,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967730242318624,
      "loss": 5.4209,
      "mean_token_accuracy": 0.15277408808469772,
      "num_tokens": 14021190.0,
      "step": 6385
    },
    {
      "entropy": 5.578460645675659,
      "epoch": 0.6341802302500993,
      "grad_norm": 0.875,
      "learning_rate": 0.000496767042210332,
      "loss": 5.3779,
      "mean_token_accuracy": 0.15559731274843216,
      "num_tokens": 14032191.0,
      "step": 6390
    },
    {
      "entropy": 5.510378646850586,
      "epoch": 0.6346764589122668,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004967610546894566,
      "loss": 5.3528,
      "mean_token_accuracy": 0.16109540909528733,
      "num_tokens": 14042817.0,
      "step": 6395
    },
    {
      "entropy": 5.5974164485931395,
      "epoch": 0.6351726875744343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496755061669385,
      "loss": 5.3724,
      "mean_token_accuracy": 0.15573742240667343,
      "num_tokens": 14052725.0,
      "step": 6400
    },
    {
      "entropy": 5.635302019119263,
      "epoch": 0.6356689162366018,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967490631502656,
      "loss": 5.4518,
      "mean_token_accuracy": 0.1458180293440819,
      "num_tokens": 14063621.0,
      "step": 6405
    },
    {
      "entropy": 5.720311689376831,
      "epoch": 0.6361651448987694,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967430591322474,
      "loss": 5.5634,
      "mean_token_accuracy": 0.15249658226966858,
      "num_tokens": 14074350.0,
      "step": 6410
    },
    {
      "entropy": 5.68701205253601,
      "epoch": 0.6366613735609369,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004967370496154791,
      "loss": 5.4985,
      "mean_token_accuracy": 0.1474001705646515,
      "num_tokens": 14084600.0,
      "step": 6415
    },
    {
      "entropy": 5.680843734741211,
      "epoch": 0.6371576022231044,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967310346001098,
      "loss": 5.4066,
      "mean_token_accuracy": 0.15360569655895234,
      "num_tokens": 14094739.0,
      "step": 6420
    },
    {
      "entropy": 5.726302623748779,
      "epoch": 0.6376538308852719,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004967250140862885,
      "loss": 5.5625,
      "mean_token_accuracy": 0.1463690847158432,
      "num_tokens": 14106020.0,
      "step": 6425
    },
    {
      "entropy": 5.69241042137146,
      "epoch": 0.6381500595474394,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967189880741647,
      "loss": 5.5479,
      "mean_token_accuracy": 0.14727798104286194,
      "num_tokens": 14117699.0,
      "step": 6430
    },
    {
      "entropy": 5.592632150650024,
      "epoch": 0.638646288209607,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004967129565638878,
      "loss": 5.4268,
      "mean_token_accuracy": 0.15378784835338594,
      "num_tokens": 14128901.0,
      "step": 6435
    },
    {
      "entropy": 5.632250547409058,
      "epoch": 0.6391425168717745,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004967069195556073,
      "loss": 5.4734,
      "mean_token_accuracy": 0.1472574934363365,
      "num_tokens": 14140832.0,
      "step": 6440
    },
    {
      "entropy": 5.67312216758728,
      "epoch": 0.6396387455339421,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496700877049473,
      "loss": 5.5058,
      "mean_token_accuracy": 0.1492530107498169,
      "num_tokens": 14151936.0,
      "step": 6445
    },
    {
      "entropy": 5.676099681854248,
      "epoch": 0.6401349741961095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966948290456345,
      "loss": 5.4887,
      "mean_token_accuracy": 0.14892208576202393,
      "num_tokens": 14160945.0,
      "step": 6450
    },
    {
      "entropy": 5.595246458053589,
      "epoch": 0.6406312028582771,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966887755442421,
      "loss": 5.3844,
      "mean_token_accuracy": 0.15149885714054107,
      "num_tokens": 14172052.0,
      "step": 6455
    },
    {
      "entropy": 5.644949722290039,
      "epoch": 0.6411274315204446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004966827165454456,
      "loss": 5.4594,
      "mean_token_accuracy": 0.1524840459227562,
      "num_tokens": 14181788.0,
      "step": 6460
    },
    {
      "entropy": 5.702589893341065,
      "epoch": 0.6416236601826122,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004966766520493956,
      "loss": 5.547,
      "mean_token_accuracy": 0.1451253831386566,
      "num_tokens": 14193668.0,
      "step": 6465
    },
    {
      "entropy": 5.690148878097534,
      "epoch": 0.6421198888447797,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004966705820562422,
      "loss": 5.4153,
      "mean_token_accuracy": 0.1531184881925583,
      "num_tokens": 14203665.0,
      "step": 6470
    },
    {
      "entropy": 5.669411087036133,
      "epoch": 0.6426161175069472,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966645065661361,
      "loss": 5.4533,
      "mean_token_accuracy": 0.14717877879738808,
      "num_tokens": 14214213.0,
      "step": 6475
    },
    {
      "entropy": 5.6454425811767575,
      "epoch": 0.6431123461691147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966584255792279,
      "loss": 5.5499,
      "mean_token_accuracy": 0.14228935688734054,
      "num_tokens": 14225754.0,
      "step": 6480
    },
    {
      "entropy": 5.636998462677002,
      "epoch": 0.6436085748312823,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966523390956683,
      "loss": 5.4336,
      "mean_token_accuracy": 0.1544865034520626,
      "num_tokens": 14236508.0,
      "step": 6485
    },
    {
      "entropy": 5.671750497817993,
      "epoch": 0.6441048034934498,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966462471156083,
      "loss": 5.4785,
      "mean_token_accuracy": 0.15358012020587922,
      "num_tokens": 14247719.0,
      "step": 6490
    },
    {
      "entropy": 5.641434288024902,
      "epoch": 0.6446010321556173,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004966401496391989,
      "loss": 5.5355,
      "mean_token_accuracy": 0.14237177222967148,
      "num_tokens": 14258667.0,
      "step": 6495
    },
    {
      "entropy": 5.652969694137573,
      "epoch": 0.6450972608177848,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966340466665915,
      "loss": 5.5292,
      "mean_token_accuracy": 0.14700583443045617,
      "num_tokens": 14269666.0,
      "step": 6500
    },
    {
      "entropy": 5.585652637481689,
      "epoch": 0.6455934894799523,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004966279381979373,
      "loss": 5.4208,
      "mean_token_accuracy": 0.1501137524843216,
      "num_tokens": 14280862.0,
      "step": 6505
    },
    {
      "entropy": 5.694407844543457,
      "epoch": 0.6460897181421199,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004966218242333879,
      "loss": 5.5629,
      "mean_token_accuracy": 0.13654801473021508,
      "num_tokens": 14292505.0,
      "step": 6510
    },
    {
      "entropy": 5.6673437595367435,
      "epoch": 0.6465859468042874,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004966157047730945,
      "loss": 5.3638,
      "mean_token_accuracy": 0.15830052644014359,
      "num_tokens": 14302547.0,
      "step": 6515
    },
    {
      "entropy": 5.581522703170776,
      "epoch": 0.647082175466455,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004966095798172095,
      "loss": 5.508,
      "mean_token_accuracy": 0.15045141503214837,
      "num_tokens": 14313585.0,
      "step": 6520
    },
    {
      "entropy": 5.560984039306641,
      "epoch": 0.6475784041286224,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004966034493658843,
      "loss": 5.3706,
      "mean_token_accuracy": 0.14977592527866362,
      "num_tokens": 14323665.0,
      "step": 6525
    },
    {
      "entropy": 5.59638671875,
      "epoch": 0.64807463279079,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004965973134192711,
      "loss": 5.4184,
      "mean_token_accuracy": 0.15560718178749083,
      "num_tokens": 14334257.0,
      "step": 6530
    },
    {
      "entropy": 5.634187316894531,
      "epoch": 0.6485708614529575,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000496591171977522,
      "loss": 5.4219,
      "mean_token_accuracy": 0.15232618302106857,
      "num_tokens": 14345350.0,
      "step": 6535
    },
    {
      "entropy": 5.650139331817627,
      "epoch": 0.6490670901151251,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965850250407893,
      "loss": 5.4412,
      "mean_token_accuracy": 0.15271503776311873,
      "num_tokens": 14355593.0,
      "step": 6540
    },
    {
      "entropy": 5.601469993591309,
      "epoch": 0.6495633187772926,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004965788726092256,
      "loss": 5.3299,
      "mean_token_accuracy": 0.1551170215010643,
      "num_tokens": 14366405.0,
      "step": 6545
    },
    {
      "entropy": 5.5965948581695555,
      "epoch": 0.6500595474394602,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004965727146829832,
      "loss": 5.4199,
      "mean_token_accuracy": 0.14973414912819863,
      "num_tokens": 14377228.0,
      "step": 6550
    },
    {
      "entropy": 5.7201292514801025,
      "epoch": 0.6505557761016276,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004965665512622148,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1467569127678871,
      "num_tokens": 14387113.0,
      "step": 6555
    },
    {
      "entropy": 5.637718248367309,
      "epoch": 0.6510520047637952,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004965603823470735,
      "loss": 5.4231,
      "mean_token_accuracy": 0.15231161266565324,
      "num_tokens": 14397980.0,
      "step": 6560
    },
    {
      "entropy": 5.580999565124512,
      "epoch": 0.6515482334259627,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004965542079377122,
      "loss": 5.3601,
      "mean_token_accuracy": 0.15634892731904984,
      "num_tokens": 14409063.0,
      "step": 6565
    },
    {
      "entropy": 5.533416891098023,
      "epoch": 0.6520444620881302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965480280342839,
      "loss": 5.3436,
      "mean_token_accuracy": 0.1610360249876976,
      "num_tokens": 14421056.0,
      "step": 6570
    },
    {
      "entropy": 5.586903047561646,
      "epoch": 0.6525406907502977,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965418426369419,
      "loss": 5.3257,
      "mean_token_accuracy": 0.15886011570692063,
      "num_tokens": 14430778.0,
      "step": 6575
    },
    {
      "entropy": 5.580356407165527,
      "epoch": 0.6530369194124652,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004965356517458395,
      "loss": 5.3546,
      "mean_token_accuracy": 0.15466159135103225,
      "num_tokens": 14441544.0,
      "step": 6580
    },
    {
      "entropy": 5.580647897720337,
      "epoch": 0.6535331480746328,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965294553611303,
      "loss": 5.5068,
      "mean_token_accuracy": 0.15017633140087128,
      "num_tokens": 14453062.0,
      "step": 6585
    },
    {
      "entropy": 5.651913642883301,
      "epoch": 0.6540293767368003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000496523253482968,
      "loss": 5.4267,
      "mean_token_accuracy": 0.15264878869056703,
      "num_tokens": 14464402.0,
      "step": 6590
    },
    {
      "entropy": 5.685528326034546,
      "epoch": 0.6545256053989679,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965170461115065,
      "loss": 5.5195,
      "mean_token_accuracy": 0.14986869543790818,
      "num_tokens": 14476595.0,
      "step": 6595
    },
    {
      "entropy": 5.687401676177979,
      "epoch": 0.6550218340611353,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004965108332468994,
      "loss": 5.5023,
      "mean_token_accuracy": 0.14899416714906694,
      "num_tokens": 14487298.0,
      "step": 6600
    },
    {
      "entropy": 5.6138396739959715,
      "epoch": 0.6555180627233029,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004965046148893011,
      "loss": 5.3939,
      "mean_token_accuracy": 0.15507970601320267,
      "num_tokens": 14499708.0,
      "step": 6605
    },
    {
      "entropy": 5.609611320495605,
      "epoch": 0.6560142913854704,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004964983910388654,
      "loss": 5.4574,
      "mean_token_accuracy": 0.15213153511285782,
      "num_tokens": 14511773.0,
      "step": 6610
    },
    {
      "entropy": 5.5898024559021,
      "epoch": 0.656510520047638,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004964921616957471,
      "loss": 5.3588,
      "mean_token_accuracy": 0.1501300022006035,
      "num_tokens": 14523024.0,
      "step": 6615
    },
    {
      "entropy": 5.6398614883422855,
      "epoch": 0.6570067487098055,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964859268601005,
      "loss": 5.4185,
      "mean_token_accuracy": 0.15466833487153053,
      "num_tokens": 14532915.0,
      "step": 6620
    },
    {
      "entropy": 5.619332218170166,
      "epoch": 0.657502977371973,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964796865320801,
      "loss": 5.4119,
      "mean_token_accuracy": 0.15669008195400239,
      "num_tokens": 14543627.0,
      "step": 6625
    },
    {
      "entropy": 5.586162900924682,
      "epoch": 0.6579992060341405,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004964734407118406,
      "loss": 5.3898,
      "mean_token_accuracy": 0.1568761721253395,
      "num_tokens": 14555181.0,
      "step": 6630
    },
    {
      "entropy": 5.605987882614135,
      "epoch": 0.6584954346963081,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964671893995372,
      "loss": 5.391,
      "mean_token_accuracy": 0.1466468259692192,
      "num_tokens": 14565542.0,
      "step": 6635
    },
    {
      "entropy": 5.657824945449829,
      "epoch": 0.6589916633584756,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964609325953245,
      "loss": 5.4511,
      "mean_token_accuracy": 0.15445198267698287,
      "num_tokens": 14575395.0,
      "step": 6640
    },
    {
      "entropy": 5.646509838104248,
      "epoch": 0.6594878920206431,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004964546702993581,
      "loss": 5.4776,
      "mean_token_accuracy": 0.14906896650791168,
      "num_tokens": 14587206.0,
      "step": 6645
    },
    {
      "entropy": 5.675960683822632,
      "epoch": 0.6599841206828106,
      "grad_norm": 0.79296875,
      "learning_rate": 0.000496448402511793,
      "loss": 5.4824,
      "mean_token_accuracy": 0.15160994827747346,
      "num_tokens": 14599048.0,
      "step": 6650
    },
    {
      "entropy": 5.665557718276977,
      "epoch": 0.6604803493449781,
      "grad_norm": 0.875,
      "learning_rate": 0.0004964421292327847,
      "loss": 5.4516,
      "mean_token_accuracy": 0.15329598039388656,
      "num_tokens": 14609633.0,
      "step": 6655
    },
    {
      "entropy": 5.635668468475342,
      "epoch": 0.6609765780071457,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004964358504624887,
      "loss": 5.3537,
      "mean_token_accuracy": 0.15587745010852813,
      "num_tokens": 14620373.0,
      "step": 6660
    },
    {
      "entropy": 5.583084583282471,
      "epoch": 0.6614728066693132,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964295662010608,
      "loss": 5.3039,
      "mean_token_accuracy": 0.1617210403084755,
      "num_tokens": 14630463.0,
      "step": 6665
    },
    {
      "entropy": 5.713048124313355,
      "epoch": 0.6619690353314808,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004964232764486568,
      "loss": 5.5537,
      "mean_token_accuracy": 0.14458150044083595,
      "num_tokens": 14641523.0,
      "step": 6670
    },
    {
      "entropy": 5.60595965385437,
      "epoch": 0.6624652639936482,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004964169812054326,
      "loss": 5.3536,
      "mean_token_accuracy": 0.15471443086862563,
      "num_tokens": 14651812.0,
      "step": 6675
    },
    {
      "entropy": 5.560608339309693,
      "epoch": 0.6629614926558158,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004964106804715444,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1587303936481476,
      "num_tokens": 14661982.0,
      "step": 6680
    },
    {
      "entropy": 5.597832489013672,
      "epoch": 0.6634577213179833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964043742471486,
      "loss": 5.3855,
      "mean_token_accuracy": 0.1622820019721985,
      "num_tokens": 14671981.0,
      "step": 6685
    },
    {
      "entropy": 5.653746891021728,
      "epoch": 0.6639539499801509,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004963980625324012,
      "loss": 5.3996,
      "mean_token_accuracy": 0.15135825723409652,
      "num_tokens": 14682478.0,
      "step": 6690
    },
    {
      "entropy": 5.660880136489868,
      "epoch": 0.6644501786423184,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000496391745327459,
      "loss": 5.4944,
      "mean_token_accuracy": 0.14261270612478255,
      "num_tokens": 14695197.0,
      "step": 6695
    },
    {
      "entropy": 5.590805959701538,
      "epoch": 0.664946407304486,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004963854226324785,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1525746315717697,
      "num_tokens": 14705702.0,
      "step": 6700
    },
    {
      "entropy": 5.616341018676758,
      "epoch": 0.6654426359666534,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004963790944476167,
      "loss": 5.4125,
      "mean_token_accuracy": 0.15438311994075776,
      "num_tokens": 14715370.0,
      "step": 6705
    },
    {
      "entropy": 5.667642116546631,
      "epoch": 0.665938864628821,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004963727607730303,
      "loss": 5.4202,
      "mean_token_accuracy": 0.15620658695697784,
      "num_tokens": 14725481.0,
      "step": 6710
    },
    {
      "entropy": 5.597318124771118,
      "epoch": 0.6664350932909885,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004963664216088764,
      "loss": 5.3422,
      "mean_token_accuracy": 0.15659624040126802,
      "num_tokens": 14735250.0,
      "step": 6715
    },
    {
      "entropy": 5.595610570907593,
      "epoch": 0.666931321953156,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963600769553124,
      "loss": 5.3644,
      "mean_token_accuracy": 0.16038820445537566,
      "num_tokens": 14744741.0,
      "step": 6720
    },
    {
      "entropy": 5.679464864730835,
      "epoch": 0.6674275506153236,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004963537268124953,
      "loss": 5.3989,
      "mean_token_accuracy": 0.15407125800848007,
      "num_tokens": 14755865.0,
      "step": 6725
    },
    {
      "entropy": 5.539860677719116,
      "epoch": 0.667923779277491,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963473711805829,
      "loss": 5.3434,
      "mean_token_accuracy": 0.15569494664669037,
      "num_tokens": 14766154.0,
      "step": 6730
    },
    {
      "entropy": 5.610124969482422,
      "epoch": 0.6684200079396586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004963410100597326,
      "loss": 5.3894,
      "mean_token_accuracy": 0.15085127651691438,
      "num_tokens": 14776940.0,
      "step": 6735
    },
    {
      "entropy": 5.598522758483886,
      "epoch": 0.6689162366018261,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004963346434501021,
      "loss": 5.499,
      "mean_token_accuracy": 0.14878811538219452,
      "num_tokens": 14788367.0,
      "step": 6740
    },
    {
      "entropy": 5.591743230819702,
      "epoch": 0.6694124652639937,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004963282713518495,
      "loss": 5.3988,
      "mean_token_accuracy": 0.15474075675010682,
      "num_tokens": 14798751.0,
      "step": 6745
    },
    {
      "entropy": 5.532715320587158,
      "epoch": 0.6699086939261611,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963218937651326,
      "loss": 5.336,
      "mean_token_accuracy": 0.15949305891990662,
      "num_tokens": 14808748.0,
      "step": 6750
    },
    {
      "entropy": 5.651119947433472,
      "epoch": 0.6704049225883287,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004963155106901098,
      "loss": 5.5111,
      "mean_token_accuracy": 0.14742460250854492,
      "num_tokens": 14820524.0,
      "step": 6755
    },
    {
      "entropy": 5.7370702743530275,
      "epoch": 0.6709011512504962,
      "grad_norm": 1.0,
      "learning_rate": 0.0004963091221269391,
      "loss": 5.541,
      "mean_token_accuracy": 0.1430164247751236,
      "num_tokens": 14832485.0,
      "step": 6760
    },
    {
      "entropy": 5.612604999542237,
      "epoch": 0.6713973799126638,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000496302728075779,
      "loss": 5.342,
      "mean_token_accuracy": 0.15255440920591354,
      "num_tokens": 14843951.0,
      "step": 6765
    },
    {
      "entropy": 5.572727155685425,
      "epoch": 0.6718936085748313,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004962963285367882,
      "loss": 5.4438,
      "mean_token_accuracy": 0.1534359022974968,
      "num_tokens": 14855502.0,
      "step": 6770
    },
    {
      "entropy": 5.588790941238403,
      "epoch": 0.6723898372369989,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004962899235101253,
      "loss": 5.3519,
      "mean_token_accuracy": 0.15730117708444596,
      "num_tokens": 14866305.0,
      "step": 6775
    },
    {
      "entropy": 5.598705005645752,
      "epoch": 0.6728860658991663,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004962835129959491,
      "loss": 5.3838,
      "mean_token_accuracy": 0.14917287975549698,
      "num_tokens": 14877222.0,
      "step": 6780
    },
    {
      "entropy": 5.568091201782226,
      "epoch": 0.6733822945613339,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962770969944186,
      "loss": 5.4506,
      "mean_token_accuracy": 0.15169098675251008,
      "num_tokens": 14888099.0,
      "step": 6785
    },
    {
      "entropy": 5.610668182373047,
      "epoch": 0.6738785232235014,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000496270675505693,
      "loss": 5.3845,
      "mean_token_accuracy": 0.1605300113558769,
      "num_tokens": 14899042.0,
      "step": 6790
    },
    {
      "entropy": 5.603315448760986,
      "epoch": 0.6743747518856689,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962642485299314,
      "loss": 5.4337,
      "mean_token_accuracy": 0.1523062467575073,
      "num_tokens": 14908837.0,
      "step": 6795
    },
    {
      "entropy": 5.597096061706543,
      "epoch": 0.6748709805478365,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004962578160672931,
      "loss": 5.3807,
      "mean_token_accuracy": 0.155751334130764,
      "num_tokens": 14920773.0,
      "step": 6800
    },
    {
      "entropy": 5.5771277904510494,
      "epoch": 0.6753672092100039,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004962513781179379,
      "loss": 5.3653,
      "mean_token_accuracy": 0.15376887023448943,
      "num_tokens": 14930970.0,
      "step": 6805
    },
    {
      "entropy": 5.584058952331543,
      "epoch": 0.6758634378721715,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004962449346820251,
      "loss": 5.4416,
      "mean_token_accuracy": 0.14662599712610244,
      "num_tokens": 14942378.0,
      "step": 6810
    },
    {
      "entropy": 5.616071653366089,
      "epoch": 0.676359666534339,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004962384857597148,
      "loss": 5.4632,
      "mean_token_accuracy": 0.14783129915595056,
      "num_tokens": 14953020.0,
      "step": 6815
    },
    {
      "entropy": 5.701265001296997,
      "epoch": 0.6768558951965066,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962320313511668,
      "loss": 5.5239,
      "mean_token_accuracy": 0.15097203254699706,
      "num_tokens": 14962340.0,
      "step": 6820
    },
    {
      "entropy": 5.677660274505615,
      "epoch": 0.677352123858674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962255714565409,
      "loss": 5.4174,
      "mean_token_accuracy": 0.15791946202516555,
      "num_tokens": 14972792.0,
      "step": 6825
    },
    {
      "entropy": 5.624913120269776,
      "epoch": 0.6778483525208416,
      "grad_norm": 0.875,
      "learning_rate": 0.0004962191060759978,
      "loss": 5.4034,
      "mean_token_accuracy": 0.15495871752500534,
      "num_tokens": 14983698.0,
      "step": 6830
    },
    {
      "entropy": 5.613697481155396,
      "epoch": 0.6783445811830091,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004962126352096973,
      "loss": 5.3589,
      "mean_token_accuracy": 0.15519290864467622,
      "num_tokens": 14994583.0,
      "step": 6835
    },
    {
      "entropy": 5.518886137008667,
      "epoch": 0.6788408098451767,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004962061588578004,
      "loss": 5.3203,
      "mean_token_accuracy": 0.15874038487672806,
      "num_tokens": 15005325.0,
      "step": 6840
    },
    {
      "entropy": 5.6368483543396,
      "epoch": 0.6793370385073442,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004961996770204673,
      "loss": 5.3836,
      "mean_token_accuracy": 0.15178438425064086,
      "num_tokens": 15015155.0,
      "step": 6845
    },
    {
      "entropy": 5.610763645172119,
      "epoch": 0.6798332671695118,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961931896978588,
      "loss": 5.4384,
      "mean_token_accuracy": 0.15444398522377015,
      "num_tokens": 15026529.0,
      "step": 6850
    },
    {
      "entropy": 5.551306867599488,
      "epoch": 0.6803294958316792,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004961866968901359,
      "loss": 5.3775,
      "mean_token_accuracy": 0.15853386521339416,
      "num_tokens": 15037632.0,
      "step": 6855
    },
    {
      "entropy": 5.577917289733887,
      "epoch": 0.6808257244938468,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004961801985974595,
      "loss": 5.4516,
      "mean_token_accuracy": 0.15589179694652558,
      "num_tokens": 15050397.0,
      "step": 6860
    },
    {
      "entropy": 5.58919587135315,
      "epoch": 0.6813219531560143,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004961736948199907,
      "loss": 5.3544,
      "mean_token_accuracy": 0.15355462580919266,
      "num_tokens": 15061505.0,
      "step": 6865
    },
    {
      "entropy": 5.689698696136475,
      "epoch": 0.6818181818181818,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004961671855578909,
      "loss": 5.5508,
      "mean_token_accuracy": 0.14556234925985337,
      "num_tokens": 15073752.0,
      "step": 6870
    },
    {
      "entropy": 5.674998950958252,
      "epoch": 0.6823144104803494,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004961606708113216,
      "loss": 5.4438,
      "mean_token_accuracy": 0.1530633293092251,
      "num_tokens": 15084730.0,
      "step": 6875
    },
    {
      "entropy": 5.617377233505249,
      "epoch": 0.6828106391425168,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004961541505804441,
      "loss": 5.3581,
      "mean_token_accuracy": 0.155072158575058,
      "num_tokens": 15095811.0,
      "step": 6880
    },
    {
      "entropy": 5.594923210144043,
      "epoch": 0.6833068678046844,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004961476248654202,
      "loss": 5.3439,
      "mean_token_accuracy": 0.1578742802143097,
      "num_tokens": 15106102.0,
      "step": 6885
    },
    {
      "entropy": 5.634280109405518,
      "epoch": 0.6838030964668519,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961410936664118,
      "loss": 5.4489,
      "mean_token_accuracy": 0.14765285551548005,
      "num_tokens": 15115907.0,
      "step": 6890
    },
    {
      "entropy": 5.602192115783692,
      "epoch": 0.6842993251290195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961345569835807,
      "loss": 5.4007,
      "mean_token_accuracy": 0.15584126561880113,
      "num_tokens": 15126788.0,
      "step": 6895
    },
    {
      "entropy": 5.581991577148438,
      "epoch": 0.684795553791187,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961280148170892,
      "loss": 5.3695,
      "mean_token_accuracy": 0.1499946653842926,
      "num_tokens": 15137814.0,
      "step": 6900
    },
    {
      "entropy": 5.674746084213257,
      "epoch": 0.6852917824533545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961214671670994,
      "loss": 5.5962,
      "mean_token_accuracy": 0.13972103893756865,
      "num_tokens": 15148597.0,
      "step": 6905
    },
    {
      "entropy": 5.6451982975006105,
      "epoch": 0.685788011115522,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961149140337737,
      "loss": 5.4007,
      "mean_token_accuracy": 0.1503000497817993,
      "num_tokens": 15158985.0,
      "step": 6910
    },
    {
      "entropy": 5.611239051818847,
      "epoch": 0.6862842397776896,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004961083554172745,
      "loss": 5.3845,
      "mean_token_accuracy": 0.151298738270998,
      "num_tokens": 15170242.0,
      "step": 6915
    },
    {
      "entropy": 5.56635684967041,
      "epoch": 0.6867804684398571,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004961017913177646,
      "loss": 5.3649,
      "mean_token_accuracy": 0.16083198934793472,
      "num_tokens": 15180353.0,
      "step": 6920
    },
    {
      "entropy": 5.529801321029663,
      "epoch": 0.6872766971020247,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004960952217354066,
      "loss": 5.3066,
      "mean_token_accuracy": 0.16144633144140244,
      "num_tokens": 15191955.0,
      "step": 6925
    },
    {
      "entropy": 5.566470813751221,
      "epoch": 0.6877729257641921,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960886466703636,
      "loss": 5.3801,
      "mean_token_accuracy": 0.15237442702054976,
      "num_tokens": 15202088.0,
      "step": 6930
    },
    {
      "entropy": 5.549537897109985,
      "epoch": 0.6882691544263597,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960820661227987,
      "loss": 5.3354,
      "mean_token_accuracy": 0.15506885349750518,
      "num_tokens": 15212291.0,
      "step": 6935
    },
    {
      "entropy": 5.561124181747436,
      "epoch": 0.6887653830885272,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960754800928748,
      "loss": 5.3739,
      "mean_token_accuracy": 0.15877065658569336,
      "num_tokens": 15223315.0,
      "step": 6940
    },
    {
      "entropy": 5.706064367294312,
      "epoch": 0.6892616117506947,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004960688885807553,
      "loss": 5.4952,
      "mean_token_accuracy": 0.14694880396127702,
      "num_tokens": 15233416.0,
      "step": 6945
    },
    {
      "entropy": 5.6424641609191895,
      "epoch": 0.6897578404128623,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004960622915866038,
      "loss": 5.3233,
      "mean_token_accuracy": 0.1593453496694565,
      "num_tokens": 15242951.0,
      "step": 6950
    },
    {
      "entropy": 5.56283597946167,
      "epoch": 0.6902540690750297,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960556891105839,
      "loss": 5.4182,
      "mean_token_accuracy": 0.15205707401037216,
      "num_tokens": 15253145.0,
      "step": 6955
    },
    {
      "entropy": 5.554887676239014,
      "epoch": 0.6907502977371973,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004960490811528591,
      "loss": 5.3611,
      "mean_token_accuracy": 0.15836797803640365,
      "num_tokens": 15263712.0,
      "step": 6960
    },
    {
      "entropy": 5.679096794128418,
      "epoch": 0.6912465263993648,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960424677135935,
      "loss": 5.4831,
      "mean_token_accuracy": 0.1556924656033516,
      "num_tokens": 15273609.0,
      "step": 6965
    },
    {
      "entropy": 5.5833167552948,
      "epoch": 0.6917427550615324,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004960358487929511,
      "loss": 5.3479,
      "mean_token_accuracy": 0.16443779617547988,
      "num_tokens": 15283934.0,
      "step": 6970
    },
    {
      "entropy": 5.582957315444946,
      "epoch": 0.6922389837236999,
      "grad_norm": 0.875,
      "learning_rate": 0.0004960292243910958,
      "loss": 5.4054,
      "mean_token_accuracy": 0.15455612689256668,
      "num_tokens": 15294394.0,
      "step": 6975
    },
    {
      "entropy": 5.585826778411866,
      "epoch": 0.6927352123858674,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496022594508192,
      "loss": 5.3709,
      "mean_token_accuracy": 0.15916962474584578,
      "num_tokens": 15304622.0,
      "step": 6980
    },
    {
      "entropy": 5.574977827072144,
      "epoch": 0.6932314410480349,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004960159591444041,
      "loss": 5.2895,
      "mean_token_accuracy": 0.16055725365877152,
      "num_tokens": 15314094.0,
      "step": 6985
    },
    {
      "entropy": 5.557066249847412,
      "epoch": 0.6937276697102025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960093182998968,
      "loss": 5.4049,
      "mean_token_accuracy": 0.1594743475317955,
      "num_tokens": 15324313.0,
      "step": 6990
    },
    {
      "entropy": 5.486831617355347,
      "epoch": 0.69422389837237,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004960026719748346,
      "loss": 5.2859,
      "mean_token_accuracy": 0.17195601314306258,
      "num_tokens": 15335195.0,
      "step": 6995
    },
    {
      "entropy": 5.699875020980835,
      "epoch": 0.6947201270345376,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004959960201693822,
      "loss": 5.5114,
      "mean_token_accuracy": 0.14829787239432335,
      "num_tokens": 15345798.0,
      "step": 7000
    },
    {
      "entropy": 5.5973199844360355,
      "epoch": 0.695216355696705,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004959893628837048,
      "loss": 5.3352,
      "mean_token_accuracy": 0.15762802958488464,
      "num_tokens": 15356340.0,
      "step": 7005
    },
    {
      "entropy": 5.6082206726074215,
      "epoch": 0.6957125843588726,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004959827001179675,
      "loss": 5.4499,
      "mean_token_accuracy": 0.14720271080732344,
      "num_tokens": 15367795.0,
      "step": 7010
    },
    {
      "entropy": 5.646885538101197,
      "epoch": 0.6962088130210401,
      "grad_norm": 1.0,
      "learning_rate": 0.0004959760318723353,
      "loss": 5.5663,
      "mean_token_accuracy": 0.14655987843871116,
      "num_tokens": 15379214.0,
      "step": 7015
    },
    {
      "entropy": 5.631622695922852,
      "epoch": 0.6967050416832076,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959693581469736,
      "loss": 5.4068,
      "mean_token_accuracy": 0.1613534316420555,
      "num_tokens": 15390242.0,
      "step": 7020
    },
    {
      "entropy": 5.606883955001831,
      "epoch": 0.6972012703453752,
      "grad_norm": 0.875,
      "learning_rate": 0.0004959626789420481,
      "loss": 5.3548,
      "mean_token_accuracy": 0.159724161028862,
      "num_tokens": 15402600.0,
      "step": 7025
    },
    {
      "entropy": 5.575189828872681,
      "epoch": 0.6976974990075426,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004959559942577243,
      "loss": 5.3843,
      "mean_token_accuracy": 0.15733335465192794,
      "num_tokens": 15412883.0,
      "step": 7030
    },
    {
      "entropy": 5.6446130752563475,
      "epoch": 0.6981937276697102,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004959493040941678,
      "loss": 5.4336,
      "mean_token_accuracy": 0.1509387031197548,
      "num_tokens": 15425919.0,
      "step": 7035
    },
    {
      "entropy": 5.577643537521363,
      "epoch": 0.6986899563318777,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004959426084515448,
      "loss": 5.3551,
      "mean_token_accuracy": 0.15517423152923585,
      "num_tokens": 15437604.0,
      "step": 7040
    },
    {
      "entropy": 5.563399171829223,
      "epoch": 0.6991861849940453,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004959359073300212,
      "loss": 5.4241,
      "mean_token_accuracy": 0.15425959154963492,
      "num_tokens": 15449023.0,
      "step": 7045
    },
    {
      "entropy": 5.6797730922698975,
      "epoch": 0.6996824136562128,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004959292007297631,
      "loss": 5.4717,
      "mean_token_accuracy": 0.1513185828924179,
      "num_tokens": 15459567.0,
      "step": 7050
    },
    {
      "entropy": 5.640298128128052,
      "epoch": 0.7001786423183803,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000495922488650937,
      "loss": 5.3914,
      "mean_token_accuracy": 0.1548449717462063,
      "num_tokens": 15470950.0,
      "step": 7055
    },
    {
      "entropy": 5.569140291213989,
      "epoch": 0.7006748709805478,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959157710937092,
      "loss": 5.3836,
      "mean_token_accuracy": 0.15649599134922026,
      "num_tokens": 15481258.0,
      "step": 7060
    },
    {
      "entropy": 5.6256061553955075,
      "epoch": 0.7011710996427154,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004959090480582462,
      "loss": 5.3191,
      "mean_token_accuracy": 0.15883412659168245,
      "num_tokens": 15491672.0,
      "step": 7065
    },
    {
      "entropy": 5.560047817230225,
      "epoch": 0.7016673283048829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004959023195447148,
      "loss": 5.3853,
      "mean_token_accuracy": 0.15852536559104918,
      "num_tokens": 15502342.0,
      "step": 7070
    },
    {
      "entropy": 5.582231235504151,
      "epoch": 0.7021635569670505,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000495895585553282,
      "loss": 5.4143,
      "mean_token_accuracy": 0.15081254094839097,
      "num_tokens": 15513430.0,
      "step": 7075
    },
    {
      "entropy": 5.781231689453125,
      "epoch": 0.7026597856292179,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004958888460841147,
      "loss": 5.5158,
      "mean_token_accuracy": 0.15422331988811494,
      "num_tokens": 15524889.0,
      "step": 7080
    },
    {
      "entropy": 5.5678489208221436,
      "epoch": 0.7031560142913855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004958821011373798,
      "loss": 5.3336,
      "mean_token_accuracy": 0.15573736429214477,
      "num_tokens": 15534998.0,
      "step": 7085
    },
    {
      "entropy": 5.543598461151123,
      "epoch": 0.703652242953553,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958753507132449,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1567114382982254,
      "num_tokens": 15544752.0,
      "step": 7090
    },
    {
      "entropy": 5.607834434509277,
      "epoch": 0.7041484716157205,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004958685948118772,
      "loss": 5.3747,
      "mean_token_accuracy": 0.15421359241008759,
      "num_tokens": 15555490.0,
      "step": 7095
    },
    {
      "entropy": 5.578544473648071,
      "epoch": 0.7046447002778881,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004958618334334443,
      "loss": 5.3159,
      "mean_token_accuracy": 0.15533821582794188,
      "num_tokens": 15565752.0,
      "step": 7100
    },
    {
      "entropy": 5.589881753921508,
      "epoch": 0.7051409289400555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004958550665781138,
      "loss": 5.4442,
      "mean_token_accuracy": 0.1503690093755722,
      "num_tokens": 15576368.0,
      "step": 7105
    },
    {
      "entropy": 5.674482822418213,
      "epoch": 0.7056371576022231,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958482942460536,
      "loss": 5.4308,
      "mean_token_accuracy": 0.15454249009490012,
      "num_tokens": 15587972.0,
      "step": 7110
    },
    {
      "entropy": 5.71412034034729,
      "epoch": 0.7061333862643906,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004958415164374316,
      "loss": 5.4671,
      "mean_token_accuracy": 0.15848549604415893,
      "num_tokens": 15599951.0,
      "step": 7115
    },
    {
      "entropy": 5.583090353012085,
      "epoch": 0.7066296149265582,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004958347331524158,
      "loss": 5.453,
      "mean_token_accuracy": 0.14946739971637726,
      "num_tokens": 15610183.0,
      "step": 7120
    },
    {
      "entropy": 5.593395709991455,
      "epoch": 0.7071258435887257,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004958279443911745,
      "loss": 5.4555,
      "mean_token_accuracy": 0.1472437486052513,
      "num_tokens": 15621074.0,
      "step": 7125
    },
    {
      "entropy": 5.695850610733032,
      "epoch": 0.7076220722508932,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000495821150153876,
      "loss": 5.4657,
      "mean_token_accuracy": 0.14712751731276513,
      "num_tokens": 15632752.0,
      "step": 7130
    },
    {
      "entropy": 5.611667346954346,
      "epoch": 0.7081183009130607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000495814350440689,
      "loss": 5.4484,
      "mean_token_accuracy": 0.1472303293645382,
      "num_tokens": 15644703.0,
      "step": 7135
    },
    {
      "entropy": 5.579992580413818,
      "epoch": 0.7086145295752283,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004958075452517816,
      "loss": 5.4164,
      "mean_token_accuracy": 0.1512021616101265,
      "num_tokens": 15655278.0,
      "step": 7140
    },
    {
      "entropy": 5.628891897201538,
      "epoch": 0.7091107582373958,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004958007345873232,
      "loss": 5.4299,
      "mean_token_accuracy": 0.15474658012390136,
      "num_tokens": 15665998.0,
      "step": 7145
    },
    {
      "entropy": 5.594137668609619,
      "epoch": 0.7096069868995634,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004957939184474821,
      "loss": 5.3061,
      "mean_token_accuracy": 0.15685224980115892,
      "num_tokens": 15676343.0,
      "step": 7150
    },
    {
      "entropy": 5.614185810089111,
      "epoch": 0.7101032155617308,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004957870968324277,
      "loss": 5.4065,
      "mean_token_accuracy": 0.15425229743123053,
      "num_tokens": 15686520.0,
      "step": 7155
    },
    {
      "entropy": 5.630849409103393,
      "epoch": 0.7105994442238984,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957802697423291,
      "loss": 5.5124,
      "mean_token_accuracy": 0.1502341777086258,
      "num_tokens": 15698302.0,
      "step": 7160
    },
    {
      "entropy": 5.546805000305175,
      "epoch": 0.7110956728860659,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004957734371773554,
      "loss": 5.2726,
      "mean_token_accuracy": 0.16012369394302367,
      "num_tokens": 15708462.0,
      "step": 7165
    },
    {
      "entropy": 5.621671199798584,
      "epoch": 0.7115919015482334,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004957665991376763,
      "loss": 5.3908,
      "mean_token_accuracy": 0.15984175652265548,
      "num_tokens": 15719234.0,
      "step": 7170
    },
    {
      "entropy": 5.569719362258911,
      "epoch": 0.712088130210401,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004957597556234613,
      "loss": 5.4155,
      "mean_token_accuracy": 0.15240118354558946,
      "num_tokens": 15732245.0,
      "step": 7175
    },
    {
      "entropy": 5.613059139251709,
      "epoch": 0.7125843588725684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004957529066348799,
      "loss": 5.3913,
      "mean_token_accuracy": 0.15765198916196824,
      "num_tokens": 15742403.0,
      "step": 7180
    },
    {
      "entropy": 5.54534420967102,
      "epoch": 0.713080587534736,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957460521721022,
      "loss": 5.2901,
      "mean_token_accuracy": 0.16082171946763993,
      "num_tokens": 15752100.0,
      "step": 7185
    },
    {
      "entropy": 5.594834232330323,
      "epoch": 0.7135768161969035,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000495739192235298,
      "loss": 5.3986,
      "mean_token_accuracy": 0.15319614708423615,
      "num_tokens": 15762315.0,
      "step": 7190
    },
    {
      "entropy": 5.5770289421081545,
      "epoch": 0.7140730448590711,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957323268246376,
      "loss": 5.4175,
      "mean_token_accuracy": 0.15682317465543746,
      "num_tokens": 15773841.0,
      "step": 7195
    },
    {
      "entropy": 5.623834276199341,
      "epoch": 0.7145692735212386,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004957254559402911,
      "loss": 5.347,
      "mean_token_accuracy": 0.1560734361410141,
      "num_tokens": 15785511.0,
      "step": 7200
    },
    {
      "entropy": 5.5824196338653564,
      "epoch": 0.7150655021834061,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004957185795824288,
      "loss": 5.3801,
      "mean_token_accuracy": 0.14989668726921082,
      "num_tokens": 15795394.0,
      "step": 7205
    },
    {
      "entropy": 5.549114561080932,
      "epoch": 0.7155617308455736,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004957116977512214,
      "loss": 5.3492,
      "mean_token_accuracy": 0.15484311580657958,
      "num_tokens": 15805894.0,
      "step": 7210
    },
    {
      "entropy": 5.64172477722168,
      "epoch": 0.7160579595077412,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004957048104468395,
      "loss": 5.4079,
      "mean_token_accuracy": 0.1467766210436821,
      "num_tokens": 15816595.0,
      "step": 7215
    },
    {
      "entropy": 5.587030124664307,
      "epoch": 0.7165541881699087,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004956979176694539,
      "loss": 5.3578,
      "mean_token_accuracy": 0.15960478484630586,
      "num_tokens": 15827638.0,
      "step": 7220
    },
    {
      "entropy": 5.576122236251831,
      "epoch": 0.7170504168320763,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004956910194192355,
      "loss": 5.4186,
      "mean_token_accuracy": 0.15302726328372956,
      "num_tokens": 15838437.0,
      "step": 7225
    },
    {
      "entropy": 5.722928905487061,
      "epoch": 0.7175466454942437,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004956841156963553,
      "loss": 5.5326,
      "mean_token_accuracy": 0.151919137686491,
      "num_tokens": 15850110.0,
      "step": 7230
    },
    {
      "entropy": 5.587578201293946,
      "epoch": 0.7180428741564113,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004956772065009846,
      "loss": 5.3805,
      "mean_token_accuracy": 0.15135951042175294,
      "num_tokens": 15861786.0,
      "step": 7235
    },
    {
      "entropy": 5.602078294754028,
      "epoch": 0.7185391028185788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956702918332948,
      "loss": 5.3673,
      "mean_token_accuracy": 0.15241504162549974,
      "num_tokens": 15871574.0,
      "step": 7240
    },
    {
      "entropy": 5.5809221267700195,
      "epoch": 0.7190353314807463,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004956633716934572,
      "loss": 5.3525,
      "mean_token_accuracy": 0.16035836488008498,
      "num_tokens": 15881944.0,
      "step": 7245
    },
    {
      "entropy": 5.5627188205719,
      "epoch": 0.7195315601429139,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004956564460816433,
      "loss": 5.3609,
      "mean_token_accuracy": 0.1546861171722412,
      "num_tokens": 15892511.0,
      "step": 7250
    },
    {
      "entropy": 5.5861094951629635,
      "epoch": 0.7200277888050813,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004956495149980252,
      "loss": 5.3873,
      "mean_token_accuracy": 0.15233663022518157,
      "num_tokens": 15903387.0,
      "step": 7255
    },
    {
      "entropy": 5.615819263458252,
      "epoch": 0.7205240174672489,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004956425784427745,
      "loss": 5.4143,
      "mean_token_accuracy": 0.15342193990945815,
      "num_tokens": 15915117.0,
      "step": 7260
    },
    {
      "entropy": 5.557542133331299,
      "epoch": 0.7210202461294164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956356364160634,
      "loss": 5.3762,
      "mean_token_accuracy": 0.15235370248556138,
      "num_tokens": 15925163.0,
      "step": 7265
    },
    {
      "entropy": 5.498903036117554,
      "epoch": 0.721516474791584,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004956286889180639,
      "loss": 5.2651,
      "mean_token_accuracy": 0.1643815517425537,
      "num_tokens": 15935021.0,
      "step": 7270
    },
    {
      "entropy": 5.539653873443603,
      "epoch": 0.7220127034537515,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004956217359489483,
      "loss": 5.3801,
      "mean_token_accuracy": 0.15152700394392013,
      "num_tokens": 15945416.0,
      "step": 7275
    },
    {
      "entropy": 5.665227031707763,
      "epoch": 0.722508932115919,
      "grad_norm": 0.921875,
      "learning_rate": 0.000495614777508889,
      "loss": 5.4898,
      "mean_token_accuracy": 0.1517569750547409,
      "num_tokens": 15955388.0,
      "step": 7280
    },
    {
      "entropy": 5.694360446929932,
      "epoch": 0.7230051607780865,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956078135980586,
      "loss": 5.5008,
      "mean_token_accuracy": 0.15092244371771812,
      "num_tokens": 15966721.0,
      "step": 7285
    },
    {
      "entropy": 5.601458549499512,
      "epoch": 0.7235013894402541,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004956008442166299,
      "loss": 5.4266,
      "mean_token_accuracy": 0.1519925594329834,
      "num_tokens": 15976388.0,
      "step": 7290
    },
    {
      "entropy": 5.5649971008300785,
      "epoch": 0.7239976181024216,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004955938693647756,
      "loss": 5.3161,
      "mean_token_accuracy": 0.1596965342760086,
      "num_tokens": 15986632.0,
      "step": 7295
    },
    {
      "entropy": 5.579634761810302,
      "epoch": 0.7244938467645892,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004955868890426686,
      "loss": 5.3718,
      "mean_token_accuracy": 0.15614695847034454,
      "num_tokens": 15996626.0,
      "step": 7300
    },
    {
      "entropy": 5.557637882232666,
      "epoch": 0.7249900754267566,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955799032504821,
      "loss": 5.34,
      "mean_token_accuracy": 0.1535207599401474,
      "num_tokens": 16007055.0,
      "step": 7305
    },
    {
      "entropy": 5.661793804168701,
      "epoch": 0.7254863040889242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004955729119883893,
      "loss": 5.4072,
      "mean_token_accuracy": 0.15194595754146575,
      "num_tokens": 16019087.0,
      "step": 7310
    },
    {
      "entropy": 5.617228221893311,
      "epoch": 0.7259825327510917,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955659152565636,
      "loss": 5.4201,
      "mean_token_accuracy": 0.15152042955160142,
      "num_tokens": 16029980.0,
      "step": 7315
    },
    {
      "entropy": 5.598032999038696,
      "epoch": 0.7264787614132592,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004955589130551784,
      "loss": 5.3854,
      "mean_token_accuracy": 0.15649388059973718,
      "num_tokens": 16040932.0,
      "step": 7320
    },
    {
      "entropy": 5.498610544204712,
      "epoch": 0.7269749900754268,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004955519053844076,
      "loss": 5.3517,
      "mean_token_accuracy": 0.15688288509845733,
      "num_tokens": 16051809.0,
      "step": 7325
    },
    {
      "entropy": 5.6194291591644285,
      "epoch": 0.7274712187375942,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004955448922444246,
      "loss": 5.3954,
      "mean_token_accuracy": 0.15169398561120034,
      "num_tokens": 16062957.0,
      "step": 7330
    },
    {
      "entropy": 5.618951511383057,
      "epoch": 0.7279674473997618,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004955378736354036,
      "loss": 5.4359,
      "mean_token_accuracy": 0.15555522292852403,
      "num_tokens": 16073592.0,
      "step": 7335
    },
    {
      "entropy": 5.5267417430877686,
      "epoch": 0.7284636760619293,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004955308495575185,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1569704942405224,
      "num_tokens": 16085054.0,
      "step": 7340
    },
    {
      "entropy": 5.6387309551239015,
      "epoch": 0.7289599047240969,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955238200109436,
      "loss": 5.4644,
      "mean_token_accuracy": 0.15437906086444855,
      "num_tokens": 16096936.0,
      "step": 7345
    },
    {
      "entropy": 5.651801013946534,
      "epoch": 0.7294561333862644,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004955167849958531,
      "loss": 5.4811,
      "mean_token_accuracy": 0.15084182769060134,
      "num_tokens": 16107673.0,
      "step": 7350
    },
    {
      "entropy": 5.607907247543335,
      "epoch": 0.729952362048432,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004955097445124215,
      "loss": 5.3368,
      "mean_token_accuracy": 0.1613706097006798,
      "num_tokens": 16117941.0,
      "step": 7355
    },
    {
      "entropy": 5.583903789520264,
      "epoch": 0.7304485907105994,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004955026985608235,
      "loss": 5.3077,
      "mean_token_accuracy": 0.16149525493383407,
      "num_tokens": 16128737.0,
      "step": 7360
    },
    {
      "entropy": 5.685665845870972,
      "epoch": 0.730944819372767,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004954956471412336,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1574809119105339,
      "num_tokens": 16139599.0,
      "step": 7365
    },
    {
      "entropy": 5.501325035095215,
      "epoch": 0.7314410480349345,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004954885902538269,
      "loss": 5.2892,
      "mean_token_accuracy": 0.15115222632884978,
      "num_tokens": 16150796.0,
      "step": 7370
    },
    {
      "entropy": 5.560714960098267,
      "epoch": 0.7319372766971021,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004954815278987781,
      "loss": 5.365,
      "mean_token_accuracy": 0.15895064026117325,
      "num_tokens": 16161236.0,
      "step": 7375
    },
    {
      "entropy": 5.66231894493103,
      "epoch": 0.7324335053592695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004954744600762627,
      "loss": 5.4619,
      "mean_token_accuracy": 0.15455430150032043,
      "num_tokens": 16172210.0,
      "step": 7380
    },
    {
      "entropy": 5.527791357040405,
      "epoch": 0.7329297340214371,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004954673867864555,
      "loss": 5.3316,
      "mean_token_accuracy": 0.16065908074378968,
      "num_tokens": 16182950.0,
      "step": 7385
    },
    {
      "entropy": 5.51721420288086,
      "epoch": 0.7334259626836046,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004954603080295323,
      "loss": 5.2524,
      "mean_token_accuracy": 0.16600435972213745,
      "num_tokens": 16193971.0,
      "step": 7390
    },
    {
      "entropy": 5.570493364334107,
      "epoch": 0.7339221913457721,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004954532238056686,
      "loss": 5.3674,
      "mean_token_accuracy": 0.15529176890850066,
      "num_tokens": 16205774.0,
      "step": 7395
    },
    {
      "entropy": 5.572083282470703,
      "epoch": 0.7344184200079397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004954461341150399,
      "loss": 5.362,
      "mean_token_accuracy": 0.1557315319776535,
      "num_tokens": 16216626.0,
      "step": 7400
    },
    {
      "entropy": 5.561210250854492,
      "epoch": 0.7349146486701071,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004954390389578222,
      "loss": 5.4363,
      "mean_token_accuracy": 0.15050967186689376,
      "num_tokens": 16227748.0,
      "step": 7405
    },
    {
      "entropy": 5.640095138549805,
      "epoch": 0.7354108773322747,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004954319383341912,
      "loss": 5.4237,
      "mean_token_accuracy": 0.15483318120241166,
      "num_tokens": 16240390.0,
      "step": 7410
    },
    {
      "entropy": 5.602118349075317,
      "epoch": 0.7359071059944422,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954248322443232,
      "loss": 5.3522,
      "mean_token_accuracy": 0.16627731472253798,
      "num_tokens": 16250563.0,
      "step": 7415
    },
    {
      "entropy": 5.582110643386841,
      "epoch": 0.7364033346566098,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004954177206883943,
      "loss": 5.4271,
      "mean_token_accuracy": 0.15135748088359832,
      "num_tokens": 16260722.0,
      "step": 7420
    },
    {
      "entropy": 5.722535419464111,
      "epoch": 0.7368995633187773,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954106036665809,
      "loss": 5.4345,
      "mean_token_accuracy": 0.1533150114119053,
      "num_tokens": 16270258.0,
      "step": 7425
    },
    {
      "entropy": 5.594661712646484,
      "epoch": 0.7373957919809448,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004954034811790594,
      "loss": 5.2944,
      "mean_token_accuracy": 0.16092931479215622,
      "num_tokens": 16279504.0,
      "step": 7430
    },
    {
      "entropy": 5.538282823562622,
      "epoch": 0.7378920206431123,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004953963532260067,
      "loss": 5.3203,
      "mean_token_accuracy": 0.1602314069867134,
      "num_tokens": 16291327.0,
      "step": 7435
    },
    {
      "entropy": 5.60131573677063,
      "epoch": 0.7383882493052799,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953892198075991,
      "loss": 5.4593,
      "mean_token_accuracy": 0.15599284321069717,
      "num_tokens": 16302778.0,
      "step": 7440
    },
    {
      "entropy": 5.641455745697021,
      "epoch": 0.7388844779674474,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004953820809240138,
      "loss": 5.399,
      "mean_token_accuracy": 0.14711376279592514,
      "num_tokens": 16314373.0,
      "step": 7445
    },
    {
      "entropy": 5.578816080093384,
      "epoch": 0.739380706629615,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953749365754279,
      "loss": 5.3695,
      "mean_token_accuracy": 0.15730762034654616,
      "num_tokens": 16324518.0,
      "step": 7450
    },
    {
      "entropy": 5.601818323135376,
      "epoch": 0.7398769352917824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004953677867620184,
      "loss": 5.3004,
      "mean_token_accuracy": 0.1593911826610565,
      "num_tokens": 16335657.0,
      "step": 7455
    },
    {
      "entropy": 5.611733150482178,
      "epoch": 0.74037316395395,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004953606314839626,
      "loss": 5.4982,
      "mean_token_accuracy": 0.14880106151103972,
      "num_tokens": 16346497.0,
      "step": 7460
    },
    {
      "entropy": 5.533340835571289,
      "epoch": 0.7408693926161175,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000495353470741438,
      "loss": 5.3701,
      "mean_token_accuracy": 0.15519127398729324,
      "num_tokens": 16358175.0,
      "step": 7465
    },
    {
      "entropy": 5.647481870651245,
      "epoch": 0.741365621278285,
      "grad_norm": 1.390625,
      "learning_rate": 0.000495346304534622,
      "loss": 5.4807,
      "mean_token_accuracy": 0.14439089670777322,
      "num_tokens": 16369803.0,
      "step": 7470
    },
    {
      "entropy": 5.54910831451416,
      "epoch": 0.7418618499404526,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004953391328636926,
      "loss": 5.1857,
      "mean_token_accuracy": 0.1628032833337784,
      "num_tokens": 16380535.0,
      "step": 7475
    },
    {
      "entropy": 5.538540410995483,
      "epoch": 0.74235807860262,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004953319557288275,
      "loss": 5.3731,
      "mean_token_accuracy": 0.15650602877140046,
      "num_tokens": 16390518.0,
      "step": 7480
    },
    {
      "entropy": 5.5667534351348875,
      "epoch": 0.7428543072647876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004953247731302046,
      "loss": 5.3834,
      "mean_token_accuracy": 0.15064625591039657,
      "num_tokens": 16400627.0,
      "step": 7485
    },
    {
      "entropy": 5.589830112457276,
      "epoch": 0.7433505359269551,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004953175850680022,
      "loss": 5.4521,
      "mean_token_accuracy": 0.15772558152675628,
      "num_tokens": 16411136.0,
      "step": 7490
    },
    {
      "entropy": 5.674487590789795,
      "epoch": 0.7438467645891227,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004953103915423983,
      "loss": 5.3776,
      "mean_token_accuracy": 0.15505346804857253,
      "num_tokens": 16422830.0,
      "step": 7495
    },
    {
      "entropy": 5.553043222427368,
      "epoch": 0.7443429932512902,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004953031925535715,
      "loss": 5.3357,
      "mean_token_accuracy": 0.15908195227384567,
      "num_tokens": 16434158.0,
      "step": 7500
    },
    {
      "entropy": 5.569729137420654,
      "epoch": 0.7448392219134577,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004952959881017003,
      "loss": 5.3596,
      "mean_token_accuracy": 0.16043863147497178,
      "num_tokens": 16444562.0,
      "step": 7505
    },
    {
      "entropy": 5.608988428115845,
      "epoch": 0.7453354505756252,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004952887781869631,
      "loss": 5.3912,
      "mean_token_accuracy": 0.15401030629873275,
      "num_tokens": 16456353.0,
      "step": 7510
    },
    {
      "entropy": 5.616450309753418,
      "epoch": 0.7458316792377928,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004952815628095391,
      "loss": 5.4121,
      "mean_token_accuracy": 0.14678518548607827,
      "num_tokens": 16465756.0,
      "step": 7515
    },
    {
      "entropy": 5.552458381652832,
      "epoch": 0.7463279078999603,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004952743419696069,
      "loss": 5.3387,
      "mean_token_accuracy": 0.15742378681898117,
      "num_tokens": 16477562.0,
      "step": 7520
    },
    {
      "entropy": 5.610256290435791,
      "epoch": 0.7468241365621279,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952671156673456,
      "loss": 5.3957,
      "mean_token_accuracy": 0.15251319259405136,
      "num_tokens": 16488134.0,
      "step": 7525
    },
    {
      "entropy": 5.637909078598023,
      "epoch": 0.7473203652242953,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004952598839029345,
      "loss": 5.3778,
      "mean_token_accuracy": 0.15602533370256425,
      "num_tokens": 16499146.0,
      "step": 7530
    },
    {
      "entropy": 5.621365642547607,
      "epoch": 0.7478165938864629,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004952526466765529,
      "loss": 5.4272,
      "mean_token_accuracy": 0.15652930215001107,
      "num_tokens": 16510334.0,
      "step": 7535
    },
    {
      "entropy": 5.573215770721435,
      "epoch": 0.7483128225486304,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004952454039883804,
      "loss": 5.2958,
      "mean_token_accuracy": 0.1592641294002533,
      "num_tokens": 16520574.0,
      "step": 7540
    },
    {
      "entropy": 5.622744464874268,
      "epoch": 0.7488090512107979,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004952381558385963,
      "loss": 5.4465,
      "mean_token_accuracy": 0.14939613193273543,
      "num_tokens": 16532246.0,
      "step": 7545
    },
    {
      "entropy": 5.7219959735870365,
      "epoch": 0.7493052798729655,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004952309022273805,
      "loss": 5.4094,
      "mean_token_accuracy": 0.14916437566280366,
      "num_tokens": 16541756.0,
      "step": 7550
    },
    {
      "entropy": 5.609796857833862,
      "epoch": 0.7498015085351329,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952236431549129,
      "loss": 5.4176,
      "mean_token_accuracy": 0.15520833283662797,
      "num_tokens": 16551061.0,
      "step": 7555
    },
    {
      "entropy": 5.556709909439087,
      "epoch": 0.7502977371973005,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004952163786213735,
      "loss": 5.3647,
      "mean_token_accuracy": 0.15090104043483735,
      "num_tokens": 16561155.0,
      "step": 7560
    },
    {
      "entropy": 5.592098903656006,
      "epoch": 0.750793965859468,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004952091086269425,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15054828077554702,
      "num_tokens": 16571843.0,
      "step": 7565
    },
    {
      "entropy": 5.604906797409058,
      "epoch": 0.7512901945216356,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952018331718001,
      "loss": 5.4003,
      "mean_token_accuracy": 0.15466262102127076,
      "num_tokens": 16583395.0,
      "step": 7570
    },
    {
      "entropy": 5.655341863632202,
      "epoch": 0.7517864231838031,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004951945522561268,
      "loss": 5.4624,
      "mean_token_accuracy": 0.1517023354768753,
      "num_tokens": 16594545.0,
      "step": 7575
    },
    {
      "entropy": 5.678570079803467,
      "epoch": 0.7522826518459707,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004951872658801029,
      "loss": 5.507,
      "mean_token_accuracy": 0.15081263929605485,
      "num_tokens": 16605901.0,
      "step": 7580
    },
    {
      "entropy": 5.6098236560821535,
      "epoch": 0.7527788805081381,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951799740439093,
      "loss": 5.4833,
      "mean_token_accuracy": 0.1473713755607605,
      "num_tokens": 16616783.0,
      "step": 7585
    },
    {
      "entropy": 5.622238349914551,
      "epoch": 0.7532751091703057,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000495172676747727,
      "loss": 5.4023,
      "mean_token_accuracy": 0.1594322606921196,
      "num_tokens": 16628655.0,
      "step": 7590
    },
    {
      "entropy": 5.606210422515869,
      "epoch": 0.7537713378324732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004951653739917366,
      "loss": 5.4016,
      "mean_token_accuracy": 0.1529039815068245,
      "num_tokens": 16639403.0,
      "step": 7595
    },
    {
      "entropy": 5.647537326812744,
      "epoch": 0.7542675664946408,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004951580657761192,
      "loss": 5.4183,
      "mean_token_accuracy": 0.16141234040260316,
      "num_tokens": 16649937.0,
      "step": 7600
    },
    {
      "entropy": 5.5170728206634525,
      "epoch": 0.7547637951568082,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004951507521010563,
      "loss": 5.3056,
      "mean_token_accuracy": 0.16003890931606293,
      "num_tokens": 16661466.0,
      "step": 7605
    },
    {
      "entropy": 5.519859409332275,
      "epoch": 0.7552600238189758,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004951434329667293,
      "loss": 5.327,
      "mean_token_accuracy": 0.15801397264003753,
      "num_tokens": 16673362.0,
      "step": 7610
    },
    {
      "entropy": 5.601373243331909,
      "epoch": 0.7557562524811433,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004951361083733192,
      "loss": 5.3705,
      "mean_token_accuracy": 0.14978507161140442,
      "num_tokens": 16685555.0,
      "step": 7615
    },
    {
      "entropy": 5.604953575134277,
      "epoch": 0.7562524811433108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951287783210082,
      "loss": 5.3466,
      "mean_token_accuracy": 0.15240003615617753,
      "num_tokens": 16696006.0,
      "step": 7620
    },
    {
      "entropy": 5.566999721527099,
      "epoch": 0.7567487098054784,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004951214428099778,
      "loss": 5.351,
      "mean_token_accuracy": 0.15930738300085068,
      "num_tokens": 16706784.0,
      "step": 7625
    },
    {
      "entropy": 5.523137903213501,
      "epoch": 0.7572449384676458,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004951141018404099,
      "loss": 5.3332,
      "mean_token_accuracy": 0.16055645644664765,
      "num_tokens": 16718492.0,
      "step": 7630
    },
    {
      "entropy": 5.5530671119689945,
      "epoch": 0.7577411671298134,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004951067554124865,
      "loss": 5.3903,
      "mean_token_accuracy": 0.15504445731639863,
      "num_tokens": 16729711.0,
      "step": 7635
    },
    {
      "entropy": 5.5830793380737305,
      "epoch": 0.7582373957919809,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00049509940352639,
      "loss": 5.3704,
      "mean_token_accuracy": 0.15670461803674698,
      "num_tokens": 16739913.0,
      "step": 7640
    },
    {
      "entropy": 5.673083305358887,
      "epoch": 0.7587336244541485,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950920461823024,
      "loss": 5.4187,
      "mean_token_accuracy": 0.15199341326951982,
      "num_tokens": 16752273.0,
      "step": 7645
    },
    {
      "entropy": 5.593998908996582,
      "epoch": 0.759229853116316,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950846833804064,
      "loss": 5.3471,
      "mean_token_accuracy": 0.15649023056030273,
      "num_tokens": 16762050.0,
      "step": 7650
    },
    {
      "entropy": 5.529874563217163,
      "epoch": 0.7597260817784836,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950773151208845,
      "loss": 5.412,
      "mean_token_accuracy": 0.15717475563287736,
      "num_tokens": 16772376.0,
      "step": 7655
    },
    {
      "entropy": 5.565098428726197,
      "epoch": 0.760222310440651,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004950699414039194,
      "loss": 5.3731,
      "mean_token_accuracy": 0.15431706309318544,
      "num_tokens": 16783307.0,
      "step": 7660
    },
    {
      "entropy": 5.602860641479492,
      "epoch": 0.7607185391028186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000495062562229694,
      "loss": 5.2476,
      "mean_token_accuracy": 0.1612496167421341,
      "num_tokens": 16793582.0,
      "step": 7665
    },
    {
      "entropy": 5.539808034896851,
      "epoch": 0.7612147677649861,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004950551775983912,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1535532459616661,
      "num_tokens": 16805160.0,
      "step": 7670
    },
    {
      "entropy": 5.559559488296509,
      "epoch": 0.7617109964271537,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950477875101942,
      "loss": 5.4269,
      "mean_token_accuracy": 0.14969281032681464,
      "num_tokens": 16816108.0,
      "step": 7675
    },
    {
      "entropy": 5.524907207489013,
      "epoch": 0.7622072250893211,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004950403919652863,
      "loss": 5.3089,
      "mean_token_accuracy": 0.16031151711940766,
      "num_tokens": 16828126.0,
      "step": 7680
    },
    {
      "entropy": 5.602362632751465,
      "epoch": 0.7627034537514887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004950329909638507,
      "loss": 5.3704,
      "mean_token_accuracy": 0.15363027155399323,
      "num_tokens": 16838186.0,
      "step": 7685
    },
    {
      "entropy": 5.5945127487182615,
      "epoch": 0.7631996824136562,
      "grad_norm": 0.8046875,
      "learning_rate": 0.000495025584506071,
      "loss": 5.3183,
      "mean_token_accuracy": 0.15459179431200026,
      "num_tokens": 16848588.0,
      "step": 7690
    },
    {
      "entropy": 5.615201568603515,
      "epoch": 0.7636959110758237,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004950181725921311,
      "loss": 5.4291,
      "mean_token_accuracy": 0.15589270889759063,
      "num_tokens": 16860269.0,
      "step": 7695
    },
    {
      "entropy": 5.568172121047974,
      "epoch": 0.7641921397379913,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004950107552222145,
      "loss": 5.4368,
      "mean_token_accuracy": 0.15411009043455123,
      "num_tokens": 16872221.0,
      "step": 7700
    },
    {
      "entropy": 5.563562679290771,
      "epoch": 0.7646883684001587,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004950033323965052,
      "loss": 5.3117,
      "mean_token_accuracy": 0.15144989490509034,
      "num_tokens": 16882588.0,
      "step": 7705
    },
    {
      "entropy": 5.619341802597046,
      "epoch": 0.7651845970623263,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949959041151875,
      "loss": 5.3733,
      "mean_token_accuracy": 0.15467335879802704,
      "num_tokens": 16894278.0,
      "step": 7710
    },
    {
      "entropy": 5.6074230670928955,
      "epoch": 0.7656808257244938,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004949884703784453,
      "loss": 5.3514,
      "mean_token_accuracy": 0.15782455503940582,
      "num_tokens": 16905448.0,
      "step": 7715
    },
    {
      "entropy": 5.541156959533692,
      "epoch": 0.7661770543866614,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000494981031186463,
      "loss": 5.3813,
      "mean_token_accuracy": 0.15570028126239777,
      "num_tokens": 16914960.0,
      "step": 7720
    },
    {
      "entropy": 5.601539230346679,
      "epoch": 0.7666732830488289,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949735865394253,
      "loss": 5.4519,
      "mean_token_accuracy": 0.15187731087207795,
      "num_tokens": 16926839.0,
      "step": 7725
    },
    {
      "entropy": 5.608623313903808,
      "epoch": 0.7671695117109965,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004949661364375167,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1634967252612114,
      "num_tokens": 16938553.0,
      "step": 7730
    },
    {
      "entropy": 5.615213394165039,
      "epoch": 0.7676657403731639,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949586808809217,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1534518599510193,
      "num_tokens": 16947987.0,
      "step": 7735
    },
    {
      "entropy": 5.53449821472168,
      "epoch": 0.7681619690353315,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004949512198698256,
      "loss": 5.3359,
      "mean_token_accuracy": 0.155488258600235,
      "num_tokens": 16959359.0,
      "step": 7740
    },
    {
      "entropy": 5.582101345062256,
      "epoch": 0.768658197697499,
      "grad_norm": 0.875,
      "learning_rate": 0.000494943753404413,
      "loss": 5.3893,
      "mean_token_accuracy": 0.1577385276556015,
      "num_tokens": 16970731.0,
      "step": 7745
    },
    {
      "entropy": 5.532139158248901,
      "epoch": 0.7691544263596666,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004949362814848693,
      "loss": 5.3404,
      "mean_token_accuracy": 0.15632505416870118,
      "num_tokens": 16981868.0,
      "step": 7750
    },
    {
      "entropy": 5.636943101882935,
      "epoch": 0.769650655021834,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004949288041113799,
      "loss": 5.3837,
      "mean_token_accuracy": 0.15388258174061775,
      "num_tokens": 16994639.0,
      "step": 7755
    },
    {
      "entropy": 5.537297868728638,
      "epoch": 0.7701468836840016,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004949213212841299,
      "loss": 5.3737,
      "mean_token_accuracy": 0.15502209067344666,
      "num_tokens": 17005067.0,
      "step": 7760
    },
    {
      "entropy": 5.5577222347259525,
      "epoch": 0.7706431123461691,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494913833003305,
      "loss": 5.2941,
      "mean_token_accuracy": 0.16221049129962922,
      "num_tokens": 17015102.0,
      "step": 7765
    },
    {
      "entropy": 5.575913810729981,
      "epoch": 0.7711393410083366,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000494906339269091,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1595125064253807,
      "num_tokens": 17026290.0,
      "step": 7770
    },
    {
      "entropy": 5.559361839294434,
      "epoch": 0.7716355696705042,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004948988400816736,
      "loss": 5.2868,
      "mean_token_accuracy": 0.1568698450922966,
      "num_tokens": 17035445.0,
      "step": 7775
    },
    {
      "entropy": 5.592072057723999,
      "epoch": 0.7721317983326716,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004948913354412388,
      "loss": 5.3964,
      "mean_token_accuracy": 0.16023921817541123,
      "num_tokens": 17046515.0,
      "step": 7780
    },
    {
      "entropy": 5.585109519958496,
      "epoch": 0.7726280269948392,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948838253479727,
      "loss": 5.3672,
      "mean_token_accuracy": 0.15563152730464935,
      "num_tokens": 17056988.0,
      "step": 7785
    },
    {
      "entropy": 5.626164293289184,
      "epoch": 0.7731242556570067,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004948763098020616,
      "loss": 5.438,
      "mean_token_accuracy": 0.1468819335103035,
      "num_tokens": 17068304.0,
      "step": 7790
    },
    {
      "entropy": 5.571609878540039,
      "epoch": 0.7736204843191743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948687888036918,
      "loss": 5.3519,
      "mean_token_accuracy": 0.15504468083381653,
      "num_tokens": 17078405.0,
      "step": 7795
    },
    {
      "entropy": 5.5191480159759525,
      "epoch": 0.7741167129813418,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004948612623530497,
      "loss": 5.244,
      "mean_token_accuracy": 0.16183137595653535,
      "num_tokens": 17087839.0,
      "step": 7800
    },
    {
      "entropy": 5.489332771301269,
      "epoch": 0.7746129416435094,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004948537304503221,
      "loss": 5.343,
      "mean_token_accuracy": 0.15664906352758406,
      "num_tokens": 17099386.0,
      "step": 7805
    },
    {
      "entropy": 5.6303938865661625,
      "epoch": 0.7751091703056768,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004948461930956959,
      "loss": 5.4147,
      "mean_token_accuracy": 0.14881821423768998,
      "num_tokens": 17111754.0,
      "step": 7810
    },
    {
      "entropy": 5.577929067611694,
      "epoch": 0.7756053989678444,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004948386502893577,
      "loss": 5.2901,
      "mean_token_accuracy": 0.16308683380484582,
      "num_tokens": 17122882.0,
      "step": 7815
    },
    {
      "entropy": 5.571791076660157,
      "epoch": 0.7761016276300119,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004948311020314946,
      "loss": 5.4017,
      "mean_token_accuracy": 0.1544287048280239,
      "num_tokens": 17133986.0,
      "step": 7820
    },
    {
      "entropy": 5.564396953582763,
      "epoch": 0.7765978562921795,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948235483222941,
      "loss": 5.3703,
      "mean_token_accuracy": 0.1608246684074402,
      "num_tokens": 17144784.0,
      "step": 7825
    },
    {
      "entropy": 5.612355756759643,
      "epoch": 0.777094084954347,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004948159891619431,
      "loss": 5.4848,
      "mean_token_accuracy": 0.14589942544698714,
      "num_tokens": 17156328.0,
      "step": 7830
    },
    {
      "entropy": 5.667614603042603,
      "epoch": 0.7775903136165145,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004948084245506293,
      "loss": 5.2969,
      "mean_token_accuracy": 0.15389322191476823,
      "num_tokens": 17168062.0,
      "step": 7835
    },
    {
      "entropy": 5.5918810844421385,
      "epoch": 0.778086542278682,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00049480085448854,
      "loss": 5.3339,
      "mean_token_accuracy": 0.15638585090637208,
      "num_tokens": 17178872.0,
      "step": 7840
    },
    {
      "entropy": 5.580067920684814,
      "epoch": 0.7785827709408495,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004947932789758635,
      "loss": 5.3759,
      "mean_token_accuracy": 0.15645520389080048,
      "num_tokens": 17189445.0,
      "step": 7845
    },
    {
      "entropy": 5.556174278259277,
      "epoch": 0.7790789996030171,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494785698012787,
      "loss": 5.3266,
      "mean_token_accuracy": 0.15869880020618438,
      "num_tokens": 17199454.0,
      "step": 7850
    },
    {
      "entropy": 5.584144449234008,
      "epoch": 0.7795752282651845,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947781115994989,
      "loss": 5.3905,
      "mean_token_accuracy": 0.15125236362218858,
      "num_tokens": 17210729.0,
      "step": 7855
    },
    {
      "entropy": 5.5718365669250485,
      "epoch": 0.7800714569273521,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004947705197361872,
      "loss": 5.4458,
      "mean_token_accuracy": 0.15393838435411453,
      "num_tokens": 17222961.0,
      "step": 7860
    },
    {
      "entropy": 5.5568482875823975,
      "epoch": 0.7805676855895196,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004947629224230401,
      "loss": 5.3284,
      "mean_token_accuracy": 0.16163438260555268,
      "num_tokens": 17234026.0,
      "step": 7865
    },
    {
      "entropy": 5.679432725906372,
      "epoch": 0.7810639142516872,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004947553196602461,
      "loss": 5.46,
      "mean_token_accuracy": 0.15723193436861038,
      "num_tokens": 17244643.0,
      "step": 7870
    },
    {
      "entropy": 5.646706056594849,
      "epoch": 0.7815601429138547,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004947477114479939,
      "loss": 5.389,
      "mean_token_accuracy": 0.15867964625358583,
      "num_tokens": 17255382.0,
      "step": 7875
    },
    {
      "entropy": 5.51731162071228,
      "epoch": 0.7820563715760223,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947400977864718,
      "loss": 5.3553,
      "mean_token_accuracy": 0.14919647872447966,
      "num_tokens": 17265860.0,
      "step": 7880
    },
    {
      "entropy": 5.615230989456177,
      "epoch": 0.7825526002381897,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947324786758688,
      "loss": 5.4431,
      "mean_token_accuracy": 0.15230426266789437,
      "num_tokens": 17277570.0,
      "step": 7885
    },
    {
      "entropy": 5.705474138259888,
      "epoch": 0.7830488289003573,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004947248541163738,
      "loss": 5.4898,
      "mean_token_accuracy": 0.14861185774207114,
      "num_tokens": 17288684.0,
      "step": 7890
    },
    {
      "entropy": 5.589676380157471,
      "epoch": 0.7835450575625248,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004947172241081759,
      "loss": 5.3253,
      "mean_token_accuracy": 0.15708291828632354,
      "num_tokens": 17299541.0,
      "step": 7895
    },
    {
      "entropy": 5.602476644515991,
      "epoch": 0.7840412862246924,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004947095886514643,
      "loss": 5.2783,
      "mean_token_accuracy": 0.15970128178596496,
      "num_tokens": 17310243.0,
      "step": 7900
    },
    {
      "entropy": 5.5937763214111325,
      "epoch": 0.7845375148868599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947019477464283,
      "loss": 5.3952,
      "mean_token_accuracy": 0.1548526644706726,
      "num_tokens": 17321167.0,
      "step": 7905
    },
    {
      "entropy": 5.506478023529053,
      "epoch": 0.7850337435490274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004946943013932574,
      "loss": 5.3667,
      "mean_token_accuracy": 0.1570984087884426,
      "num_tokens": 17331076.0,
      "step": 7910
    },
    {
      "entropy": 5.604095458984375,
      "epoch": 0.7855299722111949,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946866495921413,
      "loss": 5.3483,
      "mean_token_accuracy": 0.15030571520328523,
      "num_tokens": 17341337.0,
      "step": 7915
    },
    {
      "entropy": 5.624768924713135,
      "epoch": 0.7860262008733624,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004946789923432695,
      "loss": 5.3732,
      "mean_token_accuracy": 0.15634751468896865,
      "num_tokens": 17353055.0,
      "step": 7920
    },
    {
      "entropy": 5.501380872726441,
      "epoch": 0.78652242953553,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004946713296468322,
      "loss": 5.2738,
      "mean_token_accuracy": 0.15472376495599746,
      "num_tokens": 17364442.0,
      "step": 7925
    },
    {
      "entropy": 5.590744161605835,
      "epoch": 0.7870186581976975,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946636615030192,
      "loss": 5.3917,
      "mean_token_accuracy": 0.15338332653045655,
      "num_tokens": 17376237.0,
      "step": 7930
    },
    {
      "entropy": 5.614355754852295,
      "epoch": 0.787514886859865,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004946559879120208,
      "loss": 5.3279,
      "mean_token_accuracy": 0.1570687085390091,
      "num_tokens": 17386188.0,
      "step": 7935
    },
    {
      "entropy": 5.489083671569825,
      "epoch": 0.7880111155220325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004946483088740273,
      "loss": 5.3173,
      "mean_token_accuracy": 0.1574299693107605,
      "num_tokens": 17397058.0,
      "step": 7940
    },
    {
      "entropy": 5.623507499694824,
      "epoch": 0.7885073441842001,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946406243892289,
      "loss": 5.3819,
      "mean_token_accuracy": 0.15835516899824142,
      "num_tokens": 17408191.0,
      "step": 7945
    },
    {
      "entropy": 5.591425371170044,
      "epoch": 0.7890035728463676,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004946329344578163,
      "loss": 5.3367,
      "mean_token_accuracy": 0.15677808672189714,
      "num_tokens": 17418472.0,
      "step": 7950
    },
    {
      "entropy": 5.571155643463134,
      "epoch": 0.7894998015085352,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004946252390799803,
      "loss": 5.27,
      "mean_token_accuracy": 0.15681035667657853,
      "num_tokens": 17429556.0,
      "step": 7955
    },
    {
      "entropy": 5.59141526222229,
      "epoch": 0.7899960301707026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004946175382559116,
      "loss": 5.4496,
      "mean_token_accuracy": 0.14914974793791771,
      "num_tokens": 17440950.0,
      "step": 7960
    },
    {
      "entropy": 5.533390617370605,
      "epoch": 0.7904922588328702,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000494609831985801,
      "loss": 5.3296,
      "mean_token_accuracy": 0.16047880202531814,
      "num_tokens": 17453724.0,
      "step": 7965
    },
    {
      "entropy": 5.605449867248535,
      "epoch": 0.7909884874950377,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049460212026984,
      "loss": 5.3441,
      "mean_token_accuracy": 0.15683451890945435,
      "num_tokens": 17464623.0,
      "step": 7970
    },
    {
      "entropy": 5.566505718231201,
      "epoch": 0.7914847161572053,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004945944031082195,
      "loss": 5.3667,
      "mean_token_accuracy": 0.15867237150669097,
      "num_tokens": 17474919.0,
      "step": 7975
    },
    {
      "entropy": 5.62948169708252,
      "epoch": 0.7919809448193728,
      "grad_norm": 1.0,
      "learning_rate": 0.000494586680501131,
      "loss": 5.3576,
      "mean_token_accuracy": 0.15309137850999832,
      "num_tokens": 17484747.0,
      "step": 7980
    },
    {
      "entropy": 5.563917779922486,
      "epoch": 0.7924771734815403,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000494578952448766,
      "loss": 5.2889,
      "mean_token_accuracy": 0.15763356536626816,
      "num_tokens": 17496299.0,
      "step": 7985
    },
    {
      "entropy": 5.605212545394897,
      "epoch": 0.7929734021437078,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004945712189513161,
      "loss": 5.4543,
      "mean_token_accuracy": 0.14569606631994247,
      "num_tokens": 17507448.0,
      "step": 7990
    },
    {
      "entropy": 5.649667549133301,
      "epoch": 0.7934696308058753,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004945634800089733,
      "loss": 5.4168,
      "mean_token_accuracy": 0.15070326775312423,
      "num_tokens": 17520112.0,
      "step": 7995
    },
    {
      "entropy": 5.642276573181152,
      "epoch": 0.7939658594680429,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004945557356219291,
      "loss": 5.4167,
      "mean_token_accuracy": 0.14850203543901444,
      "num_tokens": 17531368.0,
      "step": 8000
    },
    {
      "entropy": 5.533618688583374,
      "epoch": 0.7944620881302104,
      "grad_norm": 0.875,
      "learning_rate": 0.0004945479857903758,
      "loss": 5.3558,
      "mean_token_accuracy": 0.15997211635112762,
      "num_tokens": 17542082.0,
      "step": 8005
    },
    {
      "entropy": 5.615306949615478,
      "epoch": 0.7949583167923779,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004945402305145056,
      "loss": 5.3872,
      "mean_token_accuracy": 0.15535500198602675,
      "num_tokens": 17555518.0,
      "step": 8010
    },
    {
      "entropy": 5.610823249816894,
      "epoch": 0.7954545454545454,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004945324697945107,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1580477848649025,
      "num_tokens": 17566110.0,
      "step": 8015
    },
    {
      "entropy": 5.510255002975464,
      "epoch": 0.795950774116713,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004945247036305836,
      "loss": 5.2753,
      "mean_token_accuracy": 0.16026078015565873,
      "num_tokens": 17575942.0,
      "step": 8020
    },
    {
      "entropy": 5.569670629501343,
      "epoch": 0.7964470027788805,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004945169320229169,
      "loss": 5.3462,
      "mean_token_accuracy": 0.1620591923594475,
      "num_tokens": 17586001.0,
      "step": 8025
    },
    {
      "entropy": 5.664316272735595,
      "epoch": 0.7969432314410481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945091549717033,
      "loss": 5.5056,
      "mean_token_accuracy": 0.14150234535336495,
      "num_tokens": 17596774.0,
      "step": 8030
    },
    {
      "entropy": 5.606074237823487,
      "epoch": 0.7974394601032155,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004945013724771356,
      "loss": 5.359,
      "mean_token_accuracy": 0.15726364254951478,
      "num_tokens": 17606557.0,
      "step": 8035
    },
    {
      "entropy": 5.511505603790283,
      "epoch": 0.7979356887653831,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004944935845394069,
      "loss": 5.2355,
      "mean_token_accuracy": 0.1616591840982437,
      "num_tokens": 17616465.0,
      "step": 8040
    },
    {
      "entropy": 5.617644643783569,
      "epoch": 0.7984319174275506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944857911587103,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1548914223909378,
      "num_tokens": 17626907.0,
      "step": 8045
    },
    {
      "entropy": 5.68097767829895,
      "epoch": 0.7989281460897182,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004944779923352389,
      "loss": 5.4032,
      "mean_token_accuracy": 0.1589493051171303,
      "num_tokens": 17638660.0,
      "step": 8050
    },
    {
      "entropy": 5.520690774917602,
      "epoch": 0.7994243747518857,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004944701880691864,
      "loss": 5.3229,
      "mean_token_accuracy": 0.15878796726465225,
      "num_tokens": 17650135.0,
      "step": 8055
    },
    {
      "entropy": 5.497169923782349,
      "epoch": 0.7999206034140532,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000494462378360746,
      "loss": 5.3524,
      "mean_token_accuracy": 0.1597804456949234,
      "num_tokens": 17660987.0,
      "step": 8060
    },
    {
      "entropy": 5.603233718872071,
      "epoch": 0.8004168320762207,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004944545632101117,
      "loss": 5.2721,
      "mean_token_accuracy": 0.16170316338539123,
      "num_tokens": 17672520.0,
      "step": 8065
    },
    {
      "entropy": 5.594707727432251,
      "epoch": 0.8009130607383882,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944467426174769,
      "loss": 5.3454,
      "mean_token_accuracy": 0.15994005650281906,
      "num_tokens": 17683688.0,
      "step": 8070
    },
    {
      "entropy": 5.520062732696533,
      "epoch": 0.8014092894005558,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494438916583036,
      "loss": 5.2723,
      "mean_token_accuracy": 0.16427962034940718,
      "num_tokens": 17695213.0,
      "step": 8075
    },
    {
      "entropy": 5.598856496810913,
      "epoch": 0.8019055180627233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004944310851069826,
      "loss": 5.3621,
      "mean_token_accuracy": 0.15491142719984055,
      "num_tokens": 17705545.0,
      "step": 8080
    },
    {
      "entropy": 5.572071647644043,
      "epoch": 0.8024017467248908,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004944232481895112,
      "loss": 5.3341,
      "mean_token_accuracy": 0.1653673991560936,
      "num_tokens": 17715855.0,
      "step": 8085
    },
    {
      "entropy": 5.645730257034302,
      "epoch": 0.8028979753870583,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000494415405830816,
      "loss": 5.4218,
      "mean_token_accuracy": 0.1588004395365715,
      "num_tokens": 17727764.0,
      "step": 8090
    },
    {
      "entropy": 5.595269632339478,
      "epoch": 0.8033942040492259,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004944075580310916,
      "loss": 5.4066,
      "mean_token_accuracy": 0.15369394719600676,
      "num_tokens": 17738464.0,
      "step": 8095
    },
    {
      "entropy": 5.558055639266968,
      "epoch": 0.8038904327113934,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004943997047905325,
      "loss": 5.3878,
      "mean_token_accuracy": 0.16175874471664428,
      "num_tokens": 17747502.0,
      "step": 8100
    },
    {
      "entropy": 5.543888235092163,
      "epoch": 0.804386661373561,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004943918461093335,
      "loss": 5.3816,
      "mean_token_accuracy": 0.15127424150705338,
      "num_tokens": 17759520.0,
      "step": 8105
    },
    {
      "entropy": 5.583495807647705,
      "epoch": 0.8048828900357284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943839819876894,
      "loss": 5.3465,
      "mean_token_accuracy": 0.15715339332818984,
      "num_tokens": 17770631.0,
      "step": 8110
    },
    {
      "entropy": 5.643279409408569,
      "epoch": 0.805379118697896,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943761124257954,
      "loss": 5.3202,
      "mean_token_accuracy": 0.16016861647367478,
      "num_tokens": 17781320.0,
      "step": 8115
    },
    {
      "entropy": 5.60970106124878,
      "epoch": 0.8058753473600635,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004943682374238464,
      "loss": 5.3697,
      "mean_token_accuracy": 0.15739177018404008,
      "num_tokens": 17792627.0,
      "step": 8120
    },
    {
      "entropy": 5.552402353286743,
      "epoch": 0.8063715760222311,
      "grad_norm": 0.84375,
      "learning_rate": 0.000494360356982038,
      "loss": 5.4146,
      "mean_token_accuracy": 0.15475883781909944,
      "num_tokens": 17803286.0,
      "step": 8125
    },
    {
      "entropy": 5.516264533996582,
      "epoch": 0.8068678046843986,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004943524711005653,
      "loss": 5.2066,
      "mean_token_accuracy": 0.16201590597629548,
      "num_tokens": 17813631.0,
      "step": 8130
    },
    {
      "entropy": 5.545795249938965,
      "epoch": 0.8073640333465661,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000494344579779624,
      "loss": 5.3213,
      "mean_token_accuracy": 0.16185048073530198,
      "num_tokens": 17823467.0,
      "step": 8135
    },
    {
      "entropy": 5.565887689590454,
      "epoch": 0.8078602620087336,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004943366830194098,
      "loss": 5.3243,
      "mean_token_accuracy": 0.16089937090873718,
      "num_tokens": 17833600.0,
      "step": 8140
    },
    {
      "entropy": 5.5894568920135494,
      "epoch": 0.8083564906709012,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004943287808201183,
      "loss": 5.3068,
      "mean_token_accuracy": 0.15541218370199203,
      "num_tokens": 17843112.0,
      "step": 8145
    },
    {
      "entropy": 5.582536363601685,
      "epoch": 0.8088527193330687,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004943208731819459,
      "loss": 5.3203,
      "mean_token_accuracy": 0.1597540035843849,
      "num_tokens": 17853678.0,
      "step": 8150
    },
    {
      "entropy": 5.555966567993164,
      "epoch": 0.8093489479952362,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004943129601050883,
      "loss": 5.3008,
      "mean_token_accuracy": 0.16138895601034164,
      "num_tokens": 17863537.0,
      "step": 8155
    },
    {
      "entropy": 5.573474502563476,
      "epoch": 0.8098451766574037,
      "grad_norm": 0.9375,
      "learning_rate": 0.000494305041589742,
      "loss": 5.3833,
      "mean_token_accuracy": 0.15460158437490462,
      "num_tokens": 17873511.0,
      "step": 8160
    },
    {
      "entropy": 5.548963499069214,
      "epoch": 0.8103414053195712,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942971176361031,
      "loss": 5.3039,
      "mean_token_accuracy": 0.16254584938287736,
      "num_tokens": 17884546.0,
      "step": 8165
    },
    {
      "entropy": 5.539442014694214,
      "epoch": 0.8108376339817388,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942891882443682,
      "loss": 5.267,
      "mean_token_accuracy": 0.15955083966255187,
      "num_tokens": 17895331.0,
      "step": 8170
    },
    {
      "entropy": 5.592162084579468,
      "epoch": 0.8113338626439063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494281253414734,
      "loss": 5.4307,
      "mean_token_accuracy": 0.15022534355521203,
      "num_tokens": 17906465.0,
      "step": 8175
    },
    {
      "entropy": 5.565536785125732,
      "epoch": 0.8118300913060739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004942733131473973,
      "loss": 5.348,
      "mean_token_accuracy": 0.1506285160779953,
      "num_tokens": 17916936.0,
      "step": 8180
    },
    {
      "entropy": 5.667322969436645,
      "epoch": 0.8123263199682413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004942653674425548,
      "loss": 5.3734,
      "mean_token_accuracy": 0.15105995535850525,
      "num_tokens": 17927522.0,
      "step": 8185
    },
    {
      "entropy": 5.608132314682007,
      "epoch": 0.8128225486304089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004942574163004038,
      "loss": 5.4178,
      "mean_token_accuracy": 0.15179194435477256,
      "num_tokens": 17938216.0,
      "step": 8190
    },
    {
      "entropy": 5.559127855300903,
      "epoch": 0.8133187772925764,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004942494597211412,
      "loss": 5.3114,
      "mean_token_accuracy": 0.15462345629930496,
      "num_tokens": 17949333.0,
      "step": 8195
    },
    {
      "entropy": 5.631190156936645,
      "epoch": 0.813815005954744,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004942414977049646,
      "loss": 5.3916,
      "mean_token_accuracy": 0.14411346316337587,
      "num_tokens": 17960178.0,
      "step": 8200
    },
    {
      "entropy": 5.571792364120483,
      "epoch": 0.8143112346169115,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000494233530252071,
      "loss": 5.3481,
      "mean_token_accuracy": 0.15745973587036133,
      "num_tokens": 17971180.0,
      "step": 8205
    },
    {
      "entropy": 5.599970531463623,
      "epoch": 0.814807463279079,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004942255573626585,
      "loss": 5.3791,
      "mean_token_accuracy": 0.15721147805452346,
      "num_tokens": 17982572.0,
      "step": 8210
    },
    {
      "entropy": 5.606617450714111,
      "epoch": 0.8153036919412465,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004942175790369244,
      "loss": 5.3045,
      "mean_token_accuracy": 0.15940630584955215,
      "num_tokens": 17993236.0,
      "step": 8215
    },
    {
      "entropy": 5.589914321899414,
      "epoch": 0.8157999206034141,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004942095952750669,
      "loss": 5.4241,
      "mean_token_accuracy": 0.15796806365251542,
      "num_tokens": 18004215.0,
      "step": 8220
    },
    {
      "entropy": 5.554336929321289,
      "epoch": 0.8162961492655816,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004942016060772837,
      "loss": 5.3242,
      "mean_token_accuracy": 0.16181332021951675,
      "num_tokens": 18013880.0,
      "step": 8225
    },
    {
      "entropy": 5.591228342056274,
      "epoch": 0.8167923779277491,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941936114437731,
      "loss": 5.3776,
      "mean_token_accuracy": 0.15427337735891342,
      "num_tokens": 18024928.0,
      "step": 8230
    },
    {
      "entropy": 5.581340503692627,
      "epoch": 0.8172886065899166,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941856113747331,
      "loss": 5.3282,
      "mean_token_accuracy": 0.15814435333013535,
      "num_tokens": 18035857.0,
      "step": 8235
    },
    {
      "entropy": 5.650403165817261,
      "epoch": 0.8177848352520841,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004941776058703623,
      "loss": 5.4436,
      "mean_token_accuracy": 0.15249346941709518,
      "num_tokens": 18046776.0,
      "step": 8240
    },
    {
      "entropy": 5.625113487243652,
      "epoch": 0.8182810639142517,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941695949308593,
      "loss": 5.3939,
      "mean_token_accuracy": 0.16092305779457092,
      "num_tokens": 18057167.0,
      "step": 8245
    },
    {
      "entropy": 5.615349817276001,
      "epoch": 0.8187772925764192,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941615785564224,
      "loss": 5.4091,
      "mean_token_accuracy": 0.15176156014204026,
      "num_tokens": 18066953.0,
      "step": 8250
    },
    {
      "entropy": 5.566922760009765,
      "epoch": 0.8192735212385868,
      "grad_norm": 0.875,
      "learning_rate": 0.0004941535567472506,
      "loss": 5.2438,
      "mean_token_accuracy": 0.15974170714616776,
      "num_tokens": 18077977.0,
      "step": 8255
    },
    {
      "entropy": 5.562719917297363,
      "epoch": 0.8197697499007542,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000494145529503543,
      "loss": 5.3339,
      "mean_token_accuracy": 0.15678948760032654,
      "num_tokens": 18087581.0,
      "step": 8260
    },
    {
      "entropy": 5.504536581039429,
      "epoch": 0.8202659785629218,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004941374968254985,
      "loss": 5.2478,
      "mean_token_accuracy": 0.1605113834142685,
      "num_tokens": 18098193.0,
      "step": 8265
    },
    {
      "entropy": 5.580500316619873,
      "epoch": 0.8207622072250893,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004941294587133161,
      "loss": 5.3732,
      "mean_token_accuracy": 0.16001071482896806,
      "num_tokens": 18110830.0,
      "step": 8270
    },
    {
      "entropy": 5.645088148117066,
      "epoch": 0.8212584358872569,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004941214151671954,
      "loss": 5.4013,
      "mean_token_accuracy": 0.15151395350694657,
      "num_tokens": 18122073.0,
      "step": 8275
    },
    {
      "entropy": 5.554434156417846,
      "epoch": 0.8217546645494244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004941133661873357,
      "loss": 5.363,
      "mean_token_accuracy": 0.15824461579322815,
      "num_tokens": 18132777.0,
      "step": 8280
    },
    {
      "entropy": 5.504657840728759,
      "epoch": 0.822250893211592,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004941053117739367,
      "loss": 5.2823,
      "mean_token_accuracy": 0.1562109962105751,
      "num_tokens": 18143232.0,
      "step": 8285
    },
    {
      "entropy": 5.594614744186401,
      "epoch": 0.8227471218737594,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004940972519271982,
      "loss": 5.4624,
      "mean_token_accuracy": 0.15384506583213806,
      "num_tokens": 18155225.0,
      "step": 8290
    },
    {
      "entropy": 5.612090682983398,
      "epoch": 0.823243350535927,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004940891866473199,
      "loss": 5.379,
      "mean_token_accuracy": 0.15907496511936187,
      "num_tokens": 18165945.0,
      "step": 8295
    },
    {
      "entropy": 5.647741031646729,
      "epoch": 0.8237395791980945,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004940811159345017,
      "loss": 5.4226,
      "mean_token_accuracy": 0.15485143512487412,
      "num_tokens": 18177271.0,
      "step": 8300
    },
    {
      "entropy": 5.642366933822632,
      "epoch": 0.824235807860262,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004940730397889441,
      "loss": 5.3661,
      "mean_token_accuracy": 0.15563051551580429,
      "num_tokens": 18189732.0,
      "step": 8305
    },
    {
      "entropy": 5.568066501617432,
      "epoch": 0.8247320365224295,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004940649582108472,
      "loss": 5.4571,
      "mean_token_accuracy": 0.14987728670239447,
      "num_tokens": 18200731.0,
      "step": 8310
    },
    {
      "entropy": 5.580982685089111,
      "epoch": 0.825228265184597,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004940568712004113,
      "loss": 5.3749,
      "mean_token_accuracy": 0.15586638450622559,
      "num_tokens": 18210873.0,
      "step": 8315
    },
    {
      "entropy": 5.6476280212402346,
      "epoch": 0.8257244938467646,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004940487787578369,
      "loss": 5.3635,
      "mean_token_accuracy": 0.15814482420682907,
      "num_tokens": 18222072.0,
      "step": 8320
    },
    {
      "entropy": 5.5344644546508786,
      "epoch": 0.8262207225089321,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494040680883325,
      "loss": 5.2809,
      "mean_token_accuracy": 0.1563542440533638,
      "num_tokens": 18232522.0,
      "step": 8325
    },
    {
      "entropy": 5.617687654495239,
      "epoch": 0.8267169511710997,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004940325775770762,
      "loss": 5.3801,
      "mean_token_accuracy": 0.15220721513032914,
      "num_tokens": 18243587.0,
      "step": 8330
    },
    {
      "entropy": 5.539453601837158,
      "epoch": 0.8272131798332671,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004940244688392913,
      "loss": 5.2798,
      "mean_token_accuracy": 0.16356151700019836,
      "num_tokens": 18254965.0,
      "step": 8335
    },
    {
      "entropy": 5.58465542793274,
      "epoch": 0.8277094084954347,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004940163546701716,
      "loss": 5.307,
      "mean_token_accuracy": 0.16236516535282136,
      "num_tokens": 18265892.0,
      "step": 8340
    },
    {
      "entropy": 5.576088953018188,
      "epoch": 0.8282056371576022,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004940082350699183,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1569358304142952,
      "num_tokens": 18276713.0,
      "step": 8345
    },
    {
      "entropy": 5.623620939254761,
      "epoch": 0.8287018658197698,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004940001100387325,
      "loss": 5.3682,
      "mean_token_accuracy": 0.15092188417911528,
      "num_tokens": 18288846.0,
      "step": 8350
    },
    {
      "entropy": 5.646070718765259,
      "epoch": 0.8291980944819373,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004939919795768161,
      "loss": 5.3758,
      "mean_token_accuracy": 0.16379236280918122,
      "num_tokens": 18298750.0,
      "step": 8355
    },
    {
      "entropy": 5.595037221908569,
      "epoch": 0.8296943231441049,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004939838436843704,
      "loss": 5.3585,
      "mean_token_accuracy": 0.15630945563316345,
      "num_tokens": 18310065.0,
      "step": 8360
    },
    {
      "entropy": 5.555487585067749,
      "epoch": 0.8301905518062723,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004939757023615973,
      "loss": 5.3111,
      "mean_token_accuracy": 0.15904302895069122,
      "num_tokens": 18319498.0,
      "step": 8365
    },
    {
      "entropy": 5.545468187332153,
      "epoch": 0.8306867804684399,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004939675556086986,
      "loss": 5.2788,
      "mean_token_accuracy": 0.16292627900838852,
      "num_tokens": 18329718.0,
      "step": 8370
    },
    {
      "entropy": 5.5360006332397464,
      "epoch": 0.8311830091306074,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004939594034258763,
      "loss": 5.2897,
      "mean_token_accuracy": 0.16169728934764863,
      "num_tokens": 18341628.0,
      "step": 8375
    },
    {
      "entropy": 5.558377981185913,
      "epoch": 0.8316792377927749,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004939512458133326,
      "loss": 5.3324,
      "mean_token_accuracy": 0.16465890407562256,
      "num_tokens": 18352869.0,
      "step": 8380
    },
    {
      "entropy": 5.585212898254395,
      "epoch": 0.8321754664549424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004939430827712698,
      "loss": 5.339,
      "mean_token_accuracy": 0.15363990515470505,
      "num_tokens": 18363664.0,
      "step": 8385
    },
    {
      "entropy": 5.619423341751099,
      "epoch": 0.8326716951171099,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004939349142998905,
      "loss": 5.32,
      "mean_token_accuracy": 0.15818386524915695,
      "num_tokens": 18374325.0,
      "step": 8390
    },
    {
      "entropy": 5.555478620529175,
      "epoch": 0.8331679237792775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939267403993968,
      "loss": 5.3044,
      "mean_token_accuracy": 0.16002315282821655,
      "num_tokens": 18384547.0,
      "step": 8395
    },
    {
      "entropy": 5.577031755447388,
      "epoch": 0.833664152441445,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000493918561069992,
      "loss": 5.2935,
      "mean_token_accuracy": 0.15935265123844147,
      "num_tokens": 18395141.0,
      "step": 8400
    },
    {
      "entropy": 5.555101299285889,
      "epoch": 0.8341603811036126,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004939103763118784,
      "loss": 5.3061,
      "mean_token_accuracy": 0.15835533589124678,
      "num_tokens": 18406326.0,
      "step": 8405
    },
    {
      "entropy": 5.587814140319824,
      "epoch": 0.83465660976578,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004939021861252593,
      "loss": 5.3491,
      "mean_token_accuracy": 0.16088419854640962,
      "num_tokens": 18417521.0,
      "step": 8410
    },
    {
      "entropy": 5.590854692459106,
      "epoch": 0.8351528384279476,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938939905103376,
      "loss": 5.298,
      "mean_token_accuracy": 0.1615044191479683,
      "num_tokens": 18426813.0,
      "step": 8415
    },
    {
      "entropy": 5.529473209381104,
      "epoch": 0.8356490670901151,
      "grad_norm": 0.875,
      "learning_rate": 0.0004938857894673166,
      "loss": 5.2566,
      "mean_token_accuracy": 0.1579472079873085,
      "num_tokens": 18438227.0,
      "step": 8420
    },
    {
      "entropy": 5.565154266357422,
      "epoch": 0.8361452957522827,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004938775829963997,
      "loss": 5.3825,
      "mean_token_accuracy": 0.15830719470977783,
      "num_tokens": 18448915.0,
      "step": 8425
    },
    {
      "entropy": 5.5853485584259035,
      "epoch": 0.8366415244144502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938693710977904,
      "loss": 5.2703,
      "mean_token_accuracy": 0.15719448775053024,
      "num_tokens": 18459330.0,
      "step": 8430
    },
    {
      "entropy": 5.550517416000366,
      "epoch": 0.8371377530766178,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004938611537716922,
      "loss": 5.2543,
      "mean_token_accuracy": 0.16147958487272263,
      "num_tokens": 18469813.0,
      "step": 8435
    },
    {
      "entropy": 5.512485408782959,
      "epoch": 0.8376339817387852,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004938529310183091,
      "loss": 5.3146,
      "mean_token_accuracy": 0.15576055645942688,
      "num_tokens": 18479783.0,
      "step": 8440
    },
    {
      "entropy": 5.551927137374878,
      "epoch": 0.8381302104009528,
      "grad_norm": 0.90625,
      "learning_rate": 0.000493844702837845,
      "loss": 5.3175,
      "mean_token_accuracy": 0.1590595155954361,
      "num_tokens": 18489640.0,
      "step": 8445
    },
    {
      "entropy": 5.603930139541626,
      "epoch": 0.8386264390631203,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004938364692305036,
      "loss": 5.3815,
      "mean_token_accuracy": 0.16340595334768296,
      "num_tokens": 18500744.0,
      "step": 8450
    },
    {
      "entropy": 5.572290277481079,
      "epoch": 0.8391226677252878,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004938282301964896,
      "loss": 5.2879,
      "mean_token_accuracy": 0.15781228393316268,
      "num_tokens": 18513286.0,
      "step": 8455
    },
    {
      "entropy": 5.62964940071106,
      "epoch": 0.8396188963874553,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004938199857360068,
      "loss": 5.2739,
      "mean_token_accuracy": 0.15904214680194856,
      "num_tokens": 18522934.0,
      "step": 8460
    },
    {
      "entropy": 5.545163297653199,
      "epoch": 0.8401151250496228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00049381173584926,
      "loss": 5.3349,
      "mean_token_accuracy": 0.16100945323705673,
      "num_tokens": 18534828.0,
      "step": 8465
    },
    {
      "entropy": 5.517857074737549,
      "epoch": 0.8406113537117904,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004938034805364535,
      "loss": 5.2749,
      "mean_token_accuracy": 0.16149184107780457,
      "num_tokens": 18544733.0,
      "step": 8470
    },
    {
      "entropy": 5.562247610092163,
      "epoch": 0.8411075823739579,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937952197977923,
      "loss": 5.3409,
      "mean_token_accuracy": 0.15260908007621765,
      "num_tokens": 18554583.0,
      "step": 8475
    },
    {
      "entropy": 5.671601915359497,
      "epoch": 0.8416038110361255,
      "grad_norm": 0.80859375,
      "learning_rate": 0.000493786953633481,
      "loss": 5.3843,
      "mean_token_accuracy": 0.15565856993198396,
      "num_tokens": 18566503.0,
      "step": 8480
    },
    {
      "entropy": 5.5688323974609375,
      "epoch": 0.842100039698293,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0004937786820437248,
      "loss": 5.3266,
      "mean_token_accuracy": 0.16158413887023926,
      "num_tokens": 18578161.0,
      "step": 8485
    },
    {
      "entropy": 5.5614372253417965,
      "epoch": 0.8425962683604605,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004937704050287286,
      "loss": 5.3796,
      "mean_token_accuracy": 0.1622001439332962,
      "num_tokens": 18588212.0,
      "step": 8490
    },
    {
      "entropy": 5.698554182052613,
      "epoch": 0.843092497022628,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004937621225886978,
      "loss": 5.4022,
      "mean_token_accuracy": 0.15143557637929916,
      "num_tokens": 18601293.0,
      "step": 8495
    },
    {
      "entropy": 5.591166830062866,
      "epoch": 0.8435887256847956,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937538347238378,
      "loss": 5.4323,
      "mean_token_accuracy": 0.15204148963093758,
      "num_tokens": 18613316.0,
      "step": 8500
    },
    {
      "entropy": 5.579216003417969,
      "epoch": 0.8440849543469631,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000493745541434354,
      "loss": 5.3585,
      "mean_token_accuracy": 0.15921425074338913,
      "num_tokens": 18624461.0,
      "step": 8505
    },
    {
      "entropy": 5.5923584461212155,
      "epoch": 0.8445811830091307,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004937372427204522,
      "loss": 5.2601,
      "mean_token_accuracy": 0.16073173880577088,
      "num_tokens": 18635165.0,
      "step": 8510
    },
    {
      "entropy": 5.598194313049317,
      "epoch": 0.8450774116712981,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004937289385823382,
      "loss": 5.3192,
      "mean_token_accuracy": 0.16395769491791726,
      "num_tokens": 18646649.0,
      "step": 8515
    },
    {
      "entropy": 5.5439825534820555,
      "epoch": 0.8455736403334657,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004937206290202176,
      "loss": 5.3101,
      "mean_token_accuracy": 0.15525393337011337,
      "num_tokens": 18658527.0,
      "step": 8520
    },
    {
      "entropy": 5.565108346939087,
      "epoch": 0.8460698689956332,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004937123140342969,
      "loss": 5.3045,
      "mean_token_accuracy": 0.15915210992097856,
      "num_tokens": 18669463.0,
      "step": 8525
    },
    {
      "entropy": 5.610577058792114,
      "epoch": 0.8465660976578007,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493703993624782,
      "loss": 5.3444,
      "mean_token_accuracy": 0.15823778808116912,
      "num_tokens": 18680806.0,
      "step": 8530
    },
    {
      "entropy": 5.58990159034729,
      "epoch": 0.8470623263199683,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004936956677918794,
      "loss": 5.3466,
      "mean_token_accuracy": 0.15270729959011078,
      "num_tokens": 18692124.0,
      "step": 8535
    },
    {
      "entropy": 5.6173028469085695,
      "epoch": 0.8475585549821357,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004936873365357955,
      "loss": 5.3726,
      "mean_token_accuracy": 0.15745879262685775,
      "num_tokens": 18702017.0,
      "step": 8540
    },
    {
      "entropy": 5.509934759140014,
      "epoch": 0.8480547836443033,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936789998567368,
      "loss": 5.2695,
      "mean_token_accuracy": 0.1639906033873558,
      "num_tokens": 18713205.0,
      "step": 8545
    },
    {
      "entropy": 5.571488094329834,
      "epoch": 0.8485510123064708,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004936706577549101,
      "loss": 5.2767,
      "mean_token_accuracy": 0.15899254828691484,
      "num_tokens": 18724165.0,
      "step": 8550
    },
    {
      "entropy": 5.550207042694092,
      "epoch": 0.8490472409686384,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004936623102305223,
      "loss": 5.3431,
      "mean_token_accuracy": 0.15628370493650437,
      "num_tokens": 18735155.0,
      "step": 8555
    },
    {
      "entropy": 5.584510278701782,
      "epoch": 0.8495434696308058,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936539572837804,
      "loss": 5.4349,
      "mean_token_accuracy": 0.1612741768360138,
      "num_tokens": 18746228.0,
      "step": 8560
    },
    {
      "entropy": 5.595359039306641,
      "epoch": 0.8500396982929734,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004936455989148916,
      "loss": 5.381,
      "mean_token_accuracy": 0.1569515660405159,
      "num_tokens": 18756638.0,
      "step": 8565
    },
    {
      "entropy": 5.631488227844239,
      "epoch": 0.8505359269551409,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004936372351240629,
      "loss": 5.3499,
      "mean_token_accuracy": 0.16094586849212647,
      "num_tokens": 18768651.0,
      "step": 8570
    },
    {
      "entropy": 5.663279104232788,
      "epoch": 0.8510321556173085,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000493628865911502,
      "loss": 5.4583,
      "mean_token_accuracy": 0.15369980186223983,
      "num_tokens": 18780399.0,
      "step": 8575
    },
    {
      "entropy": 5.5360759735107425,
      "epoch": 0.851528384279476,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936204912774163,
      "loss": 5.3079,
      "mean_token_accuracy": 0.1620068907737732,
      "num_tokens": 18791927.0,
      "step": 8580
    },
    {
      "entropy": 5.524154090881348,
      "epoch": 0.8520246129416436,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004936121112220135,
      "loss": 5.2886,
      "mean_token_accuracy": 0.15831017196178437,
      "num_tokens": 18802779.0,
      "step": 8585
    },
    {
      "entropy": 5.5297260761260985,
      "epoch": 0.852520841603811,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004936037257455015,
      "loss": 5.2918,
      "mean_token_accuracy": 0.15869465470314026,
      "num_tokens": 18813123.0,
      "step": 8590
    },
    {
      "entropy": 5.55844349861145,
      "epoch": 0.8530170702659786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000493595334848088,
      "loss": 5.2907,
      "mean_token_accuracy": 0.16447946578264236,
      "num_tokens": 18823699.0,
      "step": 8595
    },
    {
      "entropy": 5.609306526184082,
      "epoch": 0.8535132989281461,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004935869385299813,
      "loss": 5.4537,
      "mean_token_accuracy": 0.15290404260158538,
      "num_tokens": 18835450.0,
      "step": 8600
    },
    {
      "entropy": 5.5570789813995365,
      "epoch": 0.8540095275903136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004935785367913896,
      "loss": 5.2934,
      "mean_token_accuracy": 0.15879702270030976,
      "num_tokens": 18845919.0,
      "step": 8605
    },
    {
      "entropy": 5.637758445739746,
      "epoch": 0.8545057562524812,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004935701296325212,
      "loss": 5.4228,
      "mean_token_accuracy": 0.1513108007609844,
      "num_tokens": 18858376.0,
      "step": 8610
    },
    {
      "entropy": 5.597097539901734,
      "epoch": 0.8550019849146486,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0004935617170535845,
      "loss": 5.373,
      "mean_token_accuracy": 0.16063981056213378,
      "num_tokens": 18870456.0,
      "step": 8615
    },
    {
      "entropy": 5.575227403640747,
      "epoch": 0.8554982135768162,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004935532990547882,
      "loss": 5.3742,
      "mean_token_accuracy": 0.15272693634033202,
      "num_tokens": 18882174.0,
      "step": 8620
    },
    {
      "entropy": 5.673942804336548,
      "epoch": 0.8559944422389837,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004935448756363409,
      "loss": 5.4914,
      "mean_token_accuracy": 0.15665049701929093,
      "num_tokens": 18892982.0,
      "step": 8625
    },
    {
      "entropy": 5.618603801727295,
      "epoch": 0.8564906709011513,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935364467984518,
      "loss": 5.3199,
      "mean_token_accuracy": 0.15925368964672088,
      "num_tokens": 18903557.0,
      "step": 8630
    },
    {
      "entropy": 5.624365520477295,
      "epoch": 0.8569868995633187,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004935280125413295,
      "loss": 5.4592,
      "mean_token_accuracy": 0.1539561927318573,
      "num_tokens": 18914912.0,
      "step": 8635
    },
    {
      "entropy": 5.624571847915649,
      "epoch": 0.8574831282254863,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004935195728651836,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1531470611691475,
      "num_tokens": 18927773.0,
      "step": 8640
    },
    {
      "entropy": 5.600121927261353,
      "epoch": 0.8579793568876538,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493511127770223,
      "loss": 5.344,
      "mean_token_accuracy": 0.15515555888414384,
      "num_tokens": 18938313.0,
      "step": 8645
    },
    {
      "entropy": 5.558858919143677,
      "epoch": 0.8584755855498214,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004935026772566574,
      "loss": 5.2876,
      "mean_token_accuracy": 0.16656061112880707,
      "num_tokens": 18948600.0,
      "step": 8650
    },
    {
      "entropy": 5.626723718643189,
      "epoch": 0.8589718142119889,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004934942213246961,
      "loss": 5.3732,
      "mean_token_accuracy": 0.15782984644174575,
      "num_tokens": 18958927.0,
      "step": 8655
    },
    {
      "entropy": 5.555561113357544,
      "epoch": 0.8594680428741565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000493485759974549,
      "loss": 5.33,
      "mean_token_accuracy": 0.15172460824251174,
      "num_tokens": 18968625.0,
      "step": 8660
    },
    {
      "entropy": 5.480640649795532,
      "epoch": 0.8599642715363239,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004934772932064259,
      "loss": 5.2481,
      "mean_token_accuracy": 0.16098277866840363,
      "num_tokens": 18979772.0,
      "step": 8665
    },
    {
      "entropy": 5.589976263046265,
      "epoch": 0.8604605001984915,
      "grad_norm": 0.79296875,
      "learning_rate": 0.0004934688210205367,
      "loss": 5.4028,
      "mean_token_accuracy": 0.15752531737089157,
      "num_tokens": 18991522.0,
      "step": 8670
    },
    {
      "entropy": 5.617816972732544,
      "epoch": 0.860956728860659,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004934603434170913,
      "loss": 5.367,
      "mean_token_accuracy": 0.1583887070417404,
      "num_tokens": 19002479.0,
      "step": 8675
    },
    {
      "entropy": 5.609226655960083,
      "epoch": 0.8614529575228265,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004934518603963003,
      "loss": 5.4226,
      "mean_token_accuracy": 0.15003269016742707,
      "num_tokens": 19013254.0,
      "step": 8680
    },
    {
      "entropy": 5.5523600578308105,
      "epoch": 0.861949186184994,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004934433719583739,
      "loss": 5.2658,
      "mean_token_accuracy": 0.16132600456476212,
      "num_tokens": 19023943.0,
      "step": 8685
    },
    {
      "entropy": 5.6572980880737305,
      "epoch": 0.8624454148471615,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004934348781035226,
      "loss": 5.3806,
      "mean_token_accuracy": 0.15535594075918197,
      "num_tokens": 19034672.0,
      "step": 8690
    },
    {
      "entropy": 5.645968198776245,
      "epoch": 0.8629416435093291,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000493426378831957,
      "loss": 5.3737,
      "mean_token_accuracy": 0.16000091880559922,
      "num_tokens": 19045758.0,
      "step": 8695
    },
    {
      "entropy": 5.571767902374267,
      "epoch": 0.8634378721714966,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004934178741438877,
      "loss": 5.3639,
      "mean_token_accuracy": 0.1643066629767418,
      "num_tokens": 19057526.0,
      "step": 8700
    },
    {
      "entropy": 5.49440712928772,
      "epoch": 0.8639341008336642,
      "grad_norm": 0.828125,
      "learning_rate": 0.000493409364039526,
      "loss": 5.2286,
      "mean_token_accuracy": 0.16260730922222139,
      "num_tokens": 19069657.0,
      "step": 8705
    },
    {
      "entropy": 5.5707324028015135,
      "epoch": 0.8644303294958317,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004934008485190825,
      "loss": 5.2956,
      "mean_token_accuracy": 0.1632269486784935,
      "num_tokens": 19080699.0,
      "step": 8710
    },
    {
      "entropy": 5.5336998462677,
      "epoch": 0.8649265581579992,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004933923275827687,
      "loss": 5.334,
      "mean_token_accuracy": 0.15366750955581665,
      "num_tokens": 19093041.0,
      "step": 8715
    },
    {
      "entropy": 5.628512334823609,
      "epoch": 0.8654227868201667,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004933838012307957,
      "loss": 5.3894,
      "mean_token_accuracy": 0.1524179294705391,
      "num_tokens": 19104249.0,
      "step": 8720
    },
    {
      "entropy": 5.609440040588379,
      "epoch": 0.8659190154823343,
      "grad_norm": 1.0,
      "learning_rate": 0.000493375269463375,
      "loss": 5.3455,
      "mean_token_accuracy": 0.1559339940547943,
      "num_tokens": 19114434.0,
      "step": 8725
    },
    {
      "entropy": 5.5762366771698,
      "epoch": 0.8664152441445018,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933667322807182,
      "loss": 5.3589,
      "mean_token_accuracy": 0.15534116998314856,
      "num_tokens": 19124814.0,
      "step": 8730
    },
    {
      "entropy": 5.587624406814575,
      "epoch": 0.8669114728066694,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000493358189683037,
      "loss": 5.3837,
      "mean_token_accuracy": 0.1549019917845726,
      "num_tokens": 19135205.0,
      "step": 8735
    },
    {
      "entropy": 5.548503971099853,
      "epoch": 0.8674077014688368,
      "grad_norm": 0.7578125,
      "learning_rate": 0.0004933496416705432,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1608908250927925,
      "num_tokens": 19147117.0,
      "step": 8740
    },
    {
      "entropy": 5.550886249542236,
      "epoch": 0.8679039301310044,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004933410882434488,
      "loss": 5.3556,
      "mean_token_accuracy": 0.1573579952120781,
      "num_tokens": 19158131.0,
      "step": 8745
    },
    {
      "entropy": 5.537053728103638,
      "epoch": 0.8684001587931719,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004933325294019659,
      "loss": 5.2293,
      "mean_token_accuracy": 0.16480688899755477,
      "num_tokens": 19168820.0,
      "step": 8750
    },
    {
      "entropy": 5.595280361175537,
      "epoch": 0.8688963874553394,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004933239651463068,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1653578206896782,
      "num_tokens": 19178845.0,
      "step": 8755
    },
    {
      "entropy": 5.589409971237183,
      "epoch": 0.869392616117507,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004933153954766837,
      "loss": 5.2993,
      "mean_token_accuracy": 0.1594639912247658,
      "num_tokens": 19190830.0,
      "step": 8760
    },
    {
      "entropy": 5.431173849105835,
      "epoch": 0.8698888447796744,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004933068203933092,
      "loss": 5.1651,
      "mean_token_accuracy": 0.16443818509578706,
      "num_tokens": 19202056.0,
      "step": 8765
    },
    {
      "entropy": 5.5269123077392575,
      "epoch": 0.870385073441842,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000493298239896396,
      "loss": 5.3672,
      "mean_token_accuracy": 0.15586964040994644,
      "num_tokens": 19213497.0,
      "step": 8770
    },
    {
      "entropy": 5.6321765899658205,
      "epoch": 0.8708813021040095,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004932896539861568,
      "loss": 5.3989,
      "mean_token_accuracy": 0.15559251457452775,
      "num_tokens": 19226103.0,
      "step": 8775
    },
    {
      "entropy": 5.589433336257935,
      "epoch": 0.8713775307661771,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004932810626628047,
      "loss": 5.4099,
      "mean_token_accuracy": 0.15865050852298737,
      "num_tokens": 19237297.0,
      "step": 8780
    },
    {
      "entropy": 5.590228939056397,
      "epoch": 0.8718737594283446,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004932724659265526,
      "loss": 5.316,
      "mean_token_accuracy": 0.15872074663639069,
      "num_tokens": 19247537.0,
      "step": 8785
    },
    {
      "entropy": 5.503810262680053,
      "epoch": 0.8723699880905121,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004932638637776134,
      "loss": 5.2564,
      "mean_token_accuracy": 0.1594928875565529,
      "num_tokens": 19260465.0,
      "step": 8790
    },
    {
      "entropy": 5.576683902740479,
      "epoch": 0.8728662167526796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000493255256216201,
      "loss": 5.3346,
      "mean_token_accuracy": 0.1576985627412796,
      "num_tokens": 19270534.0,
      "step": 8795
    },
    {
      "entropy": 5.572884750366211,
      "epoch": 0.8733624454148472,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004932466432425284,
      "loss": 5.347,
      "mean_token_accuracy": 0.15596090704202653,
      "num_tokens": 19281551.0,
      "step": 8800
    },
    {
      "entropy": 5.530863094329834,
      "epoch": 0.8738586740770147,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004932380248568094,
      "loss": 5.2701,
      "mean_token_accuracy": 0.1637912079691887,
      "num_tokens": 19293300.0,
      "step": 8805
    },
    {
      "entropy": 5.523511934280395,
      "epoch": 0.8743549027391823,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004932294010592575,
      "loss": 5.2173,
      "mean_token_accuracy": 0.16384039968252181,
      "num_tokens": 19302701.0,
      "step": 8810
    },
    {
      "entropy": 5.587890577316284,
      "epoch": 0.8748511314013497,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004932207718500868,
      "loss": 5.3148,
      "mean_token_accuracy": 0.1593190163373947,
      "num_tokens": 19313024.0,
      "step": 8815
    },
    {
      "entropy": 5.602646684646606,
      "epoch": 0.8753473600635173,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004932121372295112,
      "loss": 5.3274,
      "mean_token_accuracy": 0.15838188976049422,
      "num_tokens": 19324014.0,
      "step": 8820
    },
    {
      "entropy": 5.5477033138275145,
      "epoch": 0.8758435887256848,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004932034971977448,
      "loss": 5.3528,
      "mean_token_accuracy": 0.155606609582901,
      "num_tokens": 19334226.0,
      "step": 8825
    },
    {
      "entropy": 5.597412204742431,
      "epoch": 0.8763398173878523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004931948517550019,
      "loss": 5.379,
      "mean_token_accuracy": 0.15326858907938004,
      "num_tokens": 19344363.0,
      "step": 8830
    },
    {
      "entropy": 5.609567165374756,
      "epoch": 0.8768360460500199,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004931862009014968,
      "loss": 5.3906,
      "mean_token_accuracy": 0.15667886584997176,
      "num_tokens": 19354915.0,
      "step": 8835
    },
    {
      "entropy": 5.66124472618103,
      "epoch": 0.8773322747121873,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000493177544637444,
      "loss": 5.4061,
      "mean_token_accuracy": 0.15496475100517274,
      "num_tokens": 19365570.0,
      "step": 8840
    },
    {
      "entropy": 5.623072338104248,
      "epoch": 0.8778285033743549,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004931688829630582,
      "loss": 5.3191,
      "mean_token_accuracy": 0.16266682147979736,
      "num_tokens": 19375731.0,
      "step": 8845
    },
    {
      "entropy": 5.534627151489258,
      "epoch": 0.8783247320365224,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004931602158785544,
      "loss": 5.273,
      "mean_token_accuracy": 0.16416835635900498,
      "num_tokens": 19386248.0,
      "step": 8850
    },
    {
      "entropy": 5.5298747539520265,
      "epoch": 0.87882096069869,
      "grad_norm": 0.875,
      "learning_rate": 0.0004931515433841472,
      "loss": 5.2774,
      "mean_token_accuracy": 0.16458217948675155,
      "num_tokens": 19397593.0,
      "step": 8855
    },
    {
      "entropy": 5.5762992858886715,
      "epoch": 0.8793171893608575,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004931428654800518,
      "loss": 5.3725,
      "mean_token_accuracy": 0.15380093157291413,
      "num_tokens": 19409701.0,
      "step": 8860
    },
    {
      "entropy": 5.61271448135376,
      "epoch": 0.879813418023025,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004931341821664834,
      "loss": 5.3352,
      "mean_token_accuracy": 0.1559494763612747,
      "num_tokens": 19420280.0,
      "step": 8865
    },
    {
      "entropy": 5.601084566116333,
      "epoch": 0.8803096466851925,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004931254934436573,
      "loss": 5.3221,
      "mean_token_accuracy": 0.15986086130142213,
      "num_tokens": 19432208.0,
      "step": 8870
    },
    {
      "entropy": 5.675123882293701,
      "epoch": 0.8808058753473601,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004931167993117891,
      "loss": 5.4506,
      "mean_token_accuracy": 0.15499379783868789,
      "num_tokens": 19443333.0,
      "step": 8875
    },
    {
      "entropy": 5.521538734436035,
      "epoch": 0.8813021040095276,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004931080997710942,
      "loss": 5.2132,
      "mean_token_accuracy": 0.16308022439479827,
      "num_tokens": 19452867.0,
      "step": 8880
    },
    {
      "entropy": 5.479450654983521,
      "epoch": 0.8817983326716952,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004930993948217884,
      "loss": 5.2317,
      "mean_token_accuracy": 0.16965189427137375,
      "num_tokens": 19463021.0,
      "step": 8885
    },
    {
      "entropy": 5.559165048599243,
      "epoch": 0.8822945613338626,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004930906844640876,
      "loss": 5.3715,
      "mean_token_accuracy": 0.15238905996084212,
      "num_tokens": 19474092.0,
      "step": 8890
    },
    {
      "entropy": 5.578556680679322,
      "epoch": 0.8827907899960302,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930819686982077,
      "loss": 5.3458,
      "mean_token_accuracy": 0.15323653519153596,
      "num_tokens": 19486035.0,
      "step": 8895
    },
    {
      "entropy": 5.553264188766479,
      "epoch": 0.8832870186581977,
      "grad_norm": 0.8359375,
      "learning_rate": 0.000493073247524365,
      "loss": 5.2826,
      "mean_token_accuracy": 0.15990747958421708,
      "num_tokens": 19497484.0,
      "step": 8900
    },
    {
      "entropy": 5.588619232177734,
      "epoch": 0.8837832473203652,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004930645209427756,
      "loss": 5.3885,
      "mean_token_accuracy": 0.15389522314071655,
      "num_tokens": 19508463.0,
      "step": 8905
    },
    {
      "entropy": 5.606539535522461,
      "epoch": 0.8842794759825328,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000493055788953656,
      "loss": 5.3108,
      "mean_token_accuracy": 0.16238888204097748,
      "num_tokens": 19520260.0,
      "step": 8910
    },
    {
      "entropy": 5.576453590393067,
      "epoch": 0.8847757046447002,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930470515572227,
      "loss": 5.2774,
      "mean_token_accuracy": 0.16129491478204727,
      "num_tokens": 19531621.0,
      "step": 8915
    },
    {
      "entropy": 5.5710655689239506,
      "epoch": 0.8852719333068678,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004930383087536923,
      "loss": 5.2422,
      "mean_token_accuracy": 0.15970994234085084,
      "num_tokens": 19543271.0,
      "step": 8920
    },
    {
      "entropy": 5.649300765991211,
      "epoch": 0.8857681619690353,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004930295605432818,
      "loss": 5.4159,
      "mean_token_accuracy": 0.16404466480016708,
      "num_tokens": 19555178.0,
      "step": 8925
    },
    {
      "entropy": 5.593044233322144,
      "epoch": 0.8862643906312029,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004930208069262079,
      "loss": 5.3254,
      "mean_token_accuracy": 0.15738084614276887,
      "num_tokens": 19565153.0,
      "step": 8930
    },
    {
      "entropy": 5.532147741317749,
      "epoch": 0.8867606192933704,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004930120479026878,
      "loss": 5.273,
      "mean_token_accuracy": 0.16588089168071746,
      "num_tokens": 19575608.0,
      "step": 8935
    },
    {
      "entropy": 5.575547361373902,
      "epoch": 0.8872568479555379,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004930032834729388,
      "loss": 5.2332,
      "mean_token_accuracy": 0.16844702810049056,
      "num_tokens": 19587556.0,
      "step": 8940
    },
    {
      "entropy": 5.524175024032592,
      "epoch": 0.8877530766177054,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000492994513637178,
      "loss": 5.2868,
      "mean_token_accuracy": 0.15982206910848618,
      "num_tokens": 19597341.0,
      "step": 8945
    },
    {
      "entropy": 5.550887680053711,
      "epoch": 0.888249305279873,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929857383956229,
      "loss": 5.3029,
      "mean_token_accuracy": 0.158498714864254,
      "num_tokens": 19608431.0,
      "step": 8950
    },
    {
      "entropy": 5.5870387077331545,
      "epoch": 0.8887455339420405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004929769577484914,
      "loss": 5.2941,
      "mean_token_accuracy": 0.16587130278348922,
      "num_tokens": 19619645.0,
      "step": 8955
    },
    {
      "entropy": 5.537397813796997,
      "epoch": 0.8892417626042081,
      "grad_norm": 1.0,
      "learning_rate": 0.000492968171696001,
      "loss": 5.242,
      "mean_token_accuracy": 0.1683274105191231,
      "num_tokens": 19630412.0,
      "step": 8960
    },
    {
      "entropy": 5.573724460601807,
      "epoch": 0.8897379912663755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004929593802383696,
      "loss": 5.3203,
      "mean_token_accuracy": 0.16155780255794525,
      "num_tokens": 19640055.0,
      "step": 8965
    },
    {
      "entropy": 5.560134601593018,
      "epoch": 0.8902342199285431,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929505833758152,
      "loss": 5.3184,
      "mean_token_accuracy": 0.15592312440276146,
      "num_tokens": 19650814.0,
      "step": 8970
    },
    {
      "entropy": 5.587675142288208,
      "epoch": 0.8907304485907106,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004929417811085561,
      "loss": 5.3083,
      "mean_token_accuracy": 0.1629286676645279,
      "num_tokens": 19663669.0,
      "step": 8975
    },
    {
      "entropy": 5.661772775650024,
      "epoch": 0.8912266772528781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004929329734368103,
      "loss": 5.3924,
      "mean_token_accuracy": 0.15559107288718224,
      "num_tokens": 19675851.0,
      "step": 8980
    },
    {
      "entropy": 5.59484372138977,
      "epoch": 0.8917229059150457,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004929241603607964,
      "loss": 5.2751,
      "mean_token_accuracy": 0.1661703661084175,
      "num_tokens": 19686613.0,
      "step": 8985
    },
    {
      "entropy": 5.517695951461792,
      "epoch": 0.8922191345772131,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492915341880733,
      "loss": 5.2171,
      "mean_token_accuracy": 0.16442286670207978,
      "num_tokens": 19696419.0,
      "step": 8990
    },
    {
      "entropy": 5.5443480014801025,
      "epoch": 0.8927153632393807,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004929065179968387,
      "loss": 5.3124,
      "mean_token_accuracy": 0.16015457957983018,
      "num_tokens": 19707269.0,
      "step": 8995
    },
    {
      "entropy": 5.620941972732544,
      "epoch": 0.8932115919015482,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004928976887093321,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1552276134490967,
      "num_tokens": 19716099.0,
      "step": 9000
    },
    {
      "epoch": 0.8932115919015482,
      "eval_entropy": 5.377903532375772,
      "eval_loss": 5.384802341461182,
      "eval_mean_token_accuracy": 0.16113636515685884,
      "eval_num_tokens": 19716099.0,
      "eval_runtime": 20.4104,
      "eval_samples_per_second": 1649.603,
      "eval_steps_per_second": 206.219,
      "step": 9000
    },
    {
      "entropy": 5.5677838802337645,
      "epoch": 0.8937078205637158,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928888540184325,
      "loss": 5.3382,
      "mean_token_accuracy": 0.15805651545524596,
      "num_tokens": 19726632.0,
      "step": 9005
    },
    {
      "entropy": 5.5465904712677006,
      "epoch": 0.8942040492258833,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004928800139243587,
      "loss": 5.2481,
      "mean_token_accuracy": 0.16045569777488708,
      "num_tokens": 19737929.0,
      "step": 9010
    },
    {
      "entropy": 5.570677089691162,
      "epoch": 0.8947002778880508,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004928711684273302,
      "loss": 5.2705,
      "mean_token_accuracy": 0.16451206505298616,
      "num_tokens": 19748191.0,
      "step": 9015
    },
    {
      "entropy": 5.575153779983521,
      "epoch": 0.8951965065502183,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004928623175275661,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1614478513598442,
      "num_tokens": 19758892.0,
      "step": 9020
    },
    {
      "entropy": 5.531454706192017,
      "epoch": 0.8956927352123859,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004928534612252862,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1635870710015297,
      "num_tokens": 19769120.0,
      "step": 9025
    },
    {
      "entropy": 5.508886003494263,
      "epoch": 0.8961889638745534,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004928445995207095,
      "loss": 5.2659,
      "mean_token_accuracy": 0.16378669291734696,
      "num_tokens": 19779484.0,
      "step": 9030
    },
    {
      "entropy": 5.556185865402222,
      "epoch": 0.896685192536721,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004928357324140564,
      "loss": 5.3058,
      "mean_token_accuracy": 0.16304608136415483,
      "num_tokens": 19790438.0,
      "step": 9035
    },
    {
      "entropy": 5.6387793064117435,
      "epoch": 0.8971814211988884,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004928268599055465,
      "loss": 5.3878,
      "mean_token_accuracy": 0.15672214329242706,
      "num_tokens": 19801646.0,
      "step": 9040
    },
    {
      "entropy": 5.5769758224487305,
      "epoch": 0.897677649861056,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004928179819953998,
      "loss": 5.266,
      "mean_token_accuracy": 0.16250865161418915,
      "num_tokens": 19812887.0,
      "step": 9045
    },
    {
      "entropy": 5.5438604831695555,
      "epoch": 0.8981738785232235,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928090986838366,
      "loss": 5.3244,
      "mean_token_accuracy": 0.15720922797918319,
      "num_tokens": 19822938.0,
      "step": 9050
    },
    {
      "entropy": 5.547235441207886,
      "epoch": 0.898670107185391,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492800209971077,
      "loss": 5.2842,
      "mean_token_accuracy": 0.1630757376551628,
      "num_tokens": 19833988.0,
      "step": 9055
    },
    {
      "entropy": 5.645789957046508,
      "epoch": 0.8991663358475586,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004927913158573416,
      "loss": 5.3507,
      "mean_token_accuracy": 0.15545873939990998,
      "num_tokens": 19845112.0,
      "step": 9060
    },
    {
      "entropy": 5.547437477111816,
      "epoch": 0.899662564509726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004927824163428508,
      "loss": 5.3084,
      "mean_token_accuracy": 0.15988384634256364,
      "num_tokens": 19856304.0,
      "step": 9065
    },
    {
      "entropy": 5.544822072982788,
      "epoch": 0.9001587931718936,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927735114278254,
      "loss": 5.3029,
      "mean_token_accuracy": 0.1646771177649498,
      "num_tokens": 19866776.0,
      "step": 9070
    },
    {
      "entropy": 5.587491035461426,
      "epoch": 0.9006550218340611,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004927646011124861,
      "loss": 5.2749,
      "mean_token_accuracy": 0.15796078592538834,
      "num_tokens": 19876392.0,
      "step": 9075
    },
    {
      "entropy": 5.61587176322937,
      "epoch": 0.9011512504962287,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004927556853970539,
      "loss": 5.3098,
      "mean_token_accuracy": 0.1541835442185402,
      "num_tokens": 19887362.0,
      "step": 9080
    },
    {
      "entropy": 5.544364976882934,
      "epoch": 0.9016474791583962,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004927467642817499,
      "loss": 5.2877,
      "mean_token_accuracy": 0.16370008438825606,
      "num_tokens": 19898989.0,
      "step": 9085
    },
    {
      "entropy": 5.578433322906494,
      "epoch": 0.9021437078205637,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004927378377667955,
      "loss": 5.2909,
      "mean_token_accuracy": 0.15280220210552214,
      "num_tokens": 19910112.0,
      "step": 9090
    },
    {
      "entropy": 5.6322808265686035,
      "epoch": 0.9026399364827312,
      "grad_norm": 0.921875,
      "learning_rate": 0.000492728905852412,
      "loss": 5.4105,
      "mean_token_accuracy": 0.15606809556484222,
      "num_tokens": 19922601.0,
      "step": 9095
    },
    {
      "entropy": 5.553589820861816,
      "epoch": 0.9031361651448988,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004927199685388205,
      "loss": 5.3214,
      "mean_token_accuracy": 0.16237563639879227,
      "num_tokens": 19934153.0,
      "step": 9100
    },
    {
      "entropy": 5.575612878799438,
      "epoch": 0.9036323938070663,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000492711025826243,
      "loss": 5.2896,
      "mean_token_accuracy": 0.159117966145277,
      "num_tokens": 19944572.0,
      "step": 9105
    },
    {
      "entropy": 5.595759677886963,
      "epoch": 0.9041286224692339,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004927020777149014,
      "loss": 5.3992,
      "mean_token_accuracy": 0.15489456206560134,
      "num_tokens": 19956431.0,
      "step": 9110
    },
    {
      "entropy": 5.618905305862427,
      "epoch": 0.9046248511314013,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004926931242050171,
      "loss": 5.3204,
      "mean_token_accuracy": 0.15188727378845215,
      "num_tokens": 19968118.0,
      "step": 9115
    },
    {
      "entropy": 5.552833843231201,
      "epoch": 0.9051210797935689,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004926841652968125,
      "loss": 5.3829,
      "mean_token_accuracy": 0.16006607413291932,
      "num_tokens": 19979545.0,
      "step": 9120
    },
    {
      "entropy": 5.568428373336792,
      "epoch": 0.9056173084557364,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004926752009905099,
      "loss": 5.3431,
      "mean_token_accuracy": 0.15130785554647447,
      "num_tokens": 19989832.0,
      "step": 9125
    },
    {
      "entropy": 5.577058219909668,
      "epoch": 0.9061135371179039,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004926662312863311,
      "loss": 5.3573,
      "mean_token_accuracy": 0.159266896545887,
      "num_tokens": 20000935.0,
      "step": 9130
    },
    {
      "entropy": 5.572552824020386,
      "epoch": 0.9066097657800715,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004926572561844988,
      "loss": 5.3523,
      "mean_token_accuracy": 0.15485593527555466,
      "num_tokens": 20011485.0,
      "step": 9135
    },
    {
      "entropy": 5.550427341461182,
      "epoch": 0.9071059944422389,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004926482756852356,
      "loss": 5.2149,
      "mean_token_accuracy": 0.1645887777209282,
      "num_tokens": 20023043.0,
      "step": 9140
    },
    {
      "entropy": 5.592850399017334,
      "epoch": 0.9076022231044065,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004926392897887642,
      "loss": 5.319,
      "mean_token_accuracy": 0.16620936393737792,
      "num_tokens": 20032437.0,
      "step": 9145
    },
    {
      "entropy": 5.573175287246704,
      "epoch": 0.908098451766574,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004926302984953073,
      "loss": 5.3451,
      "mean_token_accuracy": 0.15950006246566772,
      "num_tokens": 20044662.0,
      "step": 9150
    },
    {
      "entropy": 5.564566564559937,
      "epoch": 0.9085946804287416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004926213018050881,
      "loss": 5.3602,
      "mean_token_accuracy": 0.16055931150913239,
      "num_tokens": 20055924.0,
      "step": 9155
    },
    {
      "entropy": 5.520517063140869,
      "epoch": 0.9090909090909091,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004926122997183295,
      "loss": 5.3061,
      "mean_token_accuracy": 0.1581579178571701,
      "num_tokens": 20067909.0,
      "step": 9160
    },
    {
      "entropy": 5.582996940612793,
      "epoch": 0.9095871377530766,
      "grad_norm": 0.875,
      "learning_rate": 0.0004926032922352546,
      "loss": 5.2799,
      "mean_token_accuracy": 0.16499029844999313,
      "num_tokens": 20079333.0,
      "step": 9165
    },
    {
      "entropy": 5.557482290267944,
      "epoch": 0.9100833664152441,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004925942793560872,
      "loss": 5.259,
      "mean_token_accuracy": 0.16249357014894486,
      "num_tokens": 20089741.0,
      "step": 9170
    },
    {
      "entropy": 5.519131755828857,
      "epoch": 0.9105795950774117,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004925852610810504,
      "loss": 5.2856,
      "mean_token_accuracy": 0.1625542774796486,
      "num_tokens": 20100267.0,
      "step": 9175
    },
    {
      "entropy": 5.545725870132446,
      "epoch": 0.9110758237395792,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000492576237410368,
      "loss": 5.2652,
      "mean_token_accuracy": 0.159230774641037,
      "num_tokens": 20111964.0,
      "step": 9180
    },
    {
      "entropy": 5.590697431564331,
      "epoch": 0.9115720524017468,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004925672083442638,
      "loss": 5.3252,
      "mean_token_accuracy": 0.15364896208047868,
      "num_tokens": 20123132.0,
      "step": 9185
    },
    {
      "entropy": 5.513075876235962,
      "epoch": 0.9120682810639142,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004925581738829616,
      "loss": 5.2277,
      "mean_token_accuracy": 0.16437134593725206,
      "num_tokens": 20134050.0,
      "step": 9190
    },
    {
      "entropy": 5.519852924346924,
      "epoch": 0.9125645097260818,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004925491340266855,
      "loss": 5.2847,
      "mean_token_accuracy": 0.16542524099349976,
      "num_tokens": 20143814.0,
      "step": 9195
    },
    {
      "entropy": 5.595990705490112,
      "epoch": 0.9130607383882493,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004925400887756596,
      "loss": 5.3657,
      "mean_token_accuracy": 0.16596104204654694,
      "num_tokens": 20157647.0,
      "step": 9200
    },
    {
      "entropy": 5.584433937072754,
      "epoch": 0.9135569670504168,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004925310381301083,
      "loss": 5.3384,
      "mean_token_accuracy": 0.15844662338495255,
      "num_tokens": 20168366.0,
      "step": 9205
    },
    {
      "entropy": 5.564235877990723,
      "epoch": 0.9140531957125844,
      "grad_norm": 0.890625,
      "learning_rate": 0.000492521982090256,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1611117273569107,
      "num_tokens": 20179362.0,
      "step": 9210
    },
    {
      "entropy": 5.541329765319825,
      "epoch": 0.9145494243747518,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004925129206563273,
      "loss": 5.2649,
      "mean_token_accuracy": 0.16410792469978333,
      "num_tokens": 20190320.0,
      "step": 9215
    },
    {
      "entropy": 5.537449550628662,
      "epoch": 0.9150456530369194,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004925038538285467,
      "loss": 5.2765,
      "mean_token_accuracy": 0.1630534127354622,
      "num_tokens": 20201114.0,
      "step": 9220
    },
    {
      "entropy": 5.570141696929932,
      "epoch": 0.9155418816990869,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004924947816071394,
      "loss": 5.3085,
      "mean_token_accuracy": 0.1627030313014984,
      "num_tokens": 20213351.0,
      "step": 9225
    },
    {
      "entropy": 5.518476438522339,
      "epoch": 0.9160381103612545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00049248570399233,
      "loss": 5.2096,
      "mean_token_accuracy": 0.16812825202941895,
      "num_tokens": 20223376.0,
      "step": 9230
    },
    {
      "entropy": 5.584335708618164,
      "epoch": 0.916534339023422,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000492476620984344,
      "loss": 5.356,
      "mean_token_accuracy": 0.15710895806550979,
      "num_tokens": 20234248.0,
      "step": 9235
    },
    {
      "entropy": 5.526259231567383,
      "epoch": 0.9170305676855895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004924675325834063,
      "loss": 5.2808,
      "mean_token_accuracy": 0.16342011392116546,
      "num_tokens": 20245823.0,
      "step": 9240
    },
    {
      "entropy": 5.57660608291626,
      "epoch": 0.917526796347757,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004924584387897424,
      "loss": 5.3309,
      "mean_token_accuracy": 0.15736903101205826,
      "num_tokens": 20257003.0,
      "step": 9245
    },
    {
      "entropy": 5.625552701950073,
      "epoch": 0.9180230250099246,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004924493396035777,
      "loss": 5.4582,
      "mean_token_accuracy": 0.15018327981233598,
      "num_tokens": 20268500.0,
      "step": 9250
    },
    {
      "entropy": 5.632844114303589,
      "epoch": 0.9185192536720921,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000492440235025138,
      "loss": 5.3361,
      "mean_token_accuracy": 0.15995648503303528,
      "num_tokens": 20279301.0,
      "step": 9255
    },
    {
      "entropy": 5.680012178421021,
      "epoch": 0.9190154823342597,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004924311250546491,
      "loss": 5.4278,
      "mean_token_accuracy": 0.1542675033211708,
      "num_tokens": 20291263.0,
      "step": 9260
    },
    {
      "entropy": 5.561777782440186,
      "epoch": 0.9195117109964271,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004924220096923367,
      "loss": 5.3101,
      "mean_token_accuracy": 0.15981553345918656,
      "num_tokens": 20301554.0,
      "step": 9265
    },
    {
      "entropy": 5.638153553009033,
      "epoch": 0.9200079396585947,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004924128889384271,
      "loss": 5.3543,
      "mean_token_accuracy": 0.153249791264534,
      "num_tokens": 20314299.0,
      "step": 9270
    },
    {
      "entropy": 5.548483180999756,
      "epoch": 0.9205041683207622,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004924037627931463,
      "loss": 5.318,
      "mean_token_accuracy": 0.15937757194042207,
      "num_tokens": 20324595.0,
      "step": 9275
    },
    {
      "entropy": 5.5754638671875,
      "epoch": 0.9210003969829297,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004923946312567206,
      "loss": 5.2839,
      "mean_token_accuracy": 0.16126274764537812,
      "num_tokens": 20335744.0,
      "step": 9280
    },
    {
      "entropy": 5.516260099411011,
      "epoch": 0.9214966256450973,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004923854943293765,
      "loss": 5.1943,
      "mean_token_accuracy": 0.15963348597288132,
      "num_tokens": 20346072.0,
      "step": 9285
    },
    {
      "entropy": 5.544660377502441,
      "epoch": 0.9219928543072647,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0004923763520113406,
      "loss": 5.2688,
      "mean_token_accuracy": 0.17300408482551574,
      "num_tokens": 20357747.0,
      "step": 9290
    },
    {
      "entropy": 5.605404615402222,
      "epoch": 0.9224890829694323,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004923672043028395,
      "loss": 5.2568,
      "mean_token_accuracy": 0.1635786473751068,
      "num_tokens": 20368838.0,
      "step": 9295
    },
    {
      "entropy": 5.5697259426116945,
      "epoch": 0.9229853116315998,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923580512041001,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1558098614215851,
      "num_tokens": 20380844.0,
      "step": 9300
    },
    {
      "entropy": 5.588720273971558,
      "epoch": 0.9234815402937674,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004923488927153495,
      "loss": 5.3406,
      "mean_token_accuracy": 0.15699591785669326,
      "num_tokens": 20391739.0,
      "step": 9305
    },
    {
      "entropy": 5.577686023712158,
      "epoch": 0.9239777689559349,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923397288368146,
      "loss": 5.234,
      "mean_token_accuracy": 0.166792631149292,
      "num_tokens": 20402571.0,
      "step": 9310
    },
    {
      "entropy": 5.549501085281372,
      "epoch": 0.9244739976181024,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004923305595687229,
      "loss": 5.2677,
      "mean_token_accuracy": 0.15813456624746322,
      "num_tokens": 20414766.0,
      "step": 9315
    },
    {
      "entropy": 5.537642955780029,
      "epoch": 0.9249702262802699,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004923213849113015,
      "loss": 5.2378,
      "mean_token_accuracy": 0.16708897948265075,
      "num_tokens": 20425508.0,
      "step": 9320
    },
    {
      "entropy": 5.512779664993286,
      "epoch": 0.9254664549424375,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004923122048647781,
      "loss": 5.2307,
      "mean_token_accuracy": 0.16518714427947997,
      "num_tokens": 20436128.0,
      "step": 9325
    },
    {
      "entropy": 5.518051338195801,
      "epoch": 0.925962683604605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004923030194293802,
      "loss": 5.341,
      "mean_token_accuracy": 0.1588973119854927,
      "num_tokens": 20447212.0,
      "step": 9330
    },
    {
      "entropy": 5.551166772842407,
      "epoch": 0.9264589122667726,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004922938286053358,
      "loss": 5.2582,
      "mean_token_accuracy": 0.16507821083068847,
      "num_tokens": 20457770.0,
      "step": 9335
    },
    {
      "entropy": 5.583638429641724,
      "epoch": 0.92695514092894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004922846323928726,
      "loss": 5.3267,
      "mean_token_accuracy": 0.15723714977502823,
      "num_tokens": 20468172.0,
      "step": 9340
    },
    {
      "entropy": 5.5516290187835695,
      "epoch": 0.9274513695911076,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004922754307922188,
      "loss": 5.2535,
      "mean_token_accuracy": 0.16497961729764937,
      "num_tokens": 20479429.0,
      "step": 9345
    },
    {
      "entropy": 5.560518980026245,
      "epoch": 0.9279475982532751,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004922662238036024,
      "loss": 5.2257,
      "mean_token_accuracy": 0.16623591780662536,
      "num_tokens": 20490373.0,
      "step": 9350
    },
    {
      "entropy": 5.613909149169922,
      "epoch": 0.9284438269154426,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004922570114272518,
      "loss": 5.3598,
      "mean_token_accuracy": 0.15748854726552963,
      "num_tokens": 20500002.0,
      "step": 9355
    },
    {
      "entropy": 5.607706117630005,
      "epoch": 0.9289400555776102,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922477936633955,
      "loss": 5.4005,
      "mean_token_accuracy": 0.15188513100147247,
      "num_tokens": 20512990.0,
      "step": 9360
    },
    {
      "entropy": 5.525366878509521,
      "epoch": 0.9294362842397776,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922385705122621,
      "loss": 5.3008,
      "mean_token_accuracy": 0.1623746261000633,
      "num_tokens": 20522979.0,
      "step": 9365
    },
    {
      "entropy": 5.501667451858521,
      "epoch": 0.9299325129019452,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004922293419740802,
      "loss": 5.1941,
      "mean_token_accuracy": 0.16979625970125198,
      "num_tokens": 20534620.0,
      "step": 9370
    },
    {
      "entropy": 5.573191356658936,
      "epoch": 0.9304287415641127,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004922201080490787,
      "loss": 5.2859,
      "mean_token_accuracy": 0.15834615677595137,
      "num_tokens": 20545526.0,
      "step": 9375
    },
    {
      "entropy": 5.63176646232605,
      "epoch": 0.9309249702262803,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004922108687374867,
      "loss": 5.4521,
      "mean_token_accuracy": 0.15453197360038756,
      "num_tokens": 20556665.0,
      "step": 9380
    },
    {
      "entropy": 5.585740566253662,
      "epoch": 0.9314211988884478,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492201624039533,
      "loss": 5.3086,
      "mean_token_accuracy": 0.15535683333873748,
      "num_tokens": 20566054.0,
      "step": 9385
    },
    {
      "entropy": 5.65473690032959,
      "epoch": 0.9319174275506154,
      "grad_norm": 0.875,
      "learning_rate": 0.0004921923739554472,
      "loss": 5.3761,
      "mean_token_accuracy": 0.15707683861255645,
      "num_tokens": 20577922.0,
      "step": 9390
    },
    {
      "entropy": 5.516871786117553,
      "epoch": 0.9324136562127828,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004921831184854585,
      "loss": 5.1788,
      "mean_token_accuracy": 0.16971657127141954,
      "num_tokens": 20588473.0,
      "step": 9395
    },
    {
      "entropy": 5.549216651916504,
      "epoch": 0.9329098848749504,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004921738576297965,
      "loss": 5.3765,
      "mean_token_accuracy": 0.15780840292572976,
      "num_tokens": 20599371.0,
      "step": 9400
    },
    {
      "entropy": 5.515951538085938,
      "epoch": 0.9334061135371179,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921645913886907,
      "loss": 5.2486,
      "mean_token_accuracy": 0.15473873615264894,
      "num_tokens": 20610066.0,
      "step": 9405
    },
    {
      "entropy": 5.573815631866455,
      "epoch": 0.9339023421992855,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004921553197623711,
      "loss": 5.3551,
      "mean_token_accuracy": 0.15802679359912872,
      "num_tokens": 20621804.0,
      "step": 9410
    },
    {
      "entropy": 5.59134464263916,
      "epoch": 0.934398570861453,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921460427510674,
      "loss": 5.2407,
      "mean_token_accuracy": 0.1692960485816002,
      "num_tokens": 20632706.0,
      "step": 9415
    },
    {
      "entropy": 5.541371870040893,
      "epoch": 0.9348947995236205,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004921367603550098,
      "loss": 5.3549,
      "mean_token_accuracy": 0.1550799861550331,
      "num_tokens": 20643303.0,
      "step": 9420
    },
    {
      "entropy": 5.563239240646363,
      "epoch": 0.935391028185788,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004921274725744283,
      "loss": 5.3136,
      "mean_token_accuracy": 0.16914501637220383,
      "num_tokens": 20653681.0,
      "step": 9425
    },
    {
      "entropy": 5.6122636795043945,
      "epoch": 0.9358872568479555,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921181794095536,
      "loss": 5.3207,
      "mean_token_accuracy": 0.16096874624490737,
      "num_tokens": 20665599.0,
      "step": 9430
    },
    {
      "entropy": 5.501299428939819,
      "epoch": 0.9363834855101231,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004921088808606157,
      "loss": 5.2085,
      "mean_token_accuracy": 0.16178009361028672,
      "num_tokens": 20675726.0,
      "step": 9435
    },
    {
      "entropy": 5.543586921691895,
      "epoch": 0.9368797141722905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004920995769278454,
      "loss": 5.3927,
      "mean_token_accuracy": 0.1539223313331604,
      "num_tokens": 20688083.0,
      "step": 9440
    },
    {
      "entropy": 5.606049203872681,
      "epoch": 0.9373759428344581,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920902676114734,
      "loss": 5.2589,
      "mean_token_accuracy": 0.16520233750343322,
      "num_tokens": 20697642.0,
      "step": 9445
    },
    {
      "entropy": 5.6175336837768555,
      "epoch": 0.9378721714966256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004920809529117306,
      "loss": 5.3544,
      "mean_token_accuracy": 0.15562134236097336,
      "num_tokens": 20709172.0,
      "step": 9450
    },
    {
      "entropy": 5.589820766448975,
      "epoch": 0.9383684001587932,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004920716328288479,
      "loss": 5.389,
      "mean_token_accuracy": 0.15578695237636567,
      "num_tokens": 20720563.0,
      "step": 9455
    },
    {
      "entropy": 5.5717333316802975,
      "epoch": 0.9388646288209607,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004920623073630566,
      "loss": 5.2793,
      "mean_token_accuracy": 0.16606247425079346,
      "num_tokens": 20730943.0,
      "step": 9460
    },
    {
      "entropy": 5.567726087570191,
      "epoch": 0.9393608574831283,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004920529765145876,
      "loss": 5.2568,
      "mean_token_accuracy": 0.16133053004741668,
      "num_tokens": 20740903.0,
      "step": 9465
    },
    {
      "entropy": 5.572282075881958,
      "epoch": 0.9398570861452957,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004920436402836726,
      "loss": 5.3504,
      "mean_token_accuracy": 0.1623139947652817,
      "num_tokens": 20750580.0,
      "step": 9470
    },
    {
      "entropy": 5.547543144226074,
      "epoch": 0.9403533148074633,
      "grad_norm": 0.90625,
      "learning_rate": 0.000492034298670543,
      "loss": 5.2617,
      "mean_token_accuracy": 0.15864874422550201,
      "num_tokens": 20761357.0,
      "step": 9475
    },
    {
      "entropy": 5.654269504547119,
      "epoch": 0.9408495434696308,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004920249516754304,
      "loss": 5.4251,
      "mean_token_accuracy": 0.14949466288089752,
      "num_tokens": 20773273.0,
      "step": 9480
    },
    {
      "entropy": 5.564805412292481,
      "epoch": 0.9413457721317984,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004920155992985667,
      "loss": 5.2585,
      "mean_token_accuracy": 0.15717626512050628,
      "num_tokens": 20784290.0,
      "step": 9485
    },
    {
      "entropy": 5.560206651687622,
      "epoch": 0.9418420007939659,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004920062415401837,
      "loss": 5.2877,
      "mean_token_accuracy": 0.16689648032188414,
      "num_tokens": 20794764.0,
      "step": 9490
    },
    {
      "entropy": 5.612143611907959,
      "epoch": 0.9423382294561334,
      "grad_norm": 1.0,
      "learning_rate": 0.0004919968784005136,
      "loss": 5.3326,
      "mean_token_accuracy": 0.153749680519104,
      "num_tokens": 20804863.0,
      "step": 9495
    },
    {
      "entropy": 5.622992897033692,
      "epoch": 0.9428344581183009,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004919875098797886,
      "loss": 5.4145,
      "mean_token_accuracy": 0.15409638434648515,
      "num_tokens": 20816445.0,
      "step": 9500
    },
    {
      "entropy": 5.560190582275391,
      "epoch": 0.9433306867804684,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004919781359782407,
      "loss": 5.2422,
      "mean_token_accuracy": 0.16560266762971879,
      "num_tokens": 20827141.0,
      "step": 9505
    },
    {
      "entropy": 5.535797834396362,
      "epoch": 0.943826915442636,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004919687566961027,
      "loss": 5.2818,
      "mean_token_accuracy": 0.15792427062988282,
      "num_tokens": 20838291.0,
      "step": 9510
    },
    {
      "entropy": 5.599298191070557,
      "epoch": 0.9443231441048034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491959372033607,
      "loss": 5.2905,
      "mean_token_accuracy": 0.16101907938718796,
      "num_tokens": 20848839.0,
      "step": 9515
    },
    {
      "entropy": 5.598592138290405,
      "epoch": 0.944819372766971,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004919499819909865,
      "loss": 5.2809,
      "mean_token_accuracy": 0.16426681727170944,
      "num_tokens": 20859961.0,
      "step": 9520
    },
    {
      "entropy": 5.4815366744995115,
      "epoch": 0.9453156014291385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919405865684738,
      "loss": 5.2212,
      "mean_token_accuracy": 0.16402681767940522,
      "num_tokens": 20869720.0,
      "step": 9525
    },
    {
      "entropy": 5.5652608394622805,
      "epoch": 0.9458118300913061,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491931185766302,
      "loss": 5.3541,
      "mean_token_accuracy": 0.16435228139162064,
      "num_tokens": 20880513.0,
      "step": 9530
    },
    {
      "entropy": 5.62198839187622,
      "epoch": 0.9463080587534736,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004919217795847044,
      "loss": 5.2743,
      "mean_token_accuracy": 0.1658939465880394,
      "num_tokens": 20892456.0,
      "step": 9535
    },
    {
      "entropy": 5.523668050765991,
      "epoch": 0.9468042874156412,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004919123680239139,
      "loss": 5.2795,
      "mean_token_accuracy": 0.16097230613231658,
      "num_tokens": 20904500.0,
      "step": 9540
    },
    {
      "entropy": 5.615213203430176,
      "epoch": 0.9473005160778086,
      "grad_norm": 0.875,
      "learning_rate": 0.0004919029510841641,
      "loss": 5.2614,
      "mean_token_accuracy": 0.16305473148822786,
      "num_tokens": 20915133.0,
      "step": 9545
    },
    {
      "entropy": 5.506047487258911,
      "epoch": 0.9477967447399762,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004918935287656887,
      "loss": 5.2579,
      "mean_token_accuracy": 0.16558754295110703,
      "num_tokens": 20924779.0,
      "step": 9550
    },
    {
      "entropy": 5.472757720947266,
      "epoch": 0.9482929734021437,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491884101068721,
      "loss": 5.1992,
      "mean_token_accuracy": 0.16050480604171752,
      "num_tokens": 20935387.0,
      "step": 9555
    },
    {
      "entropy": 5.64392991065979,
      "epoch": 0.9487892020643113,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004918746679934949,
      "loss": 5.3804,
      "mean_token_accuracy": 0.15949872136116028,
      "num_tokens": 20945432.0,
      "step": 9560
    },
    {
      "entropy": 5.571486616134644,
      "epoch": 0.9492854307264788,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004918652295402444,
      "loss": 5.2607,
      "mean_token_accuracy": 0.1658664956688881,
      "num_tokens": 20956777.0,
      "step": 9565
    },
    {
      "entropy": 5.552468490600586,
      "epoch": 0.9497816593886463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004918557857092036,
      "loss": 5.2983,
      "mean_token_accuracy": 0.1609534278512001,
      "num_tokens": 20966057.0,
      "step": 9570
    },
    {
      "entropy": 5.567633247375488,
      "epoch": 0.9502778880508138,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004918463365006065,
      "loss": 5.3786,
      "mean_token_accuracy": 0.15900493413209915,
      "num_tokens": 20976838.0,
      "step": 9575
    },
    {
      "entropy": 5.653281021118164,
      "epoch": 0.9507741167129813,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004918368819146875,
      "loss": 5.3028,
      "mean_token_accuracy": 0.1623690605163574,
      "num_tokens": 20988553.0,
      "step": 9580
    },
    {
      "entropy": 5.568065071105957,
      "epoch": 0.9512703453751489,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004918274219516811,
      "loss": 5.3001,
      "mean_token_accuracy": 0.1599831148982048,
      "num_tokens": 20998983.0,
      "step": 9585
    },
    {
      "entropy": 5.6325348854064945,
      "epoch": 0.9517665740373163,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004918179566118218,
      "loss": 5.4075,
      "mean_token_accuracy": 0.15550417751073836,
      "num_tokens": 21010008.0,
      "step": 9590
    },
    {
      "entropy": 5.617930555343628,
      "epoch": 0.9522628026994839,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004918084858953445,
      "loss": 5.3806,
      "mean_token_accuracy": 0.16284365504980086,
      "num_tokens": 21022826.0,
      "step": 9595
    },
    {
      "entropy": 5.591018486022949,
      "epoch": 0.9527590313616514,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004917990098024839,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1552531436085701,
      "num_tokens": 21033457.0,
      "step": 9600
    },
    {
      "entropy": 5.5458667278289795,
      "epoch": 0.953255260023819,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004917895283334749,
      "loss": 5.313,
      "mean_token_accuracy": 0.15834345072507858,
      "num_tokens": 21044640.0,
      "step": 9605
    },
    {
      "entropy": 5.579812240600586,
      "epoch": 0.9537514886859865,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000491780041488553,
      "loss": 5.2988,
      "mean_token_accuracy": 0.16638065874576569,
      "num_tokens": 21055728.0,
      "step": 9610
    },
    {
      "entropy": 5.588179063796997,
      "epoch": 0.9542477173481541,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917705492679529,
      "loss": 5.2753,
      "mean_token_accuracy": 0.15793106853961944,
      "num_tokens": 21066057.0,
      "step": 9615
    },
    {
      "entropy": 5.616610956192017,
      "epoch": 0.9547439460103215,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004917610516719104,
      "loss": 5.302,
      "mean_token_accuracy": 0.15569670349359513,
      "num_tokens": 21075717.0,
      "step": 9620
    },
    {
      "entropy": 5.67068657875061,
      "epoch": 0.9552401746724891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004917515487006609,
      "loss": 5.3997,
      "mean_token_accuracy": 0.1533588156104088,
      "num_tokens": 21087120.0,
      "step": 9625
    },
    {
      "entropy": 5.618469381332398,
      "epoch": 0.9557364033346566,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00049174204035444,
      "loss": 5.4104,
      "mean_token_accuracy": 0.15602218434214593,
      "num_tokens": 21100110.0,
      "step": 9630
    },
    {
      "entropy": 5.6803868293762205,
      "epoch": 0.9562326319968242,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004917325266334836,
      "loss": 5.3214,
      "mean_token_accuracy": 0.16723915934562683,
      "num_tokens": 21110411.0,
      "step": 9635
    },
    {
      "entropy": 5.519408559799194,
      "epoch": 0.9567288606589917,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917230075380276,
      "loss": 5.2278,
      "mean_token_accuracy": 0.1674467459321022,
      "num_tokens": 21120303.0,
      "step": 9640
    },
    {
      "entropy": 5.553916358947754,
      "epoch": 0.9572250893211592,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004917134830683079,
      "loss": 5.3347,
      "mean_token_accuracy": 0.15733883827924727,
      "num_tokens": 21132312.0,
      "step": 9645
    },
    {
      "entropy": 5.630156230926514,
      "epoch": 0.9577213179833267,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004917039532245608,
      "loss": 5.3081,
      "mean_token_accuracy": 0.16051254868507386,
      "num_tokens": 21142472.0,
      "step": 9650
    },
    {
      "entropy": 5.583373022079468,
      "epoch": 0.9582175466454942,
      "grad_norm": 0.875,
      "learning_rate": 0.0004916944180070226,
      "loss": 5.3185,
      "mean_token_accuracy": 0.15896903425455094,
      "num_tokens": 21153306.0,
      "step": 9655
    },
    {
      "entropy": 5.645754480361939,
      "epoch": 0.9587137753076618,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004916848774159299,
      "loss": 5.3615,
      "mean_token_accuracy": 0.15320287495851517,
      "num_tokens": 21165606.0,
      "step": 9660
    },
    {
      "entropy": 5.5970155715942385,
      "epoch": 0.9592100039698293,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000491675331451519,
      "loss": 5.2937,
      "mean_token_accuracy": 0.16333945840597153,
      "num_tokens": 21176734.0,
      "step": 9665
    },
    {
      "entropy": 5.540189027786255,
      "epoch": 0.9597062326319968,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004916657801140269,
      "loss": 5.2296,
      "mean_token_accuracy": 0.1665262520313263,
      "num_tokens": 21187109.0,
      "step": 9670
    },
    {
      "entropy": 5.5527284145355225,
      "epoch": 0.9602024612941643,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004916562234036903,
      "loss": 5.2911,
      "mean_token_accuracy": 0.16412141174077988,
      "num_tokens": 21197281.0,
      "step": 9675
    },
    {
      "entropy": 5.555618810653686,
      "epoch": 0.9606986899563319,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004916466613207462,
      "loss": 5.3225,
      "mean_token_accuracy": 0.16170413047075272,
      "num_tokens": 21208767.0,
      "step": 9680
    },
    {
      "entropy": 5.580183744430542,
      "epoch": 0.9611949186184994,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004916370938654317,
      "loss": 5.2989,
      "mean_token_accuracy": 0.15791932493448257,
      "num_tokens": 21219328.0,
      "step": 9685
    },
    {
      "entropy": 5.650660562515259,
      "epoch": 0.961691147280667,
      "grad_norm": 0.875,
      "learning_rate": 0.0004916275210379841,
      "loss": 5.4279,
      "mean_token_accuracy": 0.1576124519109726,
      "num_tokens": 21231476.0,
      "step": 9690
    },
    {
      "entropy": 5.640081357955933,
      "epoch": 0.9621873759428344,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004916179428386408,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1626204952597618,
      "num_tokens": 21242445.0,
      "step": 9695
    },
    {
      "entropy": 5.5503541946411135,
      "epoch": 0.962683604605002,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004916083592676392,
      "loss": 5.1796,
      "mean_token_accuracy": 0.1633564606308937,
      "num_tokens": 21253179.0,
      "step": 9700
    },
    {
      "entropy": 5.5908397197723385,
      "epoch": 0.9631798332671695,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004915987703252171,
      "loss": 5.2165,
      "mean_token_accuracy": 0.17189544588327407,
      "num_tokens": 21263460.0,
      "step": 9705
    },
    {
      "entropy": 5.478401517868042,
      "epoch": 0.9636760619293371,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004915891760116123,
      "loss": 5.2808,
      "mean_token_accuracy": 0.1601521283388138,
      "num_tokens": 21274819.0,
      "step": 9710
    },
    {
      "entropy": 5.55060567855835,
      "epoch": 0.9641722905915046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004915795763270626,
      "loss": 5.2376,
      "mean_token_accuracy": 0.1643779307603836,
      "num_tokens": 21286215.0,
      "step": 9715
    },
    {
      "entropy": 5.602538585662842,
      "epoch": 0.9646685192536721,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000491569971271806,
      "loss": 5.2384,
      "mean_token_accuracy": 0.16729076206684113,
      "num_tokens": 21297033.0,
      "step": 9720
    },
    {
      "entropy": 5.5030872344970705,
      "epoch": 0.9651647479158396,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004915603608460809,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16320709586143495,
      "num_tokens": 21308019.0,
      "step": 9725
    },
    {
      "entropy": 5.548954486846924,
      "epoch": 0.9656609765780071,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004915507450501254,
      "loss": 5.3771,
      "mean_token_accuracy": 0.15592159554362298,
      "num_tokens": 21319313.0,
      "step": 9730
    },
    {
      "entropy": 5.5839368343353275,
      "epoch": 0.9661572052401747,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491541123884178,
      "loss": 5.2802,
      "mean_token_accuracy": 0.16490405350923537,
      "num_tokens": 21329950.0,
      "step": 9735
    },
    {
      "entropy": 5.582881450653076,
      "epoch": 0.9666534339023422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004915314973484775,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1669767528772354,
      "num_tokens": 21339061.0,
      "step": 9740
    },
    {
      "entropy": 5.55190224647522,
      "epoch": 0.9671496625645097,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004915218654432624,
      "loss": 5.353,
      "mean_token_accuracy": 0.15825756192207335,
      "num_tokens": 21350114.0,
      "step": 9745
    },
    {
      "entropy": 5.556118535995483,
      "epoch": 0.9676458912266772,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004915122281687716,
      "loss": 5.2662,
      "mean_token_accuracy": 0.1619737282395363,
      "num_tokens": 21361219.0,
      "step": 9750
    },
    {
      "entropy": 5.522767782211304,
      "epoch": 0.9681421198888448,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004915025855252441,
      "loss": 5.2591,
      "mean_token_accuracy": 0.16071735471487045,
      "num_tokens": 21372722.0,
      "step": 9755
    },
    {
      "entropy": 5.58786826133728,
      "epoch": 0.9686383485510123,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004914929375129188,
      "loss": 5.2894,
      "mean_token_accuracy": 0.16373575627803802,
      "num_tokens": 21383350.0,
      "step": 9760
    },
    {
      "entropy": 5.5879720687866214,
      "epoch": 0.9691345772131799,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004914832841320353,
      "loss": 5.3322,
      "mean_token_accuracy": 0.15844108909368515,
      "num_tokens": 21394656.0,
      "step": 9765
    },
    {
      "entropy": 5.6014649868011475,
      "epoch": 0.9696308058753473,
      "grad_norm": 0.82421875,
      "learning_rate": 0.000491473625382833,
      "loss": 5.3271,
      "mean_token_accuracy": 0.15667567849159242,
      "num_tokens": 21406163.0,
      "step": 9770
    },
    {
      "entropy": 5.492336273193359,
      "epoch": 0.9701270345375149,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491463961265551,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16060831397771835,
      "num_tokens": 21418703.0,
      "step": 9775
    },
    {
      "entropy": 5.545598745346069,
      "epoch": 0.9706232631996824,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004914542917804292,
      "loss": 5.2213,
      "mean_token_accuracy": 0.1705381989479065,
      "num_tokens": 21428511.0,
      "step": 9780
    },
    {
      "entropy": 5.582032346725464,
      "epoch": 0.97111949186185,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004914446169277075,
      "loss": 5.2607,
      "mean_token_accuracy": 0.16658746898174287,
      "num_tokens": 21440350.0,
      "step": 9785
    },
    {
      "entropy": 5.6102845668792725,
      "epoch": 0.9716157205240175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004914349367076258,
      "loss": 5.3491,
      "mean_token_accuracy": 0.16187367588281631,
      "num_tokens": 21451262.0,
      "step": 9790
    },
    {
      "entropy": 5.609955644607544,
      "epoch": 0.972111949186185,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004914252511204238,
      "loss": 5.3116,
      "mean_token_accuracy": 0.16031045615673065,
      "num_tokens": 21462587.0,
      "step": 9795
    },
    {
      "entropy": 5.593465280532837,
      "epoch": 0.9726081778483525,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914155601663422,
      "loss": 5.2717,
      "mean_token_accuracy": 0.16374309808015824,
      "num_tokens": 21473452.0,
      "step": 9800
    },
    {
      "entropy": 5.5712298393249515,
      "epoch": 0.97310440651052,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004914058638456208,
      "loss": 5.2355,
      "mean_token_accuracy": 0.16049509048461913,
      "num_tokens": 21484011.0,
      "step": 9805
    },
    {
      "entropy": 5.483835697174072,
      "epoch": 0.9736006351726876,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004913961621585003,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1647476702928543,
      "num_tokens": 21495959.0,
      "step": 9810
    },
    {
      "entropy": 5.511680221557617,
      "epoch": 0.974096863834855,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913864551052214,
      "loss": 5.2716,
      "mean_token_accuracy": 0.16836465895175934,
      "num_tokens": 21507496.0,
      "step": 9815
    },
    {
      "entropy": 5.632143926620484,
      "epoch": 0.9745930924970226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913767426860246,
      "loss": 5.3468,
      "mean_token_accuracy": 0.14913135692477225,
      "num_tokens": 21517532.0,
      "step": 9820
    },
    {
      "entropy": 5.519072532653809,
      "epoch": 0.9750893211591901,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004913670249011508,
      "loss": 5.1725,
      "mean_token_accuracy": 0.16684030443429948,
      "num_tokens": 21527969.0,
      "step": 9825
    },
    {
      "entropy": 5.521410036087036,
      "epoch": 0.9755855498213577,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004913573017508411,
      "loss": 5.2886,
      "mean_token_accuracy": 0.15797928869724273,
      "num_tokens": 21538243.0,
      "step": 9830
    },
    {
      "entropy": 5.570577001571655,
      "epoch": 0.9760817784835252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004913475732353365,
      "loss": 5.2814,
      "mean_token_accuracy": 0.16380224004387856,
      "num_tokens": 21548702.0,
      "step": 9835
    },
    {
      "entropy": 5.522900390625,
      "epoch": 0.9765780071456928,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913378393548784,
      "loss": 5.2546,
      "mean_token_accuracy": 0.16014898121356963,
      "num_tokens": 21558583.0,
      "step": 9840
    },
    {
      "entropy": 5.535926246643067,
      "epoch": 0.9770742358078602,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004913281001097078,
      "loss": 5.2274,
      "mean_token_accuracy": 0.16474588811397553,
      "num_tokens": 21569845.0,
      "step": 9845
    },
    {
      "entropy": 5.619068956375122,
      "epoch": 0.9775704644700278,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004913183555000667,
      "loss": 5.3743,
      "mean_token_accuracy": 0.156708575040102,
      "num_tokens": 21580540.0,
      "step": 9850
    },
    {
      "entropy": 5.605562973022461,
      "epoch": 0.9780666931321953,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913086055261963,
      "loss": 5.2408,
      "mean_token_accuracy": 0.16236970573663712,
      "num_tokens": 21590982.0,
      "step": 9855
    },
    {
      "entropy": 5.551388835906982,
      "epoch": 0.9785629217943629,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004912988501883387,
      "loss": 5.2743,
      "mean_token_accuracy": 0.16005310267210007,
      "num_tokens": 21602817.0,
      "step": 9860
    },
    {
      "entropy": 5.548692750930786,
      "epoch": 0.9790591504565304,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912890894867357,
      "loss": 5.3146,
      "mean_token_accuracy": 0.1607313111424446,
      "num_tokens": 21613422.0,
      "step": 9865
    },
    {
      "entropy": 5.548087215423584,
      "epoch": 0.9795553791186979,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004912793234216293,
      "loss": 5.3022,
      "mean_token_accuracy": 0.15625362619757652,
      "num_tokens": 21624241.0,
      "step": 9870
    },
    {
      "entropy": 5.489270830154419,
      "epoch": 0.9800516077808654,
      "grad_norm": 0.875,
      "learning_rate": 0.0004912695519932618,
      "loss": 5.2902,
      "mean_token_accuracy": 0.16378703266382216,
      "num_tokens": 21636243.0,
      "step": 9875
    },
    {
      "entropy": 5.558585739135742,
      "epoch": 0.9805478364430329,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004912597752018753,
      "loss": 5.3186,
      "mean_token_accuracy": 0.15876202583312987,
      "num_tokens": 21647918.0,
      "step": 9880
    },
    {
      "entropy": 5.617195272445679,
      "epoch": 0.9810440651052005,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912499930477126,
      "loss": 5.3007,
      "mean_token_accuracy": 0.15920585244894028,
      "num_tokens": 21659236.0,
      "step": 9885
    },
    {
      "entropy": 5.555436801910401,
      "epoch": 0.981540293767368,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004912402055310158,
      "loss": 5.2918,
      "mean_token_accuracy": 0.16367138922214508,
      "num_tokens": 21669647.0,
      "step": 9890
    },
    {
      "entropy": 5.531766033172607,
      "epoch": 0.9820365224295355,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004912304126520279,
      "loss": 5.3324,
      "mean_token_accuracy": 0.15495423674583436,
      "num_tokens": 21680369.0,
      "step": 9895
    },
    {
      "entropy": 5.6336465835571286,
      "epoch": 0.982532751091703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004912206144109917,
      "loss": 5.3452,
      "mean_token_accuracy": 0.161551433801651,
      "num_tokens": 21690942.0,
      "step": 9900
    },
    {
      "entropy": 5.571320104598999,
      "epoch": 0.9830289797538706,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912108108081503,
      "loss": 5.2708,
      "mean_token_accuracy": 0.16247555911540984,
      "num_tokens": 21701945.0,
      "step": 9905
    },
    {
      "entropy": 5.581538391113281,
      "epoch": 0.9835252084160381,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004912010018437465,
      "loss": 5.2917,
      "mean_token_accuracy": 0.16236080676317216,
      "num_tokens": 21712246.0,
      "step": 9910
    },
    {
      "entropy": 5.656294345855713,
      "epoch": 0.9840214370782057,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000491191187518024,
      "loss": 5.2988,
      "mean_token_accuracy": 0.16024024784564972,
      "num_tokens": 21721660.0,
      "step": 9915
    },
    {
      "entropy": 5.5558836460113525,
      "epoch": 0.9845176657403731,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004911813678312257,
      "loss": 5.1875,
      "mean_token_accuracy": 0.1651577889919281,
      "num_tokens": 21732157.0,
      "step": 9920
    },
    {
      "entropy": 5.549558639526367,
      "epoch": 0.9850138944025407,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911715427835953,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1567473106086254,
      "num_tokens": 21741727.0,
      "step": 9925
    },
    {
      "entropy": 5.5697094917297365,
      "epoch": 0.9855101230647082,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911617123753765,
      "loss": 5.2436,
      "mean_token_accuracy": 0.16794258058071138,
      "num_tokens": 21752596.0,
      "step": 9930
    },
    {
      "entropy": 5.509955167770386,
      "epoch": 0.9860063517268758,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004911518766068131,
      "loss": 5.1658,
      "mean_token_accuracy": 0.16699011474847794,
      "num_tokens": 21763978.0,
      "step": 9935
    },
    {
      "entropy": 5.561686944961548,
      "epoch": 0.9865025803890433,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004911420354781487,
      "loss": 5.3249,
      "mean_token_accuracy": 0.1500205598771572,
      "num_tokens": 21774961.0,
      "step": 9940
    },
    {
      "entropy": 5.572843313217163,
      "epoch": 0.9869988090512108,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004911321889896278,
      "loss": 5.2263,
      "mean_token_accuracy": 0.1687939465045929,
      "num_tokens": 21785387.0,
      "step": 9945
    },
    {
      "entropy": 5.553194618225097,
      "epoch": 0.9874950377133783,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911223371414943,
      "loss": 5.2808,
      "mean_token_accuracy": 0.16930504888296127,
      "num_tokens": 21795982.0,
      "step": 9950
    },
    {
      "entropy": 5.594301080703735,
      "epoch": 0.9879912663755459,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004911124799339924,
      "loss": 5.3108,
      "mean_token_accuracy": 0.15826882869005204,
      "num_tokens": 21806768.0,
      "step": 9955
    },
    {
      "entropy": 5.570723295211792,
      "epoch": 0.9884874950377134,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004911026173673667,
      "loss": 5.2743,
      "mean_token_accuracy": 0.1663658618927002,
      "num_tokens": 21818563.0,
      "step": 9960
    },
    {
      "entropy": 5.592010068893432,
      "epoch": 0.9889837236998809,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004910927494418619,
      "loss": 5.3402,
      "mean_token_accuracy": 0.16357164829969406,
      "num_tokens": 21830253.0,
      "step": 9965
    },
    {
      "entropy": 5.6081441879272464,
      "epoch": 0.9894799523620484,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004910828761577224,
      "loss": 5.3201,
      "mean_token_accuracy": 0.16309469789266587,
      "num_tokens": 21842353.0,
      "step": 9970
    },
    {
      "entropy": 5.541133975982666,
      "epoch": 0.9899761810242159,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004910729975151933,
      "loss": 5.2678,
      "mean_token_accuracy": 0.16653788536787034,
      "num_tokens": 21852066.0,
      "step": 9975
    },
    {
      "entropy": 5.521096467971802,
      "epoch": 0.9904724096863835,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004910631135145194,
      "loss": 5.2859,
      "mean_token_accuracy": 0.16597900539636612,
      "num_tokens": 21862754.0,
      "step": 9980
    },
    {
      "entropy": 5.479059743881225,
      "epoch": 0.990968638348551,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004910532241559458,
      "loss": 5.212,
      "mean_token_accuracy": 0.17282212823629378,
      "num_tokens": 21873531.0,
      "step": 9985
    },
    {
      "entropy": 5.468411636352539,
      "epoch": 0.9914648670107186,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004910433294397179,
      "loss": 5.1495,
      "mean_token_accuracy": 0.17061155438423156,
      "num_tokens": 21884339.0,
      "step": 9990
    },
    {
      "entropy": 5.5224603652954105,
      "epoch": 0.991961095672886,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000491033429366081,
      "loss": 5.268,
      "mean_token_accuracy": 0.16112745255231858,
      "num_tokens": 21895489.0,
      "step": 9995
    },
    {
      "entropy": 5.530955505371094,
      "epoch": 0.9924573243350536,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004910235239352804,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16834522485733033,
      "num_tokens": 21906832.0,
      "step": 10000
    },
    {
      "entropy": 5.559308671951294,
      "epoch": 0.9929535529972211,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000491013613147562,
      "loss": 5.281,
      "mean_token_accuracy": 0.1642713338136673,
      "num_tokens": 21916979.0,
      "step": 10005
    },
    {
      "entropy": 5.5602928638458256,
      "epoch": 0.9934497816593887,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004910036970031715,
      "loss": 5.1908,
      "mean_token_accuracy": 0.1753999412059784,
      "num_tokens": 21927352.0,
      "step": 10010
    },
    {
      "entropy": 5.553837537765503,
      "epoch": 0.9939460103215562,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004909937755023547,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1587557539343834,
      "num_tokens": 21937765.0,
      "step": 10015
    },
    {
      "entropy": 5.489664459228516,
      "epoch": 0.9944422389837237,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004909838486453577,
      "loss": 5.1993,
      "mean_token_accuracy": 0.16981570273637772,
      "num_tokens": 21948659.0,
      "step": 10020
    },
    {
      "entropy": 5.541535758972168,
      "epoch": 0.9949384676458912,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909739164324267,
      "loss": 5.3208,
      "mean_token_accuracy": 0.15923656076192855,
      "num_tokens": 21959356.0,
      "step": 10025
    },
    {
      "entropy": 5.659869861602783,
      "epoch": 0.9954346963080588,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004909639788638079,
      "loss": 5.4365,
      "mean_token_accuracy": 0.15047643035650254,
      "num_tokens": 21971321.0,
      "step": 10030
    },
    {
      "entropy": 5.560820531845093,
      "epoch": 0.9959309249702263,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004909540359397477,
      "loss": 5.2748,
      "mean_token_accuracy": 0.16066478043794633,
      "num_tokens": 21981252.0,
      "step": 10035
    },
    {
      "entropy": 5.607734823226929,
      "epoch": 0.9964271536323938,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004909440876604929,
      "loss": 5.2982,
      "mean_token_accuracy": 0.1596674531698227,
      "num_tokens": 21991820.0,
      "step": 10040
    },
    {
      "entropy": 5.562358331680298,
      "epoch": 0.9969233822945613,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00049093413402629,
      "loss": 5.2519,
      "mean_token_accuracy": 0.16246361434459686,
      "num_tokens": 22003619.0,
      "step": 10045
    },
    {
      "entropy": 5.532558870315552,
      "epoch": 0.9974196109567288,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909241750373858,
      "loss": 5.3056,
      "mean_token_accuracy": 0.1582287535071373,
      "num_tokens": 22016116.0,
      "step": 10050
    },
    {
      "entropy": 5.6082922458648685,
      "epoch": 0.9979158396188964,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004909142106940275,
      "loss": 5.2334,
      "mean_token_accuracy": 0.16466933190822602,
      "num_tokens": 22026115.0,
      "step": 10055
    },
    {
      "entropy": 5.59378490447998,
      "epoch": 0.9984120682810639,
      "grad_norm": 0.9375,
      "learning_rate": 0.000490904240996462,
      "loss": 5.313,
      "mean_token_accuracy": 0.1642625153064728,
      "num_tokens": 22038889.0,
      "step": 10060
    },
    {
      "entropy": 5.53775429725647,
      "epoch": 0.9989082969432315,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004908942659449364,
      "loss": 5.2703,
      "mean_token_accuracy": 0.16213991045951842,
      "num_tokens": 22050773.0,
      "step": 10065
    },
    {
      "entropy": 5.528830528259277,
      "epoch": 0.9994045256053989,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908842855396982,
      "loss": 5.275,
      "mean_token_accuracy": 0.16044630110263824,
      "num_tokens": 22062869.0,
      "step": 10070
    },
    {
      "entropy": 5.548056411743164,
      "epoch": 0.9999007542675665,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004908742997809951,
      "loss": 5.2895,
      "mean_token_accuracy": 0.1627122640609741,
      "num_tokens": 22073898.0,
      "step": 10075
    },
    {
      "entropy": 5.615260791778565,
      "epoch": 1.0003969829297341,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908643086690744,
      "loss": 5.2034,
      "mean_token_accuracy": 0.16319991126656533,
      "num_tokens": 22084383.0,
      "step": 10080
    },
    {
      "entropy": 5.5695881843566895,
      "epoch": 1.0008932115919016,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490854312204184,
      "loss": 5.233,
      "mean_token_accuracy": 0.16529347896575927,
      "num_tokens": 22094990.0,
      "step": 10085
    },
    {
      "entropy": 5.560304737091064,
      "epoch": 1.001389440254069,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004908443103865719,
      "loss": 5.2643,
      "mean_token_accuracy": 0.15825624465942384,
      "num_tokens": 22105259.0,
      "step": 10090
    },
    {
      "entropy": 5.536929941177368,
      "epoch": 1.0018856689162365,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908343032164859,
      "loss": 5.2285,
      "mean_token_accuracy": 0.16413161307573318,
      "num_tokens": 22116348.0,
      "step": 10095
    },
    {
      "entropy": 5.5739833354949955,
      "epoch": 1.0023818975784042,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004908242906941743,
      "loss": 5.2227,
      "mean_token_accuracy": 0.16226010173559188,
      "num_tokens": 22127311.0,
      "step": 10100
    },
    {
      "entropy": 5.6037389755249025,
      "epoch": 1.0028781262405717,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004908142728198853,
      "loss": 5.2516,
      "mean_token_accuracy": 0.1596440225839615,
      "num_tokens": 22137929.0,
      "step": 10105
    },
    {
      "entropy": 5.666545581817627,
      "epoch": 1.0033743549027392,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004908042495938673,
      "loss": 5.3448,
      "mean_token_accuracy": 0.16056488007307052,
      "num_tokens": 22148202.0,
      "step": 10110
    },
    {
      "entropy": 5.541648912429809,
      "epoch": 1.0038705835649067,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907942210163692,
      "loss": 5.1454,
      "mean_token_accuracy": 0.16704110503196717,
      "num_tokens": 22158955.0,
      "step": 10115
    },
    {
      "entropy": 5.5215277671813965,
      "epoch": 1.0043668122270741,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004907841870876392,
      "loss": 5.1776,
      "mean_token_accuracy": 0.163038769364357,
      "num_tokens": 22169569.0,
      "step": 10120
    },
    {
      "entropy": 5.531347036361694,
      "epoch": 1.0048630408892418,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004907741478079263,
      "loss": 5.158,
      "mean_token_accuracy": 0.16266010999679564,
      "num_tokens": 22180171.0,
      "step": 10125
    },
    {
      "entropy": 5.5394816398620605,
      "epoch": 1.0053592695514093,
      "grad_norm": 1.0,
      "learning_rate": 0.0004907641031774796,
      "loss": 5.2857,
      "mean_token_accuracy": 0.16013559252023696,
      "num_tokens": 22190531.0,
      "step": 10130
    },
    {
      "entropy": 5.5429911613464355,
      "epoch": 1.0058554982135768,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000490754053196548,
      "loss": 5.1739,
      "mean_token_accuracy": 0.16688481122255325,
      "num_tokens": 22201495.0,
      "step": 10135
    },
    {
      "entropy": 5.6280182838439945,
      "epoch": 1.0063517268757443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907439978653809,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1566721737384796,
      "num_tokens": 22212941.0,
      "step": 10140
    },
    {
      "entropy": 5.62884111404419,
      "epoch": 1.006847955537912,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004907339371842275,
      "loss": 5.2829,
      "mean_token_accuracy": 0.15863350704312323,
      "num_tokens": 22225212.0,
      "step": 10145
    },
    {
      "entropy": 5.547158908843994,
      "epoch": 1.0073441842000794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907238711533373,
      "loss": 5.2332,
      "mean_token_accuracy": 0.16370310634374619,
      "num_tokens": 22236370.0,
      "step": 10150
    },
    {
      "entropy": 5.5607482433319095,
      "epoch": 1.007840412862247,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004907137997729599,
      "loss": 5.2067,
      "mean_token_accuracy": 0.16858701258897782,
      "num_tokens": 22247641.0,
      "step": 10155
    },
    {
      "entropy": 5.507538747787476,
      "epoch": 1.0083366415244144,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000490703723043345,
      "loss": 5.25,
      "mean_token_accuracy": 0.16729840487241746,
      "num_tokens": 22258599.0,
      "step": 10160
    },
    {
      "entropy": 5.519358587265015,
      "epoch": 1.008832870186582,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004906936409647428,
      "loss": 5.1757,
      "mean_token_accuracy": 0.1641075298190117,
      "num_tokens": 22270913.0,
      "step": 10165
    },
    {
      "entropy": 5.5304711818695065,
      "epoch": 1.0093290988487496,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000490683553537403,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17113254368305206,
      "num_tokens": 22281053.0,
      "step": 10170
    },
    {
      "entropy": 5.534640741348267,
      "epoch": 1.009825327510917,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004906734607615758,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1695442095398903,
      "num_tokens": 22292570.0,
      "step": 10175
    },
    {
      "entropy": 5.519126844406128,
      "epoch": 1.0103215561730845,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004906633626375114,
      "loss": 5.1926,
      "mean_token_accuracy": 0.1635767087340355,
      "num_tokens": 22303548.0,
      "step": 10180
    },
    {
      "entropy": 5.5485076904296875,
      "epoch": 1.010817784835252,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906532591654604,
      "loss": 5.2168,
      "mean_token_accuracy": 0.1707324728369713,
      "num_tokens": 22314133.0,
      "step": 10185
    },
    {
      "entropy": 5.592421436309815,
      "epoch": 1.0113140134974197,
      "grad_norm": 0.7734375,
      "learning_rate": 0.0004906431503456733,
      "loss": 5.1789,
      "mean_token_accuracy": 0.16579480618238449,
      "num_tokens": 22326431.0,
      "step": 10190
    },
    {
      "entropy": 5.549811124801636,
      "epoch": 1.0118102421595871,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004906330361784007,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16449104696512223,
      "num_tokens": 22338507.0,
      "step": 10195
    },
    {
      "entropy": 5.5582836151123045,
      "epoch": 1.0123064708217546,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004906229166638936,
      "loss": 5.196,
      "mean_token_accuracy": 0.1639328896999359,
      "num_tokens": 22348863.0,
      "step": 10200
    },
    {
      "entropy": 5.5030683994293215,
      "epoch": 1.012802699483922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004906127918024026,
      "loss": 5.146,
      "mean_token_accuracy": 0.16409398019313812,
      "num_tokens": 22359000.0,
      "step": 10205
    },
    {
      "entropy": 5.6267993450164795,
      "epoch": 1.0132989281460898,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004906026615941791,
      "loss": 5.2851,
      "mean_token_accuracy": 0.15941507518291473,
      "num_tokens": 22369539.0,
      "step": 10210
    },
    {
      "entropy": 5.585826444625854,
      "epoch": 1.0137951568082573,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004905925260394741,
      "loss": 5.2479,
      "mean_token_accuracy": 0.16477965861558913,
      "num_tokens": 22380700.0,
      "step": 10215
    },
    {
      "entropy": 5.52986421585083,
      "epoch": 1.0142913854704247,
      "grad_norm": 0.859375,
      "learning_rate": 0.000490582385138539,
      "loss": 5.2229,
      "mean_token_accuracy": 0.15874820500612258,
      "num_tokens": 22392151.0,
      "step": 10220
    },
    {
      "entropy": 5.528717041015625,
      "epoch": 1.0147876141325922,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905722388916254,
      "loss": 5.194,
      "mean_token_accuracy": 0.17224560230970382,
      "num_tokens": 22401735.0,
      "step": 10225
    },
    {
      "entropy": 5.5665429592132565,
      "epoch": 1.01528384279476,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004905620872989846,
      "loss": 5.2467,
      "mean_token_accuracy": 0.1640242740511894,
      "num_tokens": 22412549.0,
      "step": 10230
    },
    {
      "entropy": 5.6074787139892575,
      "epoch": 1.0157800714569274,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004905519303608687,
      "loss": 5.2359,
      "mean_token_accuracy": 0.1627390593290329,
      "num_tokens": 22424728.0,
      "step": 10235
    },
    {
      "entropy": 5.494732522964478,
      "epoch": 1.0162763001190949,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004905417680775294,
      "loss": 5.1251,
      "mean_token_accuracy": 0.16549260169267654,
      "num_tokens": 22434197.0,
      "step": 10240
    },
    {
      "entropy": 5.625267219543457,
      "epoch": 1.0167725287812623,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004905316004492186,
      "loss": 5.3174,
      "mean_token_accuracy": 0.15682656317949295,
      "num_tokens": 22444671.0,
      "step": 10245
    },
    {
      "entropy": 5.525687885284424,
      "epoch": 1.01726875744343,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004905214274761885,
      "loss": 5.1862,
      "mean_token_accuracy": 0.15919533520936965,
      "num_tokens": 22456636.0,
      "step": 10250
    },
    {
      "entropy": 5.500757169723511,
      "epoch": 1.0177649861055975,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004905112491586915,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1714894562959671,
      "num_tokens": 22467011.0,
      "step": 10255
    },
    {
      "entropy": 5.537000465393066,
      "epoch": 1.018261214767765,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049050106549698,
      "loss": 5.2139,
      "mean_token_accuracy": 0.16324659287929535,
      "num_tokens": 22477630.0,
      "step": 10260
    },
    {
      "entropy": 5.516261577606201,
      "epoch": 1.0187574434299325,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004904908764913063,
      "loss": 5.1642,
      "mean_token_accuracy": 0.17469357401132585,
      "num_tokens": 22488132.0,
      "step": 10265
    },
    {
      "entropy": 5.508202934265137,
      "epoch": 1.0192536720921,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004904806821419233,
      "loss": 5.2282,
      "mean_token_accuracy": 0.16461318582296372,
      "num_tokens": 22499430.0,
      "step": 10270
    },
    {
      "entropy": 5.518159198760986,
      "epoch": 1.0197499007542676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904704824490835,
      "loss": 5.1462,
      "mean_token_accuracy": 0.16928788870573044,
      "num_tokens": 22509728.0,
      "step": 10275
    },
    {
      "entropy": 5.549171733856201,
      "epoch": 1.020246129416435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904602774130402,
      "loss": 5.2573,
      "mean_token_accuracy": 0.15980387330055237,
      "num_tokens": 22520955.0,
      "step": 10280
    },
    {
      "entropy": 5.624895429611206,
      "epoch": 1.0207423580786026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004904500670340463,
      "loss": 5.2677,
      "mean_token_accuracy": 0.1583781734108925,
      "num_tokens": 22530440.0,
      "step": 10285
    },
    {
      "entropy": 5.568074560165405,
      "epoch": 1.02123858674077,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000490439851312355,
      "loss": 5.2442,
      "mean_token_accuracy": 0.17001541554927826,
      "num_tokens": 22541539.0,
      "step": 10290
    },
    {
      "entropy": 5.53445873260498,
      "epoch": 1.0217348154029378,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004904296302482196,
      "loss": 5.2335,
      "mean_token_accuracy": 0.16549630910158158,
      "num_tokens": 22555717.0,
      "step": 10295
    },
    {
      "entropy": 5.632543134689331,
      "epoch": 1.0222310440651052,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004904194038418935,
      "loss": 5.3194,
      "mean_token_accuracy": 0.1577071651816368,
      "num_tokens": 22566938.0,
      "step": 10300
    },
    {
      "entropy": 5.645268964767456,
      "epoch": 1.0227272727272727,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004904091720936305,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16342383325099946,
      "num_tokens": 22578725.0,
      "step": 10305
    },
    {
      "entropy": 5.549882221221924,
      "epoch": 1.0232235013894402,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004903989350036841,
      "loss": 5.1151,
      "mean_token_accuracy": 0.17128328680992128,
      "num_tokens": 22588476.0,
      "step": 10310
    },
    {
      "entropy": 5.582836151123047,
      "epoch": 1.0237197300516079,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004903886925723083,
      "loss": 5.3352,
      "mean_token_accuracy": 0.1550891861319542,
      "num_tokens": 22600294.0,
      "step": 10315
    },
    {
      "entropy": 5.600665187835693,
      "epoch": 1.0242159587137754,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000490378444799757,
      "loss": 5.2121,
      "mean_token_accuracy": 0.16296615600585937,
      "num_tokens": 22610578.0,
      "step": 10320
    },
    {
      "entropy": 5.602261161804199,
      "epoch": 1.0247121873759428,
      "grad_norm": 1.0,
      "learning_rate": 0.0004903681916862844,
      "loss": 5.223,
      "mean_token_accuracy": 0.16472982317209245,
      "num_tokens": 22620988.0,
      "step": 10325
    },
    {
      "entropy": 5.543428373336792,
      "epoch": 1.0252084160381103,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004903579332321447,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1660815879702568,
      "num_tokens": 22632198.0,
      "step": 10330
    },
    {
      "entropy": 5.564525556564331,
      "epoch": 1.0257046447002778,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004903476694375924,
      "loss": 5.2143,
      "mean_token_accuracy": 0.17281053066253663,
      "num_tokens": 22642543.0,
      "step": 10335
    },
    {
      "entropy": 5.594615650177002,
      "epoch": 1.0262008733624455,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004903374003028819,
      "loss": 5.2102,
      "mean_token_accuracy": 0.16698787808418275,
      "num_tokens": 22655453.0,
      "step": 10340
    },
    {
      "entropy": 5.508903694152832,
      "epoch": 1.026697102024613,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903271258282678,
      "loss": 5.1481,
      "mean_token_accuracy": 0.17198095917701722,
      "num_tokens": 22667342.0,
      "step": 10345
    },
    {
      "entropy": 5.60483021736145,
      "epoch": 1.0271933306867804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004903168460140049,
      "loss": 5.2304,
      "mean_token_accuracy": 0.16972459405660628,
      "num_tokens": 22676890.0,
      "step": 10350
    },
    {
      "entropy": 5.556510353088379,
      "epoch": 1.027689559348948,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004903065608603484,
      "loss": 5.2264,
      "mean_token_accuracy": 0.16433410793542863,
      "num_tokens": 22687809.0,
      "step": 10355
    },
    {
      "entropy": 5.5336202621459964,
      "epoch": 1.0281857880111156,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490296270367553,
      "loss": 5.2362,
      "mean_token_accuracy": 0.16485975682735443,
      "num_tokens": 22697938.0,
      "step": 10360
    },
    {
      "entropy": 5.6364585876464846,
      "epoch": 1.028682016673283,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004902859745358741,
      "loss": 5.2284,
      "mean_token_accuracy": 0.16909437477588654,
      "num_tokens": 22708303.0,
      "step": 10365
    },
    {
      "entropy": 5.527675485610962,
      "epoch": 1.0291782453354505,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902756733655668,
      "loss": 5.1674,
      "mean_token_accuracy": 0.16882594674825668,
      "num_tokens": 22717987.0,
      "step": 10370
    },
    {
      "entropy": 5.5493913173675535,
      "epoch": 1.029674473997618,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004902653668568868,
      "loss": 5.243,
      "mean_token_accuracy": 0.1632582202553749,
      "num_tokens": 22729672.0,
      "step": 10375
    },
    {
      "entropy": 5.605344200134278,
      "epoch": 1.0301707026597857,
      "grad_norm": 0.875,
      "learning_rate": 0.0004902550550100896,
      "loss": 5.1981,
      "mean_token_accuracy": 0.1698942244052887,
      "num_tokens": 22741296.0,
      "step": 10380
    },
    {
      "entropy": 5.536325740814209,
      "epoch": 1.0306669313219532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004902447378254307,
      "loss": 5.2022,
      "mean_token_accuracy": 0.16885661333799362,
      "num_tokens": 22752337.0,
      "step": 10385
    },
    {
      "entropy": 5.570239925384522,
      "epoch": 1.0311631599841207,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004902344153031663,
      "loss": 5.2329,
      "mean_token_accuracy": 0.15916235148906707,
      "num_tokens": 22762681.0,
      "step": 10390
    },
    {
      "entropy": 5.623147439956665,
      "epoch": 1.0316593886462881,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000490224087443552,
      "loss": 5.2833,
      "mean_token_accuracy": 0.16048735976219178,
      "num_tokens": 22774275.0,
      "step": 10395
    },
    {
      "entropy": 5.532047510147095,
      "epoch": 1.0321556173084558,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902137542468442,
      "loss": 5.1514,
      "mean_token_accuracy": 0.17046201080083848,
      "num_tokens": 22785944.0,
      "step": 10400
    },
    {
      "entropy": 5.480773401260376,
      "epoch": 1.0326518459706233,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004902034157132991,
      "loss": 5.08,
      "mean_token_accuracy": 0.1674461007118225,
      "num_tokens": 22796543.0,
      "step": 10405
    },
    {
      "entropy": 5.4836280822753904,
      "epoch": 1.0331480746327908,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004901930718431729,
      "loss": 5.1898,
      "mean_token_accuracy": 0.16869719326496124,
      "num_tokens": 22807957.0,
      "step": 10410
    },
    {
      "entropy": 5.584758138656616,
      "epoch": 1.0336443032949583,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004901827226367223,
      "loss": 5.3017,
      "mean_token_accuracy": 0.16239845752716064,
      "num_tokens": 22818903.0,
      "step": 10415
    },
    {
      "entropy": 5.589465665817261,
      "epoch": 1.0341405319571257,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004901723680942038,
      "loss": 5.3319,
      "mean_token_accuracy": 0.15857019722461702,
      "num_tokens": 22829286.0,
      "step": 10420
    },
    {
      "entropy": 5.635228872299194,
      "epoch": 1.0346367606192934,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004901620082158742,
      "loss": 5.2261,
      "mean_token_accuracy": 0.16329090893268586,
      "num_tokens": 22839860.0,
      "step": 10425
    },
    {
      "entropy": 5.6458680629730225,
      "epoch": 1.035132989281461,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004901516430019905,
      "loss": 5.278,
      "mean_token_accuracy": 0.15597940385341644,
      "num_tokens": 22852238.0,
      "step": 10430
    },
    {
      "entropy": 5.584874439239502,
      "epoch": 1.0356292179436284,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004901412724528097,
      "loss": 5.248,
      "mean_token_accuracy": 0.16430449038743972,
      "num_tokens": 22863013.0,
      "step": 10435
    },
    {
      "entropy": 5.524501657485962,
      "epoch": 1.0361254466057959,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004901308965685889,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17041504383087158,
      "num_tokens": 22874233.0,
      "step": 10440
    },
    {
      "entropy": 5.561395359039307,
      "epoch": 1.0366216752679636,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004901205153495853,
      "loss": 5.2871,
      "mean_token_accuracy": 0.16226013302803038,
      "num_tokens": 22885313.0,
      "step": 10445
    },
    {
      "entropy": 5.545291185379028,
      "epoch": 1.037117903930131,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901101287960566,
      "loss": 5.1727,
      "mean_token_accuracy": 0.16494603753089904,
      "num_tokens": 22895365.0,
      "step": 10450
    },
    {
      "entropy": 5.531405639648438,
      "epoch": 1.0376141325922985,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900997369082602,
      "loss": 5.16,
      "mean_token_accuracy": 0.16615414768457412,
      "num_tokens": 22906392.0,
      "step": 10455
    },
    {
      "entropy": 5.478501319885254,
      "epoch": 1.038110361254466,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004900893396864539,
      "loss": 5.1259,
      "mean_token_accuracy": 0.1732056587934494,
      "num_tokens": 22916760.0,
      "step": 10460
    },
    {
      "entropy": 5.580978107452393,
      "epoch": 1.0386065899166337,
      "grad_norm": 0.875,
      "learning_rate": 0.0004900789371308953,
      "loss": 5.1892,
      "mean_token_accuracy": 0.16418147087097168,
      "num_tokens": 22928070.0,
      "step": 10465
    },
    {
      "entropy": 5.473230028152466,
      "epoch": 1.0391028185788012,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004900685292418425,
      "loss": 5.1668,
      "mean_token_accuracy": 0.16596991568803787,
      "num_tokens": 22938815.0,
      "step": 10470
    },
    {
      "entropy": 5.531615161895752,
      "epoch": 1.0395990472409686,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004900581160195537,
      "loss": 5.2375,
      "mean_token_accuracy": 0.15803822949528695,
      "num_tokens": 22949806.0,
      "step": 10475
    },
    {
      "entropy": 5.492841625213623,
      "epoch": 1.040095275903136,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490047697464287,
      "loss": 5.0737,
      "mean_token_accuracy": 0.16634062081575393,
      "num_tokens": 22960880.0,
      "step": 10480
    },
    {
      "entropy": 5.579438924789429,
      "epoch": 1.0405915045653038,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004900372735763009,
      "loss": 5.2289,
      "mean_token_accuracy": 0.16133745461702348,
      "num_tokens": 22972828.0,
      "step": 10485
    },
    {
      "entropy": 5.470828247070313,
      "epoch": 1.0410877332274713,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900268443558536,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1696924239397049,
      "num_tokens": 22981869.0,
      "step": 10490
    },
    {
      "entropy": 5.577334022521972,
      "epoch": 1.0415839618896388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000490016409803204,
      "loss": 5.243,
      "mean_token_accuracy": 0.16079034209251403,
      "num_tokens": 22993111.0,
      "step": 10495
    },
    {
      "entropy": 5.529567432403565,
      "epoch": 1.0420801905518062,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004900059699186108,
      "loss": 5.2202,
      "mean_token_accuracy": 0.16344015449285507,
      "num_tokens": 23003115.0,
      "step": 10500
    },
    {
      "entropy": 5.619073581695557,
      "epoch": 1.0425764192139737,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004899955247023328,
      "loss": 5.2681,
      "mean_token_accuracy": 0.16177919059991835,
      "num_tokens": 23012587.0,
      "step": 10505
    },
    {
      "entropy": 5.58408145904541,
      "epoch": 1.0430726478761414,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000489985074154629,
      "loss": 5.2665,
      "mean_token_accuracy": 0.1609119638800621,
      "num_tokens": 23024455.0,
      "step": 10510
    },
    {
      "entropy": 5.495818042755127,
      "epoch": 1.0435688765383089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004899746182757587,
      "loss": 5.1985,
      "mean_token_accuracy": 0.16518384218215942,
      "num_tokens": 23034430.0,
      "step": 10515
    },
    {
      "entropy": 5.572316694259643,
      "epoch": 1.0440651052004764,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004899641570659811,
      "loss": 5.2692,
      "mean_token_accuracy": 0.16436239555478097,
      "num_tokens": 23044811.0,
      "step": 10520
    },
    {
      "entropy": 5.611488246917725,
      "epoch": 1.0445613338626438,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004899536905255555,
      "loss": 5.2945,
      "mean_token_accuracy": 0.1609091192483902,
      "num_tokens": 23056179.0,
      "step": 10525
    },
    {
      "entropy": 5.579851865768433,
      "epoch": 1.0450575625248115,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004899432186547416,
      "loss": 5.2208,
      "mean_token_accuracy": 0.16762323379516603,
      "num_tokens": 23068331.0,
      "step": 10530
    },
    {
      "entropy": 5.548413801193237,
      "epoch": 1.045553791186979,
      "grad_norm": 1.0,
      "learning_rate": 0.0004899327414537991,
      "loss": 5.2101,
      "mean_token_accuracy": 0.16025710552930833,
      "num_tokens": 23079213.0,
      "step": 10535
    },
    {
      "entropy": 5.646221446990967,
      "epoch": 1.0460500198491465,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004899222589229877,
      "loss": 5.3083,
      "mean_token_accuracy": 0.15819277614355087,
      "num_tokens": 23089860.0,
      "step": 10540
    },
    {
      "entropy": 5.492652702331543,
      "epoch": 1.046546248511314,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004899117710625674,
      "loss": 5.1736,
      "mean_token_accuracy": 0.17650664001703262,
      "num_tokens": 23099431.0,
      "step": 10545
    },
    {
      "entropy": 5.527903509140015,
      "epoch": 1.0470424771734816,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004899012778727983,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1618702933192253,
      "num_tokens": 23110947.0,
      "step": 10550
    },
    {
      "entropy": 5.60500807762146,
      "epoch": 1.0475387058356491,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004898907793539406,
      "loss": 5.3365,
      "mean_token_accuracy": 0.16245901733636856,
      "num_tokens": 23122782.0,
      "step": 10555
    },
    {
      "entropy": 5.661748695373535,
      "epoch": 1.0480349344978166,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004898802755062546,
      "loss": 5.2339,
      "mean_token_accuracy": 0.16416132003068923,
      "num_tokens": 23134254.0,
      "step": 10560
    },
    {
      "entropy": 5.492115449905396,
      "epoch": 1.048531163159984,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004898697663300007,
      "loss": 5.0711,
      "mean_token_accuracy": 0.17328284531831742,
      "num_tokens": 23143851.0,
      "step": 10565
    },
    {
      "entropy": 5.507873058319092,
      "epoch": 1.0490273918221515,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004898592518254398,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16552965939044953,
      "num_tokens": 23155966.0,
      "step": 10570
    },
    {
      "entropy": 5.527257966995239,
      "epoch": 1.0495236204843192,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004898487319928322,
      "loss": 5.1976,
      "mean_token_accuracy": 0.15939348489046096,
      "num_tokens": 23165896.0,
      "step": 10575
    },
    {
      "entropy": 5.509977102279663,
      "epoch": 1.0500198491464867,
      "grad_norm": 0.796875,
      "learning_rate": 0.0004898382068324392,
      "loss": 5.194,
      "mean_token_accuracy": 0.1655638724565506,
      "num_tokens": 23177821.0,
      "step": 10580
    },
    {
      "entropy": 5.578458070755005,
      "epoch": 1.0505160778086542,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004898276763445215,
      "loss": 5.2494,
      "mean_token_accuracy": 0.1635458916425705,
      "num_tokens": 23188916.0,
      "step": 10585
    },
    {
      "entropy": 5.641046094894409,
      "epoch": 1.0510123064708217,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004898171405293404,
      "loss": 5.2545,
      "mean_token_accuracy": 0.16772648096084594,
      "num_tokens": 23200490.0,
      "step": 10590
    },
    {
      "entropy": 5.510408878326416,
      "epoch": 1.0515085351329894,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004898065993871571,
      "loss": 5.1964,
      "mean_token_accuracy": 0.16434268802404403,
      "num_tokens": 23212991.0,
      "step": 10595
    },
    {
      "entropy": 5.582481145858765,
      "epoch": 1.0520047637951568,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000489796052918233,
      "loss": 5.286,
      "mean_token_accuracy": 0.1607297196984291,
      "num_tokens": 23223361.0,
      "step": 10600
    },
    {
      "entropy": 5.596232318878174,
      "epoch": 1.0525009924573243,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004897855011228295,
      "loss": 5.1966,
      "mean_token_accuracy": 0.165140238404274,
      "num_tokens": 23233765.0,
      "step": 10605
    },
    {
      "entropy": 5.522750568389893,
      "epoch": 1.0529972211194918,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004897749440012085,
      "loss": 5.1486,
      "mean_token_accuracy": 0.17227747291326523,
      "num_tokens": 23246562.0,
      "step": 10610
    },
    {
      "entropy": 5.499996852874756,
      "epoch": 1.0534934497816595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897643815536316,
      "loss": 5.1664,
      "mean_token_accuracy": 0.16706736385822296,
      "num_tokens": 23256648.0,
      "step": 10615
    },
    {
      "entropy": 5.573162794113159,
      "epoch": 1.053989678443827,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000489753813780361,
      "loss": 5.2266,
      "mean_token_accuracy": 0.16290682107210158,
      "num_tokens": 23267428.0,
      "step": 10620
    },
    {
      "entropy": 5.552780866622925,
      "epoch": 1.0544859071059944,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004897432406816585,
      "loss": 5.1577,
      "mean_token_accuracy": 0.16331737488508224,
      "num_tokens": 23278440.0,
      "step": 10625
    },
    {
      "entropy": 5.533991003036499,
      "epoch": 1.054982135768162,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004897326622577863,
      "loss": 5.1988,
      "mean_token_accuracy": 0.15689337253570557,
      "num_tokens": 23290571.0,
      "step": 10630
    },
    {
      "entropy": 5.6039793491363525,
      "epoch": 1.0554783644303294,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004897220785090067,
      "loss": 5.2839,
      "mean_token_accuracy": 0.16086291521787643,
      "num_tokens": 23301634.0,
      "step": 10635
    },
    {
      "entropy": 5.497525596618653,
      "epoch": 1.055974593092497,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004897114894355824,
      "loss": 5.1524,
      "mean_token_accuracy": 0.16921741068363189,
      "num_tokens": 23313125.0,
      "step": 10640
    },
    {
      "entropy": 5.5201784610748295,
      "epoch": 1.0564708217546646,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004897008950377758,
      "loss": 5.1765,
      "mean_token_accuracy": 0.1607747569680214,
      "num_tokens": 23323610.0,
      "step": 10645
    },
    {
      "entropy": 5.514928102493286,
      "epoch": 1.056967050416832,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004896902953158497,
      "loss": 5.1493,
      "mean_token_accuracy": 0.16465729773044585,
      "num_tokens": 23333652.0,
      "step": 10650
    },
    {
      "entropy": 5.493526792526245,
      "epoch": 1.0574632790789995,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896796902700667,
      "loss": 5.1988,
      "mean_token_accuracy": 0.16942174285650252,
      "num_tokens": 23345954.0,
      "step": 10655
    },
    {
      "entropy": 5.519980478286743,
      "epoch": 1.0579595077411672,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004896690799006901,
      "loss": 5.1689,
      "mean_token_accuracy": 0.16729920357465744,
      "num_tokens": 23357052.0,
      "step": 10660
    },
    {
      "entropy": 5.552218341827393,
      "epoch": 1.0584557364033347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896584642079828,
      "loss": 5.1925,
      "mean_token_accuracy": 0.1709657207131386,
      "num_tokens": 23368837.0,
      "step": 10665
    },
    {
      "entropy": 5.523366785049438,
      "epoch": 1.0589519650655022,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004896478431922083,
      "loss": 5.1717,
      "mean_token_accuracy": 0.17029985636472703,
      "num_tokens": 23378608.0,
      "step": 10670
    },
    {
      "entropy": 5.58826060295105,
      "epoch": 1.0594481937276696,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004896372168536297,
      "loss": 5.2389,
      "mean_token_accuracy": 0.16770715564489364,
      "num_tokens": 23390094.0,
      "step": 10675
    },
    {
      "entropy": 5.494886636734009,
      "epoch": 1.0599444223898373,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004896265851925107,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17075427621603012,
      "num_tokens": 23399991.0,
      "step": 10680
    },
    {
      "entropy": 5.590870189666748,
      "epoch": 1.0604406510520048,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004896159482091148,
      "loss": 5.3007,
      "mean_token_accuracy": 0.15832749903202056,
      "num_tokens": 23410705.0,
      "step": 10685
    },
    {
      "entropy": 5.504523992538452,
      "epoch": 1.0609368797141723,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004896053059037058,
      "loss": 5.0864,
      "mean_token_accuracy": 0.18163858354091644,
      "num_tokens": 23421939.0,
      "step": 10690
    },
    {
      "entropy": 5.535493659973144,
      "epoch": 1.0614331083763398,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895946582765477,
      "loss": 5.2271,
      "mean_token_accuracy": 0.16575606912374496,
      "num_tokens": 23433252.0,
      "step": 10695
    },
    {
      "entropy": 5.643127059936523,
      "epoch": 1.0619293370385074,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004895840053279045,
      "loss": 5.2853,
      "mean_token_accuracy": 0.1574098899960518,
      "num_tokens": 23445582.0,
      "step": 10700
    },
    {
      "entropy": 5.597279691696167,
      "epoch": 1.062425565700675,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004895733470580404,
      "loss": 5.2566,
      "mean_token_accuracy": 0.15683240443468094,
      "num_tokens": 23456360.0,
      "step": 10705
    },
    {
      "entropy": 5.566844081878662,
      "epoch": 1.0629217943628424,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004895626834672196,
      "loss": 5.1444,
      "mean_token_accuracy": 0.16355169266462327,
      "num_tokens": 23466576.0,
      "step": 10710
    },
    {
      "entropy": 5.4719232559204105,
      "epoch": 1.0634180230250099,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004895520145557066,
      "loss": 5.0857,
      "mean_token_accuracy": 0.17191908359527588,
      "num_tokens": 23476200.0,
      "step": 10715
    },
    {
      "entropy": 5.553640699386596,
      "epoch": 1.0639142516871773,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004895413403237659,
      "loss": 5.2907,
      "mean_token_accuracy": 0.16775867640972136,
      "num_tokens": 23487776.0,
      "step": 10720
    },
    {
      "entropy": 5.541429090499878,
      "epoch": 1.064410480349345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895306607716624,
      "loss": 5.1536,
      "mean_token_accuracy": 0.16976266652345656,
      "num_tokens": 23497517.0,
      "step": 10725
    },
    {
      "entropy": 5.573356819152832,
      "epoch": 1.0649067090115125,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004895199758996607,
      "loss": 5.1684,
      "mean_token_accuracy": 0.17306934297084808,
      "num_tokens": 23507436.0,
      "step": 10730
    },
    {
      "entropy": 5.535901212692261,
      "epoch": 1.06540293767368,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004895092857080258,
      "loss": 5.1172,
      "mean_token_accuracy": 0.16114286333322525,
      "num_tokens": 23517170.0,
      "step": 10735
    },
    {
      "entropy": 5.508092641830444,
      "epoch": 1.0658991663358475,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894985901970229,
      "loss": 5.1262,
      "mean_token_accuracy": 0.1723567932844162,
      "num_tokens": 23527329.0,
      "step": 10740
    },
    {
      "entropy": 5.576211214065552,
      "epoch": 1.0663953949980152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004894878893669171,
      "loss": 5.2302,
      "mean_token_accuracy": 0.16117187440395356,
      "num_tokens": 23538341.0,
      "step": 10745
    },
    {
      "entropy": 5.5440410614013675,
      "epoch": 1.0668916236601826,
      "grad_norm": 0.875,
      "learning_rate": 0.0004894771832179739,
      "loss": 5.1677,
      "mean_token_accuracy": 0.16601282209157944,
      "num_tokens": 23549522.0,
      "step": 10750
    },
    {
      "entropy": 5.593019151687622,
      "epoch": 1.0673878523223501,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004894664717504587,
      "loss": 5.1946,
      "mean_token_accuracy": 0.1618312656879425,
      "num_tokens": 23559274.0,
      "step": 10755
    },
    {
      "entropy": 5.542934894561768,
      "epoch": 1.0678840809845176,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894557549646371,
      "loss": 5.1635,
      "mean_token_accuracy": 0.17135603129863738,
      "num_tokens": 23570852.0,
      "step": 10760
    },
    {
      "entropy": 5.520334196090698,
      "epoch": 1.0683803096466853,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004894450328607749,
      "loss": 5.1576,
      "mean_token_accuracy": 0.16817104071378708,
      "num_tokens": 23582918.0,
      "step": 10765
    },
    {
      "entropy": 5.501906108856201,
      "epoch": 1.0688765383088528,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004894343054391379,
      "loss": 5.1863,
      "mean_token_accuracy": 0.17586371600627898,
      "num_tokens": 23594177.0,
      "step": 10770
    },
    {
      "entropy": 5.536865615844727,
      "epoch": 1.0693727669710202,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004894235726999924,
      "loss": 5.153,
      "mean_token_accuracy": 0.1688952773809433,
      "num_tokens": 23605901.0,
      "step": 10775
    },
    {
      "entropy": 5.510081338882446,
      "epoch": 1.0698689956331877,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004894128346436042,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16656966507434845,
      "num_tokens": 23617354.0,
      "step": 10780
    },
    {
      "entropy": 5.559385919570923,
      "epoch": 1.0703652242953554,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004894020912702397,
      "loss": 5.2318,
      "mean_token_accuracy": 0.16356824785470964,
      "num_tokens": 23627603.0,
      "step": 10785
    },
    {
      "entropy": 5.461163806915283,
      "epoch": 1.0708614529575229,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004893913425801653,
      "loss": 5.0565,
      "mean_token_accuracy": 0.17585218399763108,
      "num_tokens": 23637280.0,
      "step": 10790
    },
    {
      "entropy": 5.540579557418823,
      "epoch": 1.0713576816196904,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004893805885736477,
      "loss": 5.2718,
      "mean_token_accuracy": 0.16151050329208375,
      "num_tokens": 23649326.0,
      "step": 10795
    },
    {
      "entropy": 5.614642858505249,
      "epoch": 1.0718539102818578,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004893698292509534,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16112040877342224,
      "num_tokens": 23659561.0,
      "step": 10800
    },
    {
      "entropy": 5.567789697647095,
      "epoch": 1.0723501389440253,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004893590646123492,
      "loss": 5.1692,
      "mean_token_accuracy": 0.16753435581922532,
      "num_tokens": 23673206.0,
      "step": 10805
    },
    {
      "entropy": 5.526480722427368,
      "epoch": 1.072846367606193,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004893482946581022,
      "loss": 5.2144,
      "mean_token_accuracy": 0.16125859469175338,
      "num_tokens": 23683920.0,
      "step": 10810
    },
    {
      "entropy": 5.503568696975708,
      "epoch": 1.0733425962683605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004893375193884793,
      "loss": 5.1213,
      "mean_token_accuracy": 0.17639651894569397,
      "num_tokens": 23693990.0,
      "step": 10815
    },
    {
      "entropy": 5.579636287689209,
      "epoch": 1.073838824930528,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893267388037478,
      "loss": 5.2338,
      "mean_token_accuracy": 0.161630080640316,
      "num_tokens": 23704870.0,
      "step": 10820
    },
    {
      "entropy": 5.519946813583374,
      "epoch": 1.0743350535926954,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000489315952904175,
      "loss": 5.2203,
      "mean_token_accuracy": 0.16763634085655213,
      "num_tokens": 23715159.0,
      "step": 10825
    },
    {
      "entropy": 5.538396883010864,
      "epoch": 1.0748312822548631,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004893051616900283,
      "loss": 5.1509,
      "mean_token_accuracy": 0.16700153052806854,
      "num_tokens": 23725252.0,
      "step": 10830
    },
    {
      "entropy": 5.498171281814575,
      "epoch": 1.0753275109170306,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004892943651615756,
      "loss": 5.1806,
      "mean_token_accuracy": 0.17200819700956343,
      "num_tokens": 23736472.0,
      "step": 10835
    },
    {
      "entropy": 5.553162384033203,
      "epoch": 1.075823739579198,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004892835633190843,
      "loss": 5.0808,
      "mean_token_accuracy": 0.16729693114757538,
      "num_tokens": 23747565.0,
      "step": 10840
    },
    {
      "entropy": 5.535704469680786,
      "epoch": 1.0763199682413656,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004892727561628223,
      "loss": 5.1812,
      "mean_token_accuracy": 0.16496773213148117,
      "num_tokens": 23758201.0,
      "step": 10845
    },
    {
      "entropy": 5.562637805938721,
      "epoch": 1.0768161969035333,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004892619436930577,
      "loss": 5.2426,
      "mean_token_accuracy": 0.16716550141572953,
      "num_tokens": 23769087.0,
      "step": 10850
    },
    {
      "entropy": 5.556429481506347,
      "epoch": 1.0773124255657007,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004892511259100585,
      "loss": 5.2304,
      "mean_token_accuracy": 0.1665942430496216,
      "num_tokens": 23780598.0,
      "step": 10855
    },
    {
      "entropy": 5.5416840553283695,
      "epoch": 1.0778086542278682,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004892403028140932,
      "loss": 5.1509,
      "mean_token_accuracy": 0.16183367222547532,
      "num_tokens": 23791449.0,
      "step": 10860
    },
    {
      "entropy": 5.528909730911255,
      "epoch": 1.0783048828900357,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048922947440543,
      "loss": 5.1496,
      "mean_token_accuracy": 0.16679588705301285,
      "num_tokens": 23803071.0,
      "step": 10865
    },
    {
      "entropy": 5.562570571899414,
      "epoch": 1.0788011115522032,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004892186406843374,
      "loss": 5.2116,
      "mean_token_accuracy": 0.15827475786209105,
      "num_tokens": 23813372.0,
      "step": 10870
    },
    {
      "entropy": 5.512828683853149,
      "epoch": 1.0792973402143708,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004892078016510841,
      "loss": 5.1914,
      "mean_token_accuracy": 0.1670689731836319,
      "num_tokens": 23824979.0,
      "step": 10875
    },
    {
      "entropy": 5.51799464225769,
      "epoch": 1.0797935688765383,
      "grad_norm": 0.890625,
      "learning_rate": 0.000489196957305939,
      "loss": 5.1776,
      "mean_token_accuracy": 0.16226282417774202,
      "num_tokens": 23836335.0,
      "step": 10880
    },
    {
      "entropy": 5.522289991378784,
      "epoch": 1.0802897975387058,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004891861076491709,
      "loss": 5.1675,
      "mean_token_accuracy": 0.1688612163066864,
      "num_tokens": 23848147.0,
      "step": 10885
    },
    {
      "entropy": 5.4876823902130125,
      "epoch": 1.0807860262008733,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004891752526810488,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17157595306634904,
      "num_tokens": 23859502.0,
      "step": 10890
    },
    {
      "entropy": 5.560577011108398,
      "epoch": 1.081282254863041,
      "grad_norm": 0.84375,
      "learning_rate": 0.000489164392401842,
      "loss": 5.2655,
      "mean_token_accuracy": 0.16266265660524368,
      "num_tokens": 23870433.0,
      "step": 10895
    },
    {
      "entropy": 5.506106567382813,
      "epoch": 1.0817784835252084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891535268118196,
      "loss": 5.0803,
      "mean_token_accuracy": 0.17452805042266845,
      "num_tokens": 23879581.0,
      "step": 10900
    },
    {
      "entropy": 5.457401371002197,
      "epoch": 1.082274712187376,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004891426559112514,
      "loss": 5.1571,
      "mean_token_accuracy": 0.17030167132616042,
      "num_tokens": 23890889.0,
      "step": 10905
    },
    {
      "entropy": 5.548341274261475,
      "epoch": 1.0827709408495434,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004891317797004066,
      "loss": 5.1434,
      "mean_token_accuracy": 0.1688428357243538,
      "num_tokens": 23901577.0,
      "step": 10910
    },
    {
      "entropy": 5.529557943344116,
      "epoch": 1.083267169511711,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004891208981795552,
      "loss": 5.2205,
      "mean_token_accuracy": 0.16735608577728273,
      "num_tokens": 23912698.0,
      "step": 10915
    },
    {
      "entropy": 5.45461893081665,
      "epoch": 1.0837633981738786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004891100113489669,
      "loss": 5.1064,
      "mean_token_accuracy": 0.1750051572918892,
      "num_tokens": 23923056.0,
      "step": 10920
    },
    {
      "entropy": 5.520450067520142,
      "epoch": 1.084259626836046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004890991192089116,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16565134078264238,
      "num_tokens": 23933998.0,
      "step": 10925
    },
    {
      "entropy": 5.551449298858643,
      "epoch": 1.0847558554982135,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004890882217596595,
      "loss": 5.132,
      "mean_token_accuracy": 0.17169341892004014,
      "num_tokens": 23943980.0,
      "step": 10930
    },
    {
      "entropy": 5.523691844940186,
      "epoch": 1.085252084160381,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004890773190014809,
      "loss": 5.2058,
      "mean_token_accuracy": 0.1679893061518669,
      "num_tokens": 23955646.0,
      "step": 10935
    },
    {
      "entropy": 5.49336314201355,
      "epoch": 1.0857483128225487,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000489066410934646,
      "loss": 5.172,
      "mean_token_accuracy": 0.17093188762664796,
      "num_tokens": 23966381.0,
      "step": 10940
    },
    {
      "entropy": 5.610286092758178,
      "epoch": 1.0862445414847162,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004890554975594255,
      "loss": 5.2498,
      "mean_token_accuracy": 0.15676361620426177,
      "num_tokens": 23978669.0,
      "step": 10945
    },
    {
      "entropy": 5.675099802017212,
      "epoch": 1.0867407701468836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004890445788760898,
      "loss": 5.3133,
      "mean_token_accuracy": 0.15942621231079102,
      "num_tokens": 23989467.0,
      "step": 10950
    },
    {
      "entropy": 5.512882232666016,
      "epoch": 1.0872369988090511,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004890336548849098,
      "loss": 5.1287,
      "mean_token_accuracy": 0.17198625653982164,
      "num_tokens": 24000797.0,
      "step": 10955
    },
    {
      "entropy": 5.605676507949829,
      "epoch": 1.0877332274712188,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004890227255861564,
      "loss": 5.2776,
      "mean_token_accuracy": 0.1670268550515175,
      "num_tokens": 24013182.0,
      "step": 10960
    },
    {
      "entropy": 5.563385248184204,
      "epoch": 1.0882294561333863,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890117909801005,
      "loss": 5.1626,
      "mean_token_accuracy": 0.17054594904184342,
      "num_tokens": 24024005.0,
      "step": 10965
    },
    {
      "entropy": 5.518511009216309,
      "epoch": 1.0887256847955538,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004890008510670134,
      "loss": 5.1501,
      "mean_token_accuracy": 0.16404007226228715,
      "num_tokens": 24034524.0,
      "step": 10970
    },
    {
      "entropy": 5.56888370513916,
      "epoch": 1.0892219134577212,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004889899058471663,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1660050630569458,
      "num_tokens": 24045944.0,
      "step": 10975
    },
    {
      "entropy": 5.598436737060547,
      "epoch": 1.089718142119889,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004889789553208306,
      "loss": 5.2827,
      "mean_token_accuracy": 0.1679112732410431,
      "num_tokens": 24058612.0,
      "step": 10980
    },
    {
      "entropy": 5.574249172210694,
      "epoch": 1.0902143707820564,
      "grad_norm": 1.046875,
      "learning_rate": 0.000488967999488278,
      "loss": 5.147,
      "mean_token_accuracy": 0.1689131736755371,
      "num_tokens": 24068912.0,
      "step": 10985
    },
    {
      "entropy": 5.463873386383057,
      "epoch": 1.0907105994442239,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048895703834978,
      "loss": 5.1483,
      "mean_token_accuracy": 0.1694681912660599,
      "num_tokens": 24079879.0,
      "step": 10990
    },
    {
      "entropy": 5.5168863296508786,
      "epoch": 1.0912068281063914,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004889460719056085,
      "loss": 5.1785,
      "mean_token_accuracy": 0.168193057179451,
      "num_tokens": 24091268.0,
      "step": 10995
    },
    {
      "entropy": 5.524760007858276,
      "epoch": 1.091703056768559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004889351001560354,
      "loss": 5.2017,
      "mean_token_accuracy": 0.16584371775388718,
      "num_tokens": 24101428.0,
      "step": 11000
    },
    {
      "entropy": 5.489978075027466,
      "epoch": 1.0921992854307265,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004889241231013329,
      "loss": 5.1695,
      "mean_token_accuracy": 0.16908123791217805,
      "num_tokens": 24112920.0,
      "step": 11005
    },
    {
      "entropy": 5.495450592041015,
      "epoch": 1.092695514092894,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000488913140741773,
      "loss": 5.1115,
      "mean_token_accuracy": 0.1671591192483902,
      "num_tokens": 24125731.0,
      "step": 11010
    },
    {
      "entropy": 5.505485200881958,
      "epoch": 1.0931917427550615,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004889021530776282,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17019040435552596,
      "num_tokens": 24136606.0,
      "step": 11015
    },
    {
      "entropy": 5.560671329498291,
      "epoch": 1.093687971417229,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004888911601091709,
      "loss": 5.1352,
      "mean_token_accuracy": 0.16939222663640977,
      "num_tokens": 24146783.0,
      "step": 11020
    },
    {
      "entropy": 5.512586164474487,
      "epoch": 1.0941842000793967,
      "grad_norm": 0.875,
      "learning_rate": 0.0004888801618366737,
      "loss": 5.2314,
      "mean_token_accuracy": 0.16504254639148713,
      "num_tokens": 24157412.0,
      "step": 11025
    },
    {
      "entropy": 5.524743938446045,
      "epoch": 1.0946804287415641,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888691582604095,
      "loss": 5.223,
      "mean_token_accuracy": 0.16580145359039306,
      "num_tokens": 24168077.0,
      "step": 11030
    },
    {
      "entropy": 5.590411138534546,
      "epoch": 1.0951766574037316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488858149380651,
      "loss": 5.2656,
      "mean_token_accuracy": 0.1663936644792557,
      "num_tokens": 24178913.0,
      "step": 11035
    },
    {
      "entropy": 5.647264003753662,
      "epoch": 1.095672886065899,
      "grad_norm": 0.953125,
      "learning_rate": 0.000488847135197671,
      "loss": 5.3108,
      "mean_token_accuracy": 0.15696686804294585,
      "num_tokens": 24190356.0,
      "step": 11040
    },
    {
      "entropy": 5.613994121551514,
      "epoch": 1.0961691147280668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000488836115711743,
      "loss": 5.2734,
      "mean_token_accuracy": 0.16120597571134568,
      "num_tokens": 24199772.0,
      "step": 11045
    },
    {
      "entropy": 5.584469032287598,
      "epoch": 1.0966653433902342,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004888250909231401,
      "loss": 5.1733,
      "mean_token_accuracy": 0.16744020730257034,
      "num_tokens": 24211807.0,
      "step": 11050
    },
    {
      "entropy": 5.5088718891143795,
      "epoch": 1.0971615720524017,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004888140608321357,
      "loss": 5.1589,
      "mean_token_accuracy": 0.17269312739372253,
      "num_tokens": 24223156.0,
      "step": 11055
    },
    {
      "entropy": 5.595982837677002,
      "epoch": 1.0976578007145692,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004888030254390033,
      "loss": 5.2258,
      "mean_token_accuracy": 0.1636517494916916,
      "num_tokens": 24234352.0,
      "step": 11060
    },
    {
      "entropy": 5.537501192092895,
      "epoch": 1.098154029376737,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004887919847440166,
      "loss": 5.1404,
      "mean_token_accuracy": 0.17381986677646638,
      "num_tokens": 24244137.0,
      "step": 11065
    },
    {
      "entropy": 5.479118585586548,
      "epoch": 1.0986502580389044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887809387474494,
      "loss": 5.152,
      "mean_token_accuracy": 0.16452469825744628,
      "num_tokens": 24253250.0,
      "step": 11070
    },
    {
      "entropy": 5.581023836135865,
      "epoch": 1.0991464867010718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887698874495755,
      "loss": 5.2622,
      "mean_token_accuracy": 0.16246499717235566,
      "num_tokens": 24264551.0,
      "step": 11075
    },
    {
      "entropy": 5.548129034042359,
      "epoch": 1.0996427153632393,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000488758830850669,
      "loss": 5.0826,
      "mean_token_accuracy": 0.17373690158128738,
      "num_tokens": 24275522.0,
      "step": 11080
    },
    {
      "entropy": 5.502170515060425,
      "epoch": 1.100138944025407,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004887477689510042,
      "loss": 5.2347,
      "mean_token_accuracy": 0.16451115757226945,
      "num_tokens": 24286333.0,
      "step": 11085
    },
    {
      "entropy": 5.470165205001831,
      "epoch": 1.1006351726875745,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004887367017508553,
      "loss": 5.2454,
      "mean_token_accuracy": 0.16755399256944656,
      "num_tokens": 24297627.0,
      "step": 11090
    },
    {
      "entropy": 5.520968770980835,
      "epoch": 1.101131401349742,
      "grad_norm": 0.875,
      "learning_rate": 0.0004887256292504967,
      "loss": 5.1877,
      "mean_token_accuracy": 0.1679129421710968,
      "num_tokens": 24308246.0,
      "step": 11095
    },
    {
      "entropy": 5.593477344512939,
      "epoch": 1.1016276300119094,
      "grad_norm": 1.078125,
      "learning_rate": 0.000488714551450203,
      "loss": 5.2471,
      "mean_token_accuracy": 0.16170470118522645,
      "num_tokens": 24317718.0,
      "step": 11100
    },
    {
      "entropy": 5.544014024734497,
      "epoch": 1.102123858674077,
      "grad_norm": 0.921875,
      "learning_rate": 0.000488703468350249,
      "loss": 5.1627,
      "mean_token_accuracy": 0.16755675971508027,
      "num_tokens": 24329242.0,
      "step": 11105
    },
    {
      "entropy": 5.650244617462159,
      "epoch": 1.1026200873362446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004886923799509095,
      "loss": 5.3236,
      "mean_token_accuracy": 0.15826241821050643,
      "num_tokens": 24340146.0,
      "step": 11110
    },
    {
      "entropy": 5.60164737701416,
      "epoch": 1.103116315998412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004886812862524595,
      "loss": 5.2233,
      "mean_token_accuracy": 0.16297439038753508,
      "num_tokens": 24349370.0,
      "step": 11115
    },
    {
      "entropy": 5.620865726470948,
      "epoch": 1.1036125446605796,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004886701872551739,
      "loss": 5.1788,
      "mean_token_accuracy": 0.16222323030233382,
      "num_tokens": 24359770.0,
      "step": 11120
    },
    {
      "entropy": 5.546250629425049,
      "epoch": 1.104108773322747,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004886590829593281,
      "loss": 5.2201,
      "mean_token_accuracy": 0.16503935158252717,
      "num_tokens": 24370974.0,
      "step": 11125
    },
    {
      "entropy": 5.561810684204102,
      "epoch": 1.1046050019849147,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004886479733651974,
      "loss": 5.2148,
      "mean_token_accuracy": 0.1693412810564041,
      "num_tokens": 24381771.0,
      "step": 11130
    },
    {
      "entropy": 5.525352954864502,
      "epoch": 1.1051012306470822,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004886368584730575,
      "loss": 5.1805,
      "mean_token_accuracy": 0.16449014395475386,
      "num_tokens": 24393067.0,
      "step": 11135
    },
    {
      "entropy": 5.559745740890503,
      "epoch": 1.1055974593092497,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004886257382831837,
      "loss": 5.2547,
      "mean_token_accuracy": 0.16137390434741974,
      "num_tokens": 24402866.0,
      "step": 11140
    },
    {
      "entropy": 5.611644506454468,
      "epoch": 1.1060936879714172,
      "grad_norm": 0.9375,
      "learning_rate": 0.000488614612795852,
      "loss": 5.1822,
      "mean_token_accuracy": 0.16718950718641282,
      "num_tokens": 24412956.0,
      "step": 11145
    },
    {
      "entropy": 5.6083714962005615,
      "epoch": 1.1065899166335849,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004886034820113382,
      "loss": 5.2441,
      "mean_token_accuracy": 0.16436125785112382,
      "num_tokens": 24424650.0,
      "step": 11150
    },
    {
      "entropy": 5.5847437381744385,
      "epoch": 1.1070861452957523,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004885923459299182,
      "loss": 5.2598,
      "mean_token_accuracy": 0.16341067850589752,
      "num_tokens": 24436541.0,
      "step": 11155
    },
    {
      "entropy": 5.590815925598145,
      "epoch": 1.1075823739579198,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004885812045518685,
      "loss": 5.2395,
      "mean_token_accuracy": 0.16126682460308076,
      "num_tokens": 24447588.0,
      "step": 11160
    },
    {
      "entropy": 5.577991151809693,
      "epoch": 1.1080786026200873,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000488570057877465,
      "loss": 5.1603,
      "mean_token_accuracy": 0.17224933654069902,
      "num_tokens": 24458854.0,
      "step": 11165
    },
    {
      "entropy": 5.552229261398315,
      "epoch": 1.108574831282255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885589059069842,
      "loss": 5.2228,
      "mean_token_accuracy": 0.1659489780664444,
      "num_tokens": 24469427.0,
      "step": 11170
    },
    {
      "entropy": 5.488398933410645,
      "epoch": 1.1090710599444225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004885477486407028,
      "loss": 5.2215,
      "mean_token_accuracy": 0.15970034003257752,
      "num_tokens": 24479352.0,
      "step": 11175
    },
    {
      "entropy": 5.645944118499756,
      "epoch": 1.10956728860659,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004885365860788974,
      "loss": 5.2223,
      "mean_token_accuracy": 0.16270780116319655,
      "num_tokens": 24489580.0,
      "step": 11180
    },
    {
      "entropy": 5.544403171539306,
      "epoch": 1.1100635172687574,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004885254182218448,
      "loss": 5.1657,
      "mean_token_accuracy": 0.16676835864782333,
      "num_tokens": 24501141.0,
      "step": 11185
    },
    {
      "entropy": 5.482193851470948,
      "epoch": 1.1105597459309249,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004885142450698219,
      "loss": 5.1725,
      "mean_token_accuracy": 0.16832606345415116,
      "num_tokens": 24512079.0,
      "step": 11190
    },
    {
      "entropy": 5.590605783462524,
      "epoch": 1.1110559745930926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004885030666231057,
      "loss": 5.224,
      "mean_token_accuracy": 0.16458019167184829,
      "num_tokens": 24521839.0,
      "step": 11195
    },
    {
      "entropy": 5.610052490234375,
      "epoch": 1.11155220325526,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004884918828819738,
      "loss": 5.2688,
      "mean_token_accuracy": 0.16540856808423995,
      "num_tokens": 24532704.0,
      "step": 11200
    },
    {
      "entropy": 5.559331846237183,
      "epoch": 1.1120484319174275,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000488480693846703,
      "loss": 5.179,
      "mean_token_accuracy": 0.16953056305646896,
      "num_tokens": 24543790.0,
      "step": 11205
    },
    {
      "entropy": 5.524071455001831,
      "epoch": 1.112544660579595,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004884694995175711,
      "loss": 5.2213,
      "mean_token_accuracy": 0.16467441171407698,
      "num_tokens": 24554765.0,
      "step": 11210
    },
    {
      "entropy": 5.662983703613281,
      "epoch": 1.1130408892417627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004884582998948555,
      "loss": 5.2944,
      "mean_token_accuracy": 0.15750882476568223,
      "num_tokens": 24565888.0,
      "step": 11215
    },
    {
      "entropy": 5.523851299285889,
      "epoch": 1.1135371179039302,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004884470949788342,
      "loss": 5.1549,
      "mean_token_accuracy": 0.1693478301167488,
      "num_tokens": 24575482.0,
      "step": 11220
    },
    {
      "entropy": 5.6012979507446286,
      "epoch": 1.1140333465660976,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004884358847697847,
      "loss": 5.2822,
      "mean_token_accuracy": 0.15506206154823304,
      "num_tokens": 24588249.0,
      "step": 11225
    },
    {
      "entropy": 5.6031286239624025,
      "epoch": 1.1145295752282651,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004884246692679852,
      "loss": 5.2328,
      "mean_token_accuracy": 0.16978827267885208,
      "num_tokens": 24599546.0,
      "step": 11230
    },
    {
      "entropy": 5.544774436950684,
      "epoch": 1.1150258038904326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004884134484737139,
      "loss": 5.1663,
      "mean_token_accuracy": 0.1680122897028923,
      "num_tokens": 24611777.0,
      "step": 11235
    },
    {
      "entropy": 5.515751647949219,
      "epoch": 1.1155220325526003,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004884022223872489,
      "loss": 5.1491,
      "mean_token_accuracy": 0.1688683420419693,
      "num_tokens": 24623675.0,
      "step": 11240
    },
    {
      "entropy": 5.547360420227051,
      "epoch": 1.1160182612147678,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004883909910088686,
      "loss": 5.1759,
      "mean_token_accuracy": 0.1662119060754776,
      "num_tokens": 24633929.0,
      "step": 11245
    },
    {
      "entropy": 5.516512727737426,
      "epoch": 1.1165144898769352,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004883797543388515,
      "loss": 5.1402,
      "mean_token_accuracy": 0.16952908039093018,
      "num_tokens": 24642800.0,
      "step": 11250
    },
    {
      "entropy": 5.475958967208863,
      "epoch": 1.1170107185391027,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883685123774764,
      "loss": 5.1382,
      "mean_token_accuracy": 0.1716240718960762,
      "num_tokens": 24653780.0,
      "step": 11255
    },
    {
      "entropy": 5.514026498794555,
      "epoch": 1.1175069472012704,
      "grad_norm": 0.875,
      "learning_rate": 0.0004883572651250218,
      "loss": 5.1755,
      "mean_token_accuracy": 0.16623347997665405,
      "num_tokens": 24663912.0,
      "step": 11260
    },
    {
      "entropy": 5.601355934143067,
      "epoch": 1.118003175863438,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004883460125817669,
      "loss": 5.2594,
      "mean_token_accuracy": 0.1595919281244278,
      "num_tokens": 24675222.0,
      "step": 11265
    },
    {
      "entropy": 5.6082860946655275,
      "epoch": 1.1184994045256054,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004883347547479905,
      "loss": 5.2512,
      "mean_token_accuracy": 0.16543015837669373,
      "num_tokens": 24686697.0,
      "step": 11270
    },
    {
      "entropy": 5.602556467056274,
      "epoch": 1.1189956331877728,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488323491623972,
      "loss": 5.1857,
      "mean_token_accuracy": 0.17011242508888244,
      "num_tokens": 24697067.0,
      "step": 11275
    },
    {
      "entropy": 5.568862771987915,
      "epoch": 1.1194918618499405,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004883122232099905,
      "loss": 5.1876,
      "mean_token_accuracy": 0.1691440835595131,
      "num_tokens": 24707989.0,
      "step": 11280
    },
    {
      "entropy": 5.478446197509766,
      "epoch": 1.119988090512108,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004883009495063256,
      "loss": 5.1298,
      "mean_token_accuracy": 0.17486218363046646,
      "num_tokens": 24718826.0,
      "step": 11285
    },
    {
      "entropy": 5.58632607460022,
      "epoch": 1.1204843191742755,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004882896705132566,
      "loss": 5.241,
      "mean_token_accuracy": 0.15880665481090545,
      "num_tokens": 24729913.0,
      "step": 11290
    },
    {
      "entropy": 5.496412849426269,
      "epoch": 1.120980547836443,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048827838623106355,
      "loss": 5.0954,
      "mean_token_accuracy": 0.1684751346707344,
      "num_tokens": 24741621.0,
      "step": 11295
    },
    {
      "entropy": 5.5385157585144045,
      "epoch": 1.1214767764986107,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000488267096660026,
      "loss": 5.1469,
      "mean_token_accuracy": 0.16987870633602142,
      "num_tokens": 24753427.0,
      "step": 11300
    },
    {
      "entropy": 5.560443687438965,
      "epoch": 1.1219730051607781,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048825580180042406,
      "loss": 5.197,
      "mean_token_accuracy": 0.1681511402130127,
      "num_tokens": 24763580.0,
      "step": 11305
    },
    {
      "entropy": 5.529833984375,
      "epoch": 1.1224692338229456,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004882445016525377,
      "loss": 5.1842,
      "mean_token_accuracy": 0.1707780435681343,
      "num_tokens": 24773955.0,
      "step": 11310
    },
    {
      "entropy": 5.533229684829712,
      "epoch": 1.122965462485113,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048823319621664723,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17385689914226532,
      "num_tokens": 24784039.0,
      "step": 11315
    },
    {
      "entropy": 5.5877148628234865,
      "epoch": 1.1234616911472806,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048822188549303296,
      "loss": 5.2179,
      "mean_token_accuracy": 0.16231290400028228,
      "num_tokens": 24794630.0,
      "step": 11320
    },
    {
      "entropy": 5.55251898765564,
      "epoch": 1.1239579198094483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882105694819754,
      "loss": 5.264,
      "mean_token_accuracy": 0.1597123622894287,
      "num_tokens": 24805625.0,
      "step": 11325
    },
    {
      "entropy": 5.594995737075806,
      "epoch": 1.1244541484716157,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004881992481837551,
      "loss": 5.1922,
      "mean_token_accuracy": 0.17375536561012267,
      "num_tokens": 24817356.0,
      "step": 11330
    },
    {
      "entropy": 5.558232355117798,
      "epoch": 1.1249503771337832,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004881879215986529,
      "loss": 5.1697,
      "mean_token_accuracy": 0.16297870576381684,
      "num_tokens": 24828349.0,
      "step": 11335
    },
    {
      "entropy": 5.557170486450195,
      "epoch": 1.1254466057959507,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048817658972694967,
      "loss": 5.2103,
      "mean_token_accuracy": 0.1661566138267517,
      "num_tokens": 24840138.0,
      "step": 11340
    },
    {
      "entropy": 5.566293525695801,
      "epoch": 1.1259428344581184,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004881652525689263,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1644575923681259,
      "num_tokens": 24851570.0,
      "step": 11345
    },
    {
      "entropy": 5.4471972465515135,
      "epoch": 1.1264390631202859,
      "grad_norm": 0.890625,
      "learning_rate": 0.000488153910124864,
      "loss": 5.0931,
      "mean_token_accuracy": 0.17195719182491304,
      "num_tokens": 24861396.0,
      "step": 11350
    },
    {
      "entropy": 5.559403133392334,
      "epoch": 1.1269352917824533,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000488142562395044,
      "loss": 5.2059,
      "mean_token_accuracy": 0.16284462660551072,
      "num_tokens": 24873519.0,
      "step": 11355
    },
    {
      "entropy": 5.588962078094482,
      "epoch": 1.1274315204446208,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00048813120937974774,
      "loss": 5.2285,
      "mean_token_accuracy": 0.16117286533117295,
      "num_tokens": 24886856.0,
      "step": 11360
    },
    {
      "entropy": 5.530372047424317,
      "epoch": 1.1279277491067885,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048811985107925684,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16211043894290925,
      "num_tokens": 24897292.0,
      "step": 11365
    },
    {
      "entropy": 5.557800579071045,
      "epoch": 1.128423977768956,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048810848749385286,
      "loss": 5.1972,
      "mean_token_accuracy": 0.16763793230056762,
      "num_tokens": 24907757.0,
      "step": 11370
    },
    {
      "entropy": 5.482391214370727,
      "epoch": 1.1289202064311235,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048809711862381753,
      "loss": 5.1716,
      "mean_token_accuracy": 0.1716945141553879,
      "num_tokens": 24918071.0,
      "step": 11375
    },
    {
      "entropy": 5.517623376846314,
      "epoch": 1.129416435093291,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048808574446943274,
      "loss": 5.1425,
      "mean_token_accuracy": 0.17389156967401503,
      "num_tokens": 24927967.0,
      "step": 11380
    },
    {
      "entropy": 5.591787481307984,
      "epoch": 1.1299126637554586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004880743650309808,
      "loss": 5.2156,
      "mean_token_accuracy": 0.17123477309942245,
      "num_tokens": 24939170.0,
      "step": 11385
    },
    {
      "entropy": 5.549411678314209,
      "epoch": 1.130408892417626,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048806298030874373,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16855986118316652,
      "num_tokens": 24950937.0,
      "step": 11390
    },
    {
      "entropy": 5.5512247562408445,
      "epoch": 1.1309051210797936,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048805159030300383,
      "loss": 5.2018,
      "mean_token_accuracy": 0.1645895555615425,
      "num_tokens": 24962813.0,
      "step": 11395
    },
    {
      "entropy": 5.529220294952393,
      "epoch": 1.131401349741961,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048804019501404357,
      "loss": 5.0998,
      "mean_token_accuracy": 0.17063148319721222,
      "num_tokens": 24974192.0,
      "step": 11400
    },
    {
      "entropy": 5.5139471054077145,
      "epoch": 1.1318975784041285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048802879444214553,
      "loss": 5.0912,
      "mean_token_accuracy": 0.17605119049549103,
      "num_tokens": 24985353.0,
      "step": 11405
    },
    {
      "entropy": 5.529029798507691,
      "epoch": 1.1323938070662962,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048801738858759243,
      "loss": 5.2012,
      "mean_token_accuracy": 0.17208236306905747,
      "num_tokens": 24997259.0,
      "step": 11410
    },
    {
      "entropy": 5.518797588348389,
      "epoch": 1.1328900357284637,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004880059774506671,
      "loss": 5.1851,
      "mean_token_accuracy": 0.16762298941612244,
      "num_tokens": 25008222.0,
      "step": 11415
    },
    {
      "entropy": 5.531159543991089,
      "epoch": 1.1333862643906312,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004879945610316525,
      "loss": 5.2392,
      "mean_token_accuracy": 0.16421402394771575,
      "num_tokens": 25020267.0,
      "step": 11420
    },
    {
      "entropy": 5.537470388412475,
      "epoch": 1.1338824930527986,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048798313933083175,
      "loss": 5.1914,
      "mean_token_accuracy": 0.16577899008989333,
      "num_tokens": 25031212.0,
      "step": 11425
    },
    {
      "entropy": 5.536048221588135,
      "epoch": 1.1343787217149663,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004879717123484882,
      "loss": 5.1614,
      "mean_token_accuracy": 0.16391245275735855,
      "num_tokens": 25042204.0,
      "step": 11430
    },
    {
      "entropy": 5.569738435745239,
      "epoch": 1.1348749503771338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000487960280084905,
      "loss": 5.1813,
      "mean_token_accuracy": 0.17035711854696273,
      "num_tokens": 25052010.0,
      "step": 11435
    },
    {
      "entropy": 5.590561199188232,
      "epoch": 1.1353711790393013,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004879488425403657,
      "loss": 5.2445,
      "mean_token_accuracy": 0.1614553339779377,
      "num_tokens": 25062209.0,
      "step": 11440
    },
    {
      "entropy": 5.542884969711304,
      "epoch": 1.1358674077014688,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048793739971515407,
      "loss": 5.2064,
      "mean_token_accuracy": 0.16749328523874282,
      "num_tokens": 25072758.0,
      "step": 11445
    },
    {
      "entropy": 5.540720224380493,
      "epoch": 1.1363636363636362,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048792595160955374,
      "loss": 5.2773,
      "mean_token_accuracy": 0.1688552036881447,
      "num_tokens": 25084318.0,
      "step": 11450
    },
    {
      "entropy": 5.557518386840821,
      "epoch": 1.136859865025804,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048791449822384856,
      "loss": 5.1715,
      "mean_token_accuracy": 0.1697297677397728,
      "num_tokens": 25095715.0,
      "step": 11455
    },
    {
      "entropy": 5.582264518737793,
      "epoch": 1.1373560936879714,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004879030395583226,
      "loss": 5.1903,
      "mean_token_accuracy": 0.1683877557516098,
      "num_tokens": 25106163.0,
      "step": 11460
    },
    {
      "entropy": 5.597955894470215,
      "epoch": 1.137852322350139,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004878915756132601,
      "loss": 5.2568,
      "mean_token_accuracy": 0.16355849504470826,
      "num_tokens": 25118745.0,
      "step": 11465
    },
    {
      "entropy": 5.554415702819824,
      "epoch": 1.1383485510123066,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048788010638894523,
      "loss": 5.2363,
      "mean_token_accuracy": 0.15971914529800416,
      "num_tokens": 25130066.0,
      "step": 11470
    },
    {
      "entropy": 5.518828821182251,
      "epoch": 1.138844779674474,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004878686318856624,
      "loss": 5.0928,
      "mean_token_accuracy": 0.17169537246227265,
      "num_tokens": 25139895.0,
      "step": 11475
    },
    {
      "entropy": 5.633216190338135,
      "epoch": 1.1393410083366415,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048785715210369635,
      "loss": 5.3116,
      "mean_token_accuracy": 0.15535952746868134,
      "num_tokens": 25150644.0,
      "step": 11480
    },
    {
      "entropy": 5.599082803726196,
      "epoch": 1.139837236998809,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004878456670433314,
      "loss": 5.2421,
      "mean_token_accuracy": 0.16517586261034012,
      "num_tokens": 25163038.0,
      "step": 11485
    },
    {
      "entropy": 5.540990543365479,
      "epoch": 1.1403334656609765,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004878341767048527,
      "loss": 5.137,
      "mean_token_accuracy": 0.1713399201631546,
      "num_tokens": 25174154.0,
      "step": 11490
    },
    {
      "entropy": 5.595341634750366,
      "epoch": 1.1408296943231442,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048782268108854494,
      "loss": 5.2542,
      "mean_token_accuracy": 0.16223768442869185,
      "num_tokens": 25185369.0,
      "step": 11495
    },
    {
      "entropy": 5.543664932250977,
      "epoch": 1.1413259229853117,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004878111801946933,
      "loss": 5.2414,
      "mean_token_accuracy": 0.16035244315862657,
      "num_tokens": 25197875.0,
      "step": 11500
    },
    {
      "entropy": 5.484067106246949,
      "epoch": 1.1418221516474791,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048779967402358296,
      "loss": 5.1325,
      "mean_token_accuracy": 0.17858723998069764,
      "num_tokens": 25208362.0,
      "step": 11505
    },
    {
      "entropy": 5.611044120788574,
      "epoch": 1.1423183803096466,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0004877881625754993,
      "loss": 5.3243,
      "mean_token_accuracy": 0.15847403258085252,
      "num_tokens": 25220441.0,
      "step": 11510
    },
    {
      "entropy": 5.524960613250732,
      "epoch": 1.1428146089718143,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048777664585072765,
      "loss": 5.0971,
      "mean_token_accuracy": 0.17184018343687057,
      "num_tokens": 25231134.0,
      "step": 11515
    },
    {
      "entropy": 5.637229776382446,
      "epoch": 1.1433108376339818,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004877651238495537,
      "loss": 5.3504,
      "mean_token_accuracy": 0.15819707438349723,
      "num_tokens": 25242327.0,
      "step": 11520
    },
    {
      "entropy": 5.573619413375854,
      "epoch": 1.1438070662961493,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048775359657226316,
      "loss": 5.1936,
      "mean_token_accuracy": 0.16693514585494995,
      "num_tokens": 25253571.0,
      "step": 11525
    },
    {
      "entropy": 5.4986457347869875,
      "epoch": 1.1443032949583167,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004877420640191418,
      "loss": 5.1527,
      "mean_token_accuracy": 0.16866503655910492,
      "num_tokens": 25264573.0,
      "step": 11530
    },
    {
      "entropy": 5.570827388763428,
      "epoch": 1.1447995236204842,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004877305261904758,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1627132326364517,
      "num_tokens": 25275197.0,
      "step": 11535
    },
    {
      "entropy": 5.534627151489258,
      "epoch": 1.145295752282652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048771898308655107,
      "loss": 5.1092,
      "mean_token_accuracy": 0.1760156750679016,
      "num_tokens": 25286564.0,
      "step": 11540
    },
    {
      "entropy": 5.567119979858399,
      "epoch": 1.1457919809448194,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004877074347076539,
      "loss": 5.1634,
      "mean_token_accuracy": 0.16287363469600677,
      "num_tokens": 25297874.0,
      "step": 11545
    },
    {
      "entropy": 5.584882736206055,
      "epoch": 1.1462882096069869,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048769588105407075,
      "loss": 5.232,
      "mean_token_accuracy": 0.16478545665740968,
      "num_tokens": 25308047.0,
      "step": 11550
    },
    {
      "entropy": 5.6375589847564695,
      "epoch": 1.1467844382691545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048768432212608807,
      "loss": 5.2702,
      "mean_token_accuracy": 0.1580920323729515,
      "num_tokens": 25318527.0,
      "step": 11555
    },
    {
      "entropy": 5.552266836166382,
      "epoch": 1.147280666931322,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004876727579239925,
      "loss": 5.1806,
      "mean_token_accuracy": 0.16882868111133575,
      "num_tokens": 25328242.0,
      "step": 11560
    },
    {
      "entropy": 5.5429071426391605,
      "epoch": 1.1477768955934895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004876611884480708,
      "loss": 5.172,
      "mean_token_accuracy": 0.17069576382637025,
      "num_tokens": 25338214.0,
      "step": 11565
    },
    {
      "entropy": 5.563593864440918,
      "epoch": 1.148273124255657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048764961369860977,
      "loss": 5.2109,
      "mean_token_accuracy": 0.16251052021980286,
      "num_tokens": 25349762.0,
      "step": 11570
    },
    {
      "entropy": 5.549267530441284,
      "epoch": 1.1487693529178244,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048763803367589663,
      "loss": 5.2099,
      "mean_token_accuracy": 0.16652089953422547,
      "num_tokens": 25361241.0,
      "step": 11575
    },
    {
      "entropy": 5.591844129562378,
      "epoch": 1.1492655815799921,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048762644838021843,
      "loss": 5.2425,
      "mean_token_accuracy": 0.1604083314538002,
      "num_tokens": 25373846.0,
      "step": 11580
    },
    {
      "entropy": 5.541387033462525,
      "epoch": 1.1497618102421596,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048761485781186243,
      "loss": 5.1595,
      "mean_token_accuracy": 0.1678536430001259,
      "num_tokens": 25385331.0,
      "step": 11585
    },
    {
      "entropy": 5.487461900711059,
      "epoch": 1.150258038904327,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004876032619711162,
      "loss": 5.1826,
      "mean_token_accuracy": 0.1736022263765335,
      "num_tokens": 25395140.0,
      "step": 11590
    },
    {
      "entropy": 5.617384815216065,
      "epoch": 1.1507542675664946,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048759166085826714,
      "loss": 5.229,
      "mean_token_accuracy": 0.1560715064406395,
      "num_tokens": 25405455.0,
      "step": 11595
    },
    {
      "entropy": 5.569248247146606,
      "epoch": 1.1512504962286623,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000487580054473603,
      "loss": 5.2346,
      "mean_token_accuracy": 0.16681553572416305,
      "num_tokens": 25417000.0,
      "step": 11600
    },
    {
      "entropy": 5.4715711116790775,
      "epoch": 1.1517467248908297,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048756844281741156,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17374543398618697,
      "num_tokens": 25427808.0,
      "step": 11605
    },
    {
      "entropy": 5.541740751266479,
      "epoch": 1.1522429535529972,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048755682588998085,
      "loss": 5.1795,
      "mean_token_accuracy": 0.1596010819077492,
      "num_tokens": 25438302.0,
      "step": 11610
    },
    {
      "entropy": 5.465975332260132,
      "epoch": 1.1527391822151647,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004875452036915987,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1697206899523735,
      "num_tokens": 25449335.0,
      "step": 11615
    },
    {
      "entropy": 5.652761650085449,
      "epoch": 1.1532354108773322,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004875335762225536,
      "loss": 5.2644,
      "mean_token_accuracy": 0.15948119908571243,
      "num_tokens": 25461173.0,
      "step": 11620
    },
    {
      "entropy": 5.566653060913086,
      "epoch": 1.1537316395394999,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048752194348313383,
      "loss": 5.1748,
      "mean_token_accuracy": 0.16580819934606553,
      "num_tokens": 25472538.0,
      "step": 11625
    },
    {
      "entropy": 5.512246561050415,
      "epoch": 1.1542278682016673,
      "grad_norm": 0.875,
      "learning_rate": 0.0004875103054736278,
      "loss": 5.1965,
      "mean_token_accuracy": 0.16607460677623748,
      "num_tokens": 25483386.0,
      "step": 11630
    },
    {
      "entropy": 5.594605493545532,
      "epoch": 1.1547240968638348,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004874986621943241,
      "loss": 5.2145,
      "mean_token_accuracy": 0.16811052113771438,
      "num_tokens": 25495505.0,
      "step": 11635
    },
    {
      "entropy": 5.549631881713867,
      "epoch": 1.1552203255260023,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048748701364551147,
      "loss": 5.1989,
      "mean_token_accuracy": 0.16849933862686156,
      "num_tokens": 25506144.0,
      "step": 11640
    },
    {
      "entropy": 5.552102661132812,
      "epoch": 1.15571655418817,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004874753598274787,
      "loss": 5.1913,
      "mean_token_accuracy": 0.1623314380645752,
      "num_tokens": 25516185.0,
      "step": 11645
    },
    {
      "entropy": 5.648763370513916,
      "epoch": 1.1562127828503375,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004874637007405149,
      "loss": 5.2719,
      "mean_token_accuracy": 0.16492184698581697,
      "num_tokens": 25526657.0,
      "step": 11650
    },
    {
      "entropy": 5.538891077041626,
      "epoch": 1.156709011512505,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048745203638490917,
      "loss": 5.1362,
      "mean_token_accuracy": 0.17551746368408203,
      "num_tokens": 25537101.0,
      "step": 11655
    },
    {
      "entropy": 5.557717561721802,
      "epoch": 1.1572052401746724,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048744036676095067,
      "loss": 5.2329,
      "mean_token_accuracy": 0.16107953637838363,
      "num_tokens": 25547672.0,
      "step": 11660
    },
    {
      "entropy": 5.624847745895385,
      "epoch": 1.15770146883684,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874286918689289,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1660443887114525,
      "num_tokens": 25558408.0,
      "step": 11665
    },
    {
      "entropy": 5.5998766899108885,
      "epoch": 1.1581976974990076,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048741701170913326,
      "loss": 5.1698,
      "mean_token_accuracy": 0.16549035161733627,
      "num_tokens": 25569823.0,
      "step": 11670
    },
    {
      "entropy": 5.53923020362854,
      "epoch": 1.158693926161175,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048740532628185353,
      "loss": 5.1806,
      "mean_token_accuracy": 0.16889911592006684,
      "num_tokens": 25579705.0,
      "step": 11675
    },
    {
      "entropy": 5.554683828353882,
      "epoch": 1.1591901548233425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004873936355873793,
      "loss": 5.2593,
      "mean_token_accuracy": 0.16012637466192245,
      "num_tokens": 25590573.0,
      "step": 11680
    },
    {
      "entropy": 5.557940149307251,
      "epoch": 1.1596863834855102,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004873819396260007,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1703796088695526,
      "num_tokens": 25601418.0,
      "step": 11685
    },
    {
      "entropy": 5.551283645629883,
      "epoch": 1.1601826121476777,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048737023839800754,
      "loss": 5.2396,
      "mean_token_accuracy": 0.16557013988494873,
      "num_tokens": 25612409.0,
      "step": 11690
    },
    {
      "entropy": 5.536490821838379,
      "epoch": 1.1606788408098452,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004873585319036901,
      "loss": 5.143,
      "mean_token_accuracy": 0.16696803122758866,
      "num_tokens": 25623547.0,
      "step": 11695
    },
    {
      "entropy": 5.59570541381836,
      "epoch": 1.1611750694720127,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004873468201433387,
      "loss": 5.3194,
      "mean_token_accuracy": 0.16170730739831923,
      "num_tokens": 25636537.0,
      "step": 11700
    },
    {
      "entropy": 5.567227935791015,
      "epoch": 1.1616712981341801,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004873351031172437,
      "loss": 5.1597,
      "mean_token_accuracy": 0.17085832059383393,
      "num_tokens": 25646734.0,
      "step": 11705
    },
    {
      "entropy": 5.61855115890503,
      "epoch": 1.1621675267963478,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048732338082569564,
      "loss": 5.2529,
      "mean_token_accuracy": 0.16774247884750365,
      "num_tokens": 25658058.0,
      "step": 11710
    },
    {
      "entropy": 5.529183292388916,
      "epoch": 1.1626637554585153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004873116532689853,
      "loss": 5.091,
      "mean_token_accuracy": 0.17300632148981093,
      "num_tokens": 25669630.0,
      "step": 11715
    },
    {
      "entropy": 5.480577182769776,
      "epoch": 1.1631599841206828,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872999204474035,
      "loss": 5.112,
      "mean_token_accuracy": 0.16797398328781127,
      "num_tokens": 25681928.0,
      "step": 11720
    },
    {
      "entropy": 5.450588989257812,
      "epoch": 1.1636562127828503,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487288182361241,
      "loss": 5.1239,
      "mean_token_accuracy": 0.16811102628707886,
      "num_tokens": 25690653.0,
      "step": 11725
    },
    {
      "entropy": 5.585180568695068,
      "epoch": 1.164152441445018,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004872764390107891,
      "loss": 5.2264,
      "mean_token_accuracy": 0.16265956163406373,
      "num_tokens": 25701201.0,
      "step": 11730
    },
    {
      "entropy": 5.505923223495484,
      "epoch": 1.1646486701071854,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048726469039633895,
      "loss": 5.1372,
      "mean_token_accuracy": 0.17007317394018173,
      "num_tokens": 25710927.0,
      "step": 11735
    },
    {
      "entropy": 5.575149059295654,
      "epoch": 1.165144898769353,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048725293651818183,
      "loss": 5.1836,
      "mean_token_accuracy": 0.16883752197027208,
      "num_tokens": 25722136.0,
      "step": 11740
    },
    {
      "entropy": 5.535172462463379,
      "epoch": 1.1656411274315204,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004872411773766093,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17482638508081436,
      "num_tokens": 25733386.0,
      "step": 11745
    },
    {
      "entropy": 5.508816528320312,
      "epoch": 1.1661373560936878,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048722941297191284,
      "loss": 5.1758,
      "mean_token_accuracy": 0.17097684144973754,
      "num_tokens": 25742801.0,
      "step": 11750
    },
    {
      "entropy": 5.634187936782837,
      "epoch": 1.1666335847558555,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004872176433043843,
      "loss": 5.2634,
      "mean_token_accuracy": 0.16353974491357803,
      "num_tokens": 25755594.0,
      "step": 11755
    },
    {
      "entropy": 5.574434423446656,
      "epoch": 1.167129813418023,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004872058683743155,
      "loss": 5.1991,
      "mean_token_accuracy": 0.16636495441198348,
      "num_tokens": 25766271.0,
      "step": 11760
    },
    {
      "entropy": 5.5830302238464355,
      "epoch": 1.1676260420801905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048719408818199836,
      "loss": 5.244,
      "mean_token_accuracy": 0.1652930870652199,
      "num_tokens": 25777338.0,
      "step": 11765
    },
    {
      "entropy": 5.5581199645996096,
      "epoch": 1.1681222707423582,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048718230272772517,
      "loss": 5.1228,
      "mean_token_accuracy": 0.16831648647785186,
      "num_tokens": 25787968.0,
      "step": 11770
    },
    {
      "entropy": 5.572164487838745,
      "epoch": 1.1686184994045257,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048717051201178804,
      "loss": 5.2154,
      "mean_token_accuracy": 0.16895880103111266,
      "num_tokens": 25798274.0,
      "step": 11775
    },
    {
      "entropy": 5.496921396255493,
      "epoch": 1.1691147280666931,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004871587160344794,
      "loss": 5.1758,
      "mean_token_accuracy": 0.15743285417556763,
      "num_tokens": 25809140.0,
      "step": 11780
    },
    {
      "entropy": 5.587840747833252,
      "epoch": 1.1696109567288606,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004871469147960917,
      "loss": 5.2088,
      "mean_token_accuracy": 0.16665082424879074,
      "num_tokens": 25821502.0,
      "step": 11785
    },
    {
      "entropy": 5.561322736740112,
      "epoch": 1.170107185391028,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004871351082969177,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1721177577972412,
      "num_tokens": 25833231.0,
      "step": 11790
    },
    {
      "entropy": 5.592991495132447,
      "epoch": 1.1706034140531958,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004871232965372501,
      "loss": 5.2456,
      "mean_token_accuracy": 0.16383754462003708,
      "num_tokens": 25844867.0,
      "step": 11795
    },
    {
      "entropy": 5.563653039932251,
      "epoch": 1.1710996427153633,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004871114795173819,
      "loss": 5.23,
      "mean_token_accuracy": 0.17096350491046905,
      "num_tokens": 25855959.0,
      "step": 11800
    },
    {
      "entropy": 5.617068481445313,
      "epoch": 1.1715958713775307,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004870996572376061,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1634581446647644,
      "num_tokens": 25866926.0,
      "step": 11805
    },
    {
      "entropy": 5.626061153411865,
      "epoch": 1.1720921000396982,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004870878296982157,
      "loss": 5.2376,
      "mean_token_accuracy": 0.16707187741994858,
      "num_tokens": 25877699.0,
      "step": 11810
    },
    {
      "entropy": 5.631240701675415,
      "epoch": 1.172588328701866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004870759968995042,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1688520312309265,
      "num_tokens": 25888640.0,
      "step": 11815
    },
    {
      "entropy": 5.53049602508545,
      "epoch": 1.1730845573640334,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048706415884176493,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1651050016283989,
      "num_tokens": 25900651.0,
      "step": 11820
    },
    {
      "entropy": 5.53585205078125,
      "epoch": 1.1735807860262009,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004870523155252915,
      "loss": 5.1608,
      "mean_token_accuracy": 0.16988168358802797,
      "num_tokens": 25911775.0,
      "step": 11825
    },
    {
      "entropy": 5.546398544311524,
      "epoch": 1.1740770146883683,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004870404669503776,
      "loss": 5.1268,
      "mean_token_accuracy": 0.17013687938451766,
      "num_tokens": 25922126.0,
      "step": 11830
    },
    {
      "entropy": 5.603577375411987,
      "epoch": 1.1745732433505358,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000487028613117317,
      "loss": 5.1874,
      "mean_token_accuracy": 0.16847057342529298,
      "num_tokens": 25932945.0,
      "step": 11835
    },
    {
      "entropy": 5.6142493724823,
      "epoch": 1.1750694720127035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004870167540264037,
      "loss": 5.2506,
      "mean_token_accuracy": 0.16007343232631682,
      "num_tokens": 25944277.0,
      "step": 11840
    },
    {
      "entropy": 5.487193775177002,
      "epoch": 1.175565700674871,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004870048896779317,
      "loss": 5.1177,
      "mean_token_accuracy": 0.1748425081372261,
      "num_tokens": 25956838.0,
      "step": 11845
    },
    {
      "entropy": 5.534640550613403,
      "epoch": 1.1760619293370385,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048699302007219533,
      "loss": 5.1751,
      "mean_token_accuracy": 0.1639681041240692,
      "num_tokens": 25966639.0,
      "step": 11850
    },
    {
      "entropy": 5.555876302719116,
      "epoch": 1.1765581579992062,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004869811452094889,
      "loss": 5.2107,
      "mean_token_accuracy": 0.17486705034971237,
      "num_tokens": 25978209.0,
      "step": 11855
    },
    {
      "entropy": 5.508122348785401,
      "epoch": 1.1770543866613736,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048696926509010673,
      "loss": 5.1534,
      "mean_token_accuracy": 0.17129939198493957,
      "num_tokens": 25989306.0,
      "step": 11860
    },
    {
      "entropy": 5.519138336181641,
      "epoch": 1.177550615323541,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004869573797143437,
      "loss": 5.2363,
      "mean_token_accuracy": 0.16127496510744094,
      "num_tokens": 26000235.0,
      "step": 11865
    },
    {
      "entropy": 5.533761739730835,
      "epoch": 1.1780468439857086,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004869454890824942,
      "loss": 5.1487,
      "mean_token_accuracy": 0.16898459941148758,
      "num_tokens": 26010616.0,
      "step": 11870
    },
    {
      "entropy": 5.574197101593017,
      "epoch": 1.178543072647876,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00048693359319485337,
      "loss": 5.2159,
      "mean_token_accuracy": 0.15998696237802507,
      "num_tokens": 26021303.0,
      "step": 11875
    },
    {
      "entropy": 5.53950834274292,
      "epoch": 1.1790393013100438,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004869216920517161,
      "loss": 5.1464,
      "mean_token_accuracy": 0.16840660572052002,
      "num_tokens": 26032386.0,
      "step": 11880
    },
    {
      "entropy": 5.462571096420288,
      "epoch": 1.1795355299722112,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048690978565337753,
      "loss": 5.1044,
      "mean_token_accuracy": 0.16511938124895095,
      "num_tokens": 26042889.0,
      "step": 11885
    },
    {
      "entropy": 5.532870578765869,
      "epoch": 1.1800317586343787,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048689787400013287,
      "loss": 5.2149,
      "mean_token_accuracy": 0.16369595974683762,
      "num_tokens": 26054299.0,
      "step": 11890
    },
    {
      "entropy": 5.640064001083374,
      "epoch": 1.1805279872965462,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004868859570922775,
      "loss": 5.256,
      "mean_token_accuracy": 0.16323316544294358,
      "num_tokens": 26064587.0,
      "step": 11895
    },
    {
      "entropy": 5.603106069564819,
      "epoch": 1.1810242159587139,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004868740349301071,
      "loss": 5.1792,
      "mean_token_accuracy": 0.17820848971605302,
      "num_tokens": 26076495.0,
      "step": 11900
    },
    {
      "entropy": 5.542065668106079,
      "epoch": 1.1815204446208813,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000486862107513917,
      "loss": 5.191,
      "mean_token_accuracy": 0.16448932141065598,
      "num_tokens": 26086496.0,
      "step": 11905
    },
    {
      "entropy": 5.520664834976197,
      "epoch": 1.1820166732830488,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004868501748440033,
      "loss": 5.1686,
      "mean_token_accuracy": 0.1638784795999527,
      "num_tokens": 26097263.0,
      "step": 11910
    },
    {
      "entropy": 5.5795886516571045,
      "epoch": 1.1825129019452163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004868382369206617,
      "loss": 5.1793,
      "mean_token_accuracy": 0.16616731733083726,
      "num_tokens": 26108264.0,
      "step": 11915
    },
    {
      "entropy": 5.576518821716308,
      "epoch": 1.1830091306073838,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004868262937441883,
      "loss": 5.2434,
      "mean_token_accuracy": 0.17230187803506852,
      "num_tokens": 26119730.0,
      "step": 11920
    },
    {
      "entropy": 5.559617614746093,
      "epoch": 1.1835053592695515,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048681434531487925,
      "loss": 5.1491,
      "mean_token_accuracy": 0.16586946099996566,
      "num_tokens": 26131055.0,
      "step": 11925
    },
    {
      "entropy": 5.4719689846038815,
      "epoch": 1.184001587931719,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048680239163303085,
      "loss": 5.1336,
      "mean_token_accuracy": 0.1763986498117447,
      "num_tokens": 26141577.0,
      "step": 11930
    },
    {
      "entropy": 5.554886674880981,
      "epoch": 1.1844978165938864,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004867904326989396,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1657906174659729,
      "num_tokens": 26152733.0,
      "step": 11935
    },
    {
      "entropy": 5.608040142059326,
      "epoch": 1.184994045256054,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004867784685129019,
      "loss": 5.2227,
      "mean_token_accuracy": 0.16342762112617493,
      "num_tokens": 26164763.0,
      "step": 11940
    },
    {
      "entropy": 5.522825860977173,
      "epoch": 1.1854902739182216,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048676649907521455,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17121067643165588,
      "num_tokens": 26174391.0,
      "step": 11945
    },
    {
      "entropy": 5.442882394790649,
      "epoch": 1.185986502580389,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048675452438617434,
      "loss": 5.0739,
      "mean_token_accuracy": 0.17718868851661682,
      "num_tokens": 26184772.0,
      "step": 11950
    },
    {
      "entropy": 5.6077008724212645,
      "epoch": 1.1864827312425565,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004867425444460782,
      "loss": 5.2382,
      "mean_token_accuracy": 0.15622446835041046,
      "num_tokens": 26195411.0,
      "step": 11955
    },
    {
      "entropy": 5.532578325271606,
      "epoch": 1.186978959904724,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004867305592552233,
      "loss": 5.1528,
      "mean_token_accuracy": 0.1711677059531212,
      "num_tokens": 26206719.0,
      "step": 11960
    },
    {
      "entropy": 5.500160360336304,
      "epoch": 1.1874751885668917,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004867185688139067,
      "loss": 5.1069,
      "mean_token_accuracy": 0.17365608215332032,
      "num_tokens": 26217469.0,
      "step": 11965
    },
    {
      "entropy": 5.452420663833618,
      "epoch": 1.1879714172290592,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004867065731224258,
      "loss": 5.1236,
      "mean_token_accuracy": 0.16789592653512955,
      "num_tokens": 26228556.0,
      "step": 11970
    },
    {
      "entropy": 5.534000635147095,
      "epoch": 1.1884676458912267,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004866945721810781,
      "loss": 5.2326,
      "mean_token_accuracy": 0.16941561698913574,
      "num_tokens": 26239891.0,
      "step": 11975
    },
    {
      "entropy": 5.556078386306763,
      "epoch": 1.1889638745533941,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004866825659901612,
      "loss": 5.185,
      "mean_token_accuracy": 0.17289711833000182,
      "num_tokens": 26251793.0,
      "step": 11980
    },
    {
      "entropy": 5.47785964012146,
      "epoch": 1.1894601032155618,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048667055454997276,
      "loss": 5.0839,
      "mean_token_accuracy": 0.16918738335371017,
      "num_tokens": 26262367.0,
      "step": 11985
    },
    {
      "entropy": 5.524373245239258,
      "epoch": 1.1899563318777293,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048665853786081063,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16755572259426116,
      "num_tokens": 26272681.0,
      "step": 11990
    },
    {
      "entropy": 5.5546791553497314,
      "epoch": 1.1904525605398968,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048664651592297294,
      "loss": 5.2363,
      "mean_token_accuracy": 0.1671360492706299,
      "num_tokens": 26281194.0,
      "step": 11995
    },
    {
      "entropy": 5.56120491027832,
      "epoch": 1.1909487892020643,
      "grad_norm": 0.875,
      "learning_rate": 0.00048663448873675766,
      "loss": 5.1272,
      "mean_token_accuracy": 0.16934439539909363,
      "num_tokens": 26292682.0,
      "step": 12000
    },
    {
      "epoch": 1.1909487892020643,
      "eval_entropy": 5.414201732938209,
      "eval_loss": 5.304666042327881,
      "eval_mean_token_accuracy": 0.16682653438346803,
      "eval_num_tokens": 26292682.0,
      "eval_runtime": 20.5344,
      "eval_samples_per_second": 1639.638,
      "eval_steps_per_second": 204.973,
      "step": 12000
    },
    {
      "entropy": 5.5849744319915775,
      "epoch": 1.1914450178642317,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004866224563024631,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16255748718976976,
      "num_tokens": 26305500.0,
      "step": 12005
    },
    {
      "entropy": 5.553773212432861,
      "epoch": 1.1919412465263994,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004866104186203877,
      "loss": 5.1336,
      "mean_token_accuracy": 0.16356791704893112,
      "num_tokens": 26317207.0,
      "step": 12010
    },
    {
      "entropy": 5.54699330329895,
      "epoch": 1.192437475188567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048659837569082977,
      "loss": 5.1451,
      "mean_token_accuracy": 0.17224695682525634,
      "num_tokens": 26328728.0,
      "step": 12015
    },
    {
      "entropy": 5.598548412322998,
      "epoch": 1.1929337038507344,
      "grad_norm": 1.0,
      "learning_rate": 0.0004865863275140882,
      "loss": 5.1923,
      "mean_token_accuracy": 0.17023126482963563,
      "num_tokens": 26338622.0,
      "step": 12020
    },
    {
      "entropy": 5.581063365936279,
      "epoch": 1.1934299325129019,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004865742740904615,
      "loss": 5.1759,
      "mean_token_accuracy": 0.16627230048179625,
      "num_tokens": 26348807.0,
      "step": 12025
    },
    {
      "entropy": 5.499698305130005,
      "epoch": 1.1939261611750696,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004865622154202488,
      "loss": 5.1297,
      "mean_token_accuracy": 0.17258941382169724,
      "num_tokens": 26358481.0,
      "step": 12030
    },
    {
      "entropy": 5.549474573135376,
      "epoch": 1.194422389837237,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048655015150374904,
      "loss": 5.2079,
      "mean_token_accuracy": 0.16275927275419236,
      "num_tokens": 26370519.0,
      "step": 12035
    },
    {
      "entropy": 5.592999124526978,
      "epoch": 1.1949186184994045,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004865380823412613,
      "loss": 5.2059,
      "mean_token_accuracy": 0.16864195764064788,
      "num_tokens": 26380988.0,
      "step": 12040
    },
    {
      "entropy": 5.514132308959961,
      "epoch": 1.195414847161572,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000486526007933085,
      "loss": 5.1029,
      "mean_token_accuracy": 0.17362844198942184,
      "num_tokens": 26390308.0,
      "step": 12045
    },
    {
      "entropy": 5.563156223297119,
      "epoch": 1.1959110758237395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048651392827951944,
      "loss": 5.2264,
      "mean_token_accuracy": 0.1584816172719002,
      "num_tokens": 26401145.0,
      "step": 12050
    },
    {
      "entropy": 5.5552891254425045,
      "epoch": 1.1964073044859072,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048650184338086436,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16377270519733428,
      "num_tokens": 26411930.0,
      "step": 12055
    },
    {
      "entropy": 5.5329185962677006,
      "epoch": 1.1969035331480746,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048648975323741916,
      "loss": 5.1034,
      "mean_token_accuracy": 0.16410244554281234,
      "num_tokens": 26422320.0,
      "step": 12060
    },
    {
      "entropy": 5.646494770050049,
      "epoch": 1.197399761810242,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048647765784948395,
      "loss": 5.2335,
      "mean_token_accuracy": 0.16120383888483047,
      "num_tokens": 26433760.0,
      "step": 12065
    },
    {
      "entropy": 5.604353284835815,
      "epoch": 1.1978959904724098,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004864655572173584,
      "loss": 5.1475,
      "mean_token_accuracy": 0.17304044514894484,
      "num_tokens": 26444256.0,
      "step": 12070
    },
    {
      "entropy": 5.531704473495483,
      "epoch": 1.1983922191345773,
      "grad_norm": 0.875,
      "learning_rate": 0.00048645345134134274,
      "loss": 5.234,
      "mean_token_accuracy": 0.17082027941942216,
      "num_tokens": 26456433.0,
      "step": 12075
    },
    {
      "entropy": 5.593945074081421,
      "epoch": 1.1988884477967447,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048644134022173707,
      "loss": 5.2045,
      "mean_token_accuracy": 0.16808937042951583,
      "num_tokens": 26466909.0,
      "step": 12080
    },
    {
      "entropy": 5.59465651512146,
      "epoch": 1.1993846764589122,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004864292238588419,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1706068903207779,
      "num_tokens": 26478785.0,
      "step": 12085
    },
    {
      "entropy": 5.538610792160034,
      "epoch": 1.1998809051210797,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048641710225295746,
      "loss": 5.2172,
      "mean_token_accuracy": 0.1649825766682625,
      "num_tokens": 26489596.0,
      "step": 12090
    },
    {
      "entropy": 5.510122871398925,
      "epoch": 1.2003771337832474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048640497540438455,
      "loss": 5.1419,
      "mean_token_accuracy": 0.1724795088171959,
      "num_tokens": 26500376.0,
      "step": 12095
    },
    {
      "entropy": 5.514937829971314,
      "epoch": 1.2008733624454149,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004863928433134237,
      "loss": 5.0938,
      "mean_token_accuracy": 0.1706473022699356,
      "num_tokens": 26510339.0,
      "step": 12100
    },
    {
      "entropy": 5.5595794200897215,
      "epoch": 1.2013695911075823,
      "grad_norm": 0.859375,
      "learning_rate": 0.00048638070598037585,
      "loss": 5.191,
      "mean_token_accuracy": 0.16458145380020142,
      "num_tokens": 26521359.0,
      "step": 12105
    },
    {
      "entropy": 5.603921556472779,
      "epoch": 1.2018658197697498,
      "grad_norm": 1.03125,
      "learning_rate": 0.000486368563405542,
      "loss": 5.2614,
      "mean_token_accuracy": 0.16664886921644212,
      "num_tokens": 26532716.0,
      "step": 12110
    },
    {
      "entropy": 5.508515930175781,
      "epoch": 1.2023620484319175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004863564155892232,
      "loss": 5.1021,
      "mean_token_accuracy": 0.16863590925931932,
      "num_tokens": 26542624.0,
      "step": 12115
    },
    {
      "entropy": 5.599541330337525,
      "epoch": 1.202858277094085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004863442625317208,
      "loss": 5.2112,
      "mean_token_accuracy": 0.16777480244636536,
      "num_tokens": 26552907.0,
      "step": 12120
    },
    {
      "entropy": 5.51020450592041,
      "epoch": 1.2033545057562525,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048633210423333607,
      "loss": 5.1634,
      "mean_token_accuracy": 0.1694307416677475,
      "num_tokens": 26564871.0,
      "step": 12125
    },
    {
      "entropy": 5.502919864654541,
      "epoch": 1.20385073441842,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004863199406943705,
      "loss": 5.1607,
      "mean_token_accuracy": 0.16642336547374725,
      "num_tokens": 26576254.0,
      "step": 12130
    },
    {
      "entropy": 5.570493936538696,
      "epoch": 1.2043469630805874,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048630777191512584,
      "loss": 5.1612,
      "mean_token_accuracy": 0.17370194643735887,
      "num_tokens": 26587518.0,
      "step": 12135
    },
    {
      "entropy": 5.524411582946778,
      "epoch": 1.2048431917427551,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048629559789590366,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17298709452152253,
      "num_tokens": 26597485.0,
      "step": 12140
    },
    {
      "entropy": 5.514684438705444,
      "epoch": 1.2053394204049226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000486283418637006,
      "loss": 5.1372,
      "mean_token_accuracy": 0.1743522822856903,
      "num_tokens": 26607696.0,
      "step": 12145
    },
    {
      "entropy": 5.567519044876098,
      "epoch": 1.20583564906709,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004862712341387348,
      "loss": 5.1703,
      "mean_token_accuracy": 0.1669094979763031,
      "num_tokens": 26618974.0,
      "step": 12150
    },
    {
      "entropy": 5.602524089813232,
      "epoch": 1.2063318777292578,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004862590444013923,
      "loss": 5.2194,
      "mean_token_accuracy": 0.16477175503969194,
      "num_tokens": 26629949.0,
      "step": 12155
    },
    {
      "entropy": 5.578515720367432,
      "epoch": 1.2068281063914252,
      "grad_norm": 0.875,
      "learning_rate": 0.0004862468494252806,
      "loss": 5.161,
      "mean_token_accuracy": 0.1681686148047447,
      "num_tokens": 26641460.0,
      "step": 12160
    },
    {
      "entropy": 5.578083133697509,
      "epoch": 1.2073243350535927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004862346492107024,
      "loss": 5.2465,
      "mean_token_accuracy": 0.16091246753931046,
      "num_tokens": 26652553.0,
      "step": 12165
    },
    {
      "entropy": 5.506343650817871,
      "epoch": 1.2078205637157602,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048622244375796,
      "loss": 5.1641,
      "mean_token_accuracy": 0.1625804618000984,
      "num_tokens": 26662479.0,
      "step": 12170
    },
    {
      "entropy": 5.57989993095398,
      "epoch": 1.2083167923779277,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004862102330673561,
      "loss": 5.2157,
      "mean_token_accuracy": 0.1617538034915924,
      "num_tokens": 26673326.0,
      "step": 12175
    },
    {
      "entropy": 5.60931658744812,
      "epoch": 1.2088130210400954,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048619801713919357,
      "loss": 5.2592,
      "mean_token_accuracy": 0.15975237935781478,
      "num_tokens": 26684189.0,
      "step": 12180
    },
    {
      "entropy": 5.636198043823242,
      "epoch": 1.2093092497022628,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004861857959737753,
      "loss": 5.2661,
      "mean_token_accuracy": 0.16838395297527314,
      "num_tokens": 26695399.0,
      "step": 12185
    },
    {
      "entropy": 5.623047590255737,
      "epoch": 1.2098054783644303,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861735695714044,
      "loss": 5.1969,
      "mean_token_accuracy": 0.16281272768974303,
      "num_tokens": 26705993.0,
      "step": 12190
    },
    {
      "entropy": 5.56933913230896,
      "epoch": 1.2103017070265978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048616133793238394,
      "loss": 5.1674,
      "mean_token_accuracy": 0.1705382525920868,
      "num_tokens": 26716120.0,
      "step": 12195
    },
    {
      "entropy": 5.602939414978027,
      "epoch": 1.2107979356887655,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048614910105701737,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1544528305530548,
      "num_tokens": 26727736.0,
      "step": 12200
    },
    {
      "entropy": 5.597620153427124,
      "epoch": 1.211294164350933,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004861368589456081,
      "loss": 5.1466,
      "mean_token_accuracy": 0.17210688143968583,
      "num_tokens": 26737825.0,
      "step": 12205
    },
    {
      "entropy": 5.632390689849854,
      "epoch": 1.2117903930131004,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004861246115984596,
      "loss": 5.2431,
      "mean_token_accuracy": 0.16130630820989608,
      "num_tokens": 26748035.0,
      "step": 12210
    },
    {
      "entropy": 5.62108039855957,
      "epoch": 1.212286621675268,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00048611235901587575,
      "loss": 5.1446,
      "mean_token_accuracy": 0.17094327360391617,
      "num_tokens": 26759899.0,
      "step": 12215
    },
    {
      "entropy": 5.538790559768676,
      "epoch": 1.2127828503374354,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004861001011981602,
      "loss": 5.114,
      "mean_token_accuracy": 0.17405298203229905,
      "num_tokens": 26771562.0,
      "step": 12220
    },
    {
      "entropy": 5.592156267166137,
      "epoch": 1.213279078999603,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004860878381456171,
      "loss": 5.2338,
      "mean_token_accuracy": 0.17127795964479448,
      "num_tokens": 26783002.0,
      "step": 12225
    },
    {
      "entropy": 5.529238748550415,
      "epoch": 1.2137753076617706,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048607556985855047,
      "loss": 5.2131,
      "mean_token_accuracy": 0.17079030871391296,
      "num_tokens": 26794136.0,
      "step": 12230
    },
    {
      "entropy": 5.4902318000793455,
      "epoch": 1.214271536323938,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048606329633726447,
      "loss": 5.182,
      "mean_token_accuracy": 0.16759084463119506,
      "num_tokens": 26804776.0,
      "step": 12235
    },
    {
      "entropy": 5.511506509780884,
      "epoch": 1.2147677649861055,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004860510175820636,
      "loss": 5.1141,
      "mean_token_accuracy": 0.17307940274477004,
      "num_tokens": 26816696.0,
      "step": 12240
    },
    {
      "entropy": 5.505301904678345,
      "epoch": 1.2152639936482732,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004860387335932522,
      "loss": 5.1628,
      "mean_token_accuracy": 0.16943667083978653,
      "num_tokens": 26827789.0,
      "step": 12245
    },
    {
      "entropy": 5.599620819091797,
      "epoch": 1.2157602223104407,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004860264443711351,
      "loss": 5.1253,
      "mean_token_accuracy": 0.16838203221559525,
      "num_tokens": 26838763.0,
      "step": 12250
    },
    {
      "entropy": 5.500844097137451,
      "epoch": 1.2162564509726081,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004860141499160167,
      "loss": 5.1567,
      "mean_token_accuracy": 0.1611418232321739,
      "num_tokens": 26849304.0,
      "step": 12255
    },
    {
      "entropy": 5.536780452728271,
      "epoch": 1.2167526796347756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004860018502282022,
      "loss": 5.2208,
      "mean_token_accuracy": 0.16770486682653427,
      "num_tokens": 26858777.0,
      "step": 12260
    },
    {
      "entropy": 5.616679096221924,
      "epoch": 1.2172489082969433,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00048598954530799644,
      "loss": 5.1838,
      "mean_token_accuracy": 0.1628362149000168,
      "num_tokens": 26870286.0,
      "step": 12265
    },
    {
      "entropy": 5.6162299633026125,
      "epoch": 1.2177451369591108,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048597723515570465,
      "loss": 5.1614,
      "mean_token_accuracy": 0.17089560031890869,
      "num_tokens": 26880485.0,
      "step": 12270
    },
    {
      "entropy": 5.451477670669556,
      "epoch": 1.2182413656212783,
      "grad_norm": 0.8125,
      "learning_rate": 0.000485964919771632,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17539387345314025,
      "num_tokens": 26893556.0,
      "step": 12275
    },
    {
      "entropy": 5.531761598587036,
      "epoch": 1.2187375942834457,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048595259915608396,
      "loss": 5.1579,
      "mean_token_accuracy": 0.16286447048187255,
      "num_tokens": 26905699.0,
      "step": 12280
    },
    {
      "entropy": 5.566030502319336,
      "epoch": 1.2192338229456134,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048594027330936594,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1677615836262703,
      "num_tokens": 26917188.0,
      "step": 12285
    },
    {
      "entropy": 5.5651613712310795,
      "epoch": 1.219730051607781,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048592794223178374,
      "loss": 5.1598,
      "mean_token_accuracy": 0.16558319330215454,
      "num_tokens": 26927905.0,
      "step": 12290
    },
    {
      "entropy": 5.461297655105591,
      "epoch": 1.2202262802699484,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048591560592364314,
      "loss": 5.1362,
      "mean_token_accuracy": 0.1729421690106392,
      "num_tokens": 26938665.0,
      "step": 12295
    },
    {
      "entropy": 5.551711702346802,
      "epoch": 1.2207225089321159,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004859032643852499,
      "loss": 5.2336,
      "mean_token_accuracy": 0.16760626584291458,
      "num_tokens": 26950516.0,
      "step": 12300
    },
    {
      "entropy": 5.633113002777099,
      "epoch": 1.2212187375942833,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004858909176169102,
      "loss": 5.224,
      "mean_token_accuracy": 0.17405097782611847,
      "num_tokens": 26960018.0,
      "step": 12305
    },
    {
      "entropy": 5.568445873260498,
      "epoch": 1.221714966256451,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048587856561893014,
      "loss": 5.1871,
      "mean_token_accuracy": 0.16415326595306395,
      "num_tokens": 26969825.0,
      "step": 12310
    },
    {
      "entropy": 5.612930393218994,
      "epoch": 1.2222111949186185,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004858662083916161,
      "loss": 5.2816,
      "mean_token_accuracy": 0.16341308802366256,
      "num_tokens": 26979873.0,
      "step": 12315
    },
    {
      "entropy": 5.629591846466065,
      "epoch": 1.222707423580786,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004858538459352744,
      "loss": 5.2507,
      "mean_token_accuracy": 0.16443151906132697,
      "num_tokens": 26990577.0,
      "step": 12320
    },
    {
      "entropy": 5.571669816970825,
      "epoch": 1.2232036522429535,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004858414782502117,
      "loss": 5.0997,
      "mean_token_accuracy": 0.17373906821012497,
      "num_tokens": 27001598.0,
      "step": 12325
    },
    {
      "entropy": 5.520775318145752,
      "epoch": 1.2236998809051212,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048582910533673474,
      "loss": 5.1037,
      "mean_token_accuracy": 0.1740128368139267,
      "num_tokens": 27012535.0,
      "step": 12330
    },
    {
      "entropy": 5.550115776062012,
      "epoch": 1.2241961095672886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004858167271951502,
      "loss": 5.1868,
      "mean_token_accuracy": 0.15938595980405806,
      "num_tokens": 27021867.0,
      "step": 12335
    },
    {
      "entropy": 5.515055227279663,
      "epoch": 1.2246923382294561,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004858043438257651,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17048090994358062,
      "num_tokens": 27032540.0,
      "step": 12340
    },
    {
      "entropy": 5.490121221542358,
      "epoch": 1.2251885668916236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004857919552288865,
      "loss": 5.1208,
      "mean_token_accuracy": 0.1718556210398674,
      "num_tokens": 27044008.0,
      "step": 12345
    },
    {
      "entropy": 5.5762200355529785,
      "epoch": 1.225684795553791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004857795614048217,
      "loss": 5.1532,
      "mean_token_accuracy": 0.1708926096558571,
      "num_tokens": 27053646.0,
      "step": 12350
    },
    {
      "entropy": 5.62473430633545,
      "epoch": 1.2261810242159588,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048576716235387786,
      "loss": 5.2765,
      "mean_token_accuracy": 0.16109593212604523,
      "num_tokens": 27065834.0,
      "step": 12355
    },
    {
      "entropy": 5.569438838958741,
      "epoch": 1.2266772528781262,
      "grad_norm": 1.0,
      "learning_rate": 0.00048575475807636267,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16590052396059035,
      "num_tokens": 27077544.0,
      "step": 12360
    },
    {
      "entropy": 5.62896409034729,
      "epoch": 1.2271734815402937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048574234857258356,
      "loss": 5.2601,
      "mean_token_accuracy": 0.16581766158342362,
      "num_tokens": 27088294.0,
      "step": 12365
    },
    {
      "entropy": 5.580973100662232,
      "epoch": 1.2276697102024614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048572993384284833,
      "loss": 5.1976,
      "mean_token_accuracy": 0.16449857056140899,
      "num_tokens": 27098848.0,
      "step": 12370
    },
    {
      "entropy": 5.511201238632202,
      "epoch": 1.2281659388646289,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004857175138874648,
      "loss": 5.1546,
      "mean_token_accuracy": 0.16925866901874542,
      "num_tokens": 27112319.0,
      "step": 12375
    },
    {
      "entropy": 5.569800758361817,
      "epoch": 1.2286621675267964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048570508870674097,
      "loss": 5.2245,
      "mean_token_accuracy": 0.1722547009587288,
      "num_tokens": 27121840.0,
      "step": 12380
    },
    {
      "entropy": 5.572824764251709,
      "epoch": 1.2291583961889638,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00048569265830098493,
      "loss": 5.2261,
      "mean_token_accuracy": 0.1685919389128685,
      "num_tokens": 27133619.0,
      "step": 12385
    },
    {
      "entropy": 5.55884690284729,
      "epoch": 1.2296546248511313,
      "grad_norm": 0.953125,
      "learning_rate": 0.000485680222670505,
      "loss": 5.2015,
      "mean_token_accuracy": 0.16535966396331786,
      "num_tokens": 27144161.0,
      "step": 12390
    },
    {
      "entropy": 5.550874900817871,
      "epoch": 1.230150853513299,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004856677818156096,
      "loss": 5.1314,
      "mean_token_accuracy": 0.1696668431162834,
      "num_tokens": 27154906.0,
      "step": 12395
    },
    {
      "entropy": 5.533469009399414,
      "epoch": 1.2306470821754665,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048565533573660707,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17145265340805055,
      "num_tokens": 27165746.0,
      "step": 12400
    },
    {
      "entropy": 5.515801048278808,
      "epoch": 1.231143310837634,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004856428844338062,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17261476069688797,
      "num_tokens": 27176108.0,
      "step": 12405
    },
    {
      "entropy": 5.539755249023438,
      "epoch": 1.2316395394998014,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004856304279075156,
      "loss": 5.2076,
      "mean_token_accuracy": 0.1685943767428398,
      "num_tokens": 27186721.0,
      "step": 12410
    },
    {
      "entropy": 5.532007312774658,
      "epoch": 1.2321357681619691,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004856179661580444,
      "loss": 5.0885,
      "mean_token_accuracy": 0.17227238416671753,
      "num_tokens": 27196120.0,
      "step": 12415
    },
    {
      "entropy": 5.569309186935425,
      "epoch": 1.2326319968241366,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004856054991857013,
      "loss": 5.2154,
      "mean_token_accuracy": 0.16027677357196807,
      "num_tokens": 27207431.0,
      "step": 12420
    },
    {
      "entropy": 5.515983438491821,
      "epoch": 1.233128225486304,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00048559302699079576,
      "loss": 5.0996,
      "mean_token_accuracy": 0.1788653701543808,
      "num_tokens": 27218489.0,
      "step": 12425
    },
    {
      "entropy": 5.603562736511231,
      "epoch": 1.2336244541484715,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004855805495736369,
      "loss": 5.2307,
      "mean_token_accuracy": 0.16810899674892427,
      "num_tokens": 27230849.0,
      "step": 12430
    },
    {
      "entropy": 5.587864589691162,
      "epoch": 1.234120682810639,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004855680669345342,
      "loss": 5.1849,
      "mean_token_accuracy": 0.16571822464466096,
      "num_tokens": 27242838.0,
      "step": 12435
    },
    {
      "entropy": 5.5964422702789305,
      "epoch": 1.2346169114728067,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00048555557907379714,
      "loss": 5.2052,
      "mean_token_accuracy": 0.17497518360614778,
      "num_tokens": 27253945.0,
      "step": 12440
    },
    {
      "entropy": 5.549628257751465,
      "epoch": 1.2351131401349742,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048554308599173544,
      "loss": 5.2706,
      "mean_token_accuracy": 0.1588539093732834,
      "num_tokens": 27263999.0,
      "step": 12445
    },
    {
      "entropy": 5.596800470352173,
      "epoch": 1.2356093687971417,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004855305876886589,
      "loss": 5.1168,
      "mean_token_accuracy": 0.17225392609834672,
      "num_tokens": 27274677.0,
      "step": 12450
    },
    {
      "entropy": 5.578957366943359,
      "epoch": 1.2361055974593094,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00048551808416487743,
      "loss": 5.2045,
      "mean_token_accuracy": 0.1649247393012047,
      "num_tokens": 27286512.0,
      "step": 12455
    },
    {
      "entropy": 5.547616958618164,
      "epoch": 1.2366018261214768,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004855055754207011,
      "loss": 5.191,
      "mean_token_accuracy": 0.17151665389537812,
      "num_tokens": 27297620.0,
      "step": 12460
    },
    {
      "entropy": 5.580215167999268,
      "epoch": 1.2370980547836443,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004854930614564401,
      "loss": 5.1994,
      "mean_token_accuracy": 0.16559920459985733,
      "num_tokens": 27309963.0,
      "step": 12465
    },
    {
      "entropy": 5.650233221054077,
      "epoch": 1.2375942834458118,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004854805422724047,
      "loss": 5.2974,
      "mean_token_accuracy": 0.1608201175928116,
      "num_tokens": 27320740.0,
      "step": 12470
    },
    {
      "entropy": 5.535214757919311,
      "epoch": 1.2380905121079793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004854680178689055,
      "loss": 5.1191,
      "mean_token_accuracy": 0.1717436745762825,
      "num_tokens": 27331543.0,
      "step": 12475
    },
    {
      "entropy": 5.487401676177979,
      "epoch": 1.238586740770147,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048545548824625287,
      "loss": 5.096,
      "mean_token_accuracy": 0.17066734582185744,
      "num_tokens": 27343222.0,
      "step": 12480
    },
    {
      "entropy": 5.5664068222045895,
      "epoch": 1.2390829694323144,
      "grad_norm": 1.0,
      "learning_rate": 0.0004854429534047577,
      "loss": 5.1532,
      "mean_token_accuracy": 0.1702620878815651,
      "num_tokens": 27354694.0,
      "step": 12485
    },
    {
      "entropy": 5.511900520324707,
      "epoch": 1.239579198094482,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004854304133447307,
      "loss": 5.1408,
      "mean_token_accuracy": 0.16547005921602248,
      "num_tokens": 27365462.0,
      "step": 12490
    },
    {
      "entropy": 5.481397199630737,
      "epoch": 1.2400754267566494,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048541786806648293,
      "loss": 5.0741,
      "mean_token_accuracy": 0.17340039163827897,
      "num_tokens": 27377548.0,
      "step": 12495
    },
    {
      "entropy": 5.590326404571533,
      "epoch": 1.240571655418817,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004854053175703254,
      "loss": 5.2023,
      "mean_token_accuracy": 0.16784068048000336,
      "num_tokens": 27388221.0,
      "step": 12500
    },
    {
      "entropy": 5.583331727981568,
      "epoch": 1.2410678840809846,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004853927618565694,
      "loss": 5.1848,
      "mean_token_accuracy": 0.17188149988651275,
      "num_tokens": 27400497.0,
      "step": 12505
    },
    {
      "entropy": 5.518936014175415,
      "epoch": 1.241564112743152,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048538020092552626,
      "loss": 5.0847,
      "mean_token_accuracy": 0.1720809280872345,
      "num_tokens": 27410700.0,
      "step": 12510
    },
    {
      "entropy": 5.560768508911133,
      "epoch": 1.2420603414053195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048536763477750747,
      "loss": 5.2219,
      "mean_token_accuracy": 0.16980680227279663,
      "num_tokens": 27421286.0,
      "step": 12515
    },
    {
      "entropy": 5.6083423614501955,
      "epoch": 1.242556570067487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004853550634128246,
      "loss": 5.2275,
      "mean_token_accuracy": 0.16515350490808486,
      "num_tokens": 27433922.0,
      "step": 12520
    },
    {
      "entropy": 5.542487382888794,
      "epoch": 1.2430527987296547,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004853424868317895,
      "loss": 5.1577,
      "mean_token_accuracy": 0.16695666015148164,
      "num_tokens": 27444684.0,
      "step": 12525
    },
    {
      "entropy": 5.543584489822388,
      "epoch": 1.2435490273918222,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004853299050347139,
      "loss": 5.149,
      "mean_token_accuracy": 0.16552979946136476,
      "num_tokens": 27455494.0,
      "step": 12530
    },
    {
      "entropy": 5.6271607875823975,
      "epoch": 1.2440452560539896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048531731802190994,
      "loss": 5.1838,
      "mean_token_accuracy": 0.164121213555336,
      "num_tokens": 27464927.0,
      "step": 12535
    },
    {
      "entropy": 5.601613140106201,
      "epoch": 1.244541484716157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004853047257936897,
      "loss": 5.2086,
      "mean_token_accuracy": 0.16313580572605133,
      "num_tokens": 27476228.0,
      "step": 12540
    },
    {
      "entropy": 5.568379831314087,
      "epoch": 1.2450377133783248,
      "grad_norm": 0.875,
      "learning_rate": 0.0004852921283503653,
      "loss": 5.2025,
      "mean_token_accuracy": 0.1664481446146965,
      "num_tokens": 27488835.0,
      "step": 12545
    },
    {
      "entropy": 5.554850149154663,
      "epoch": 1.2455339420404923,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004852795256922494,
      "loss": 5.1687,
      "mean_token_accuracy": 0.16842327564954757,
      "num_tokens": 27501481.0,
      "step": 12550
    },
    {
      "entropy": 5.590260553359985,
      "epoch": 1.2460301707026598,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048526691781965425,
      "loss": 5.1851,
      "mean_token_accuracy": 0.1662691742181778,
      "num_tokens": 27512234.0,
      "step": 12555
    },
    {
      "entropy": 5.612431621551513,
      "epoch": 1.2465263993648272,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048525430473289274,
      "loss": 5.2399,
      "mean_token_accuracy": 0.16261722445487975,
      "num_tokens": 27523756.0,
      "step": 12560
    },
    {
      "entropy": 5.590975046157837,
      "epoch": 1.247022628026995,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048524168643227747,
      "loss": 5.2304,
      "mean_token_accuracy": 0.16207225173711776,
      "num_tokens": 27534579.0,
      "step": 12565
    },
    {
      "entropy": 5.566108274459839,
      "epoch": 1.2475188566891624,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048522906291812145,
      "loss": 5.2665,
      "mean_token_accuracy": 0.1622200131416321,
      "num_tokens": 27546420.0,
      "step": 12570
    },
    {
      "entropy": 5.597119903564453,
      "epoch": 1.2480150853513299,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048521643419073763,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1666351541876793,
      "num_tokens": 27557701.0,
      "step": 12575
    },
    {
      "entropy": 5.517369985580444,
      "epoch": 1.2485113140134974,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004852038002504392,
      "loss": 5.1468,
      "mean_token_accuracy": 0.17534344345331193,
      "num_tokens": 27568592.0,
      "step": 12580
    },
    {
      "entropy": 5.553262090682983,
      "epoch": 1.249007542675665,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048519116109753955,
      "loss": 5.1926,
      "mean_token_accuracy": 0.17634792029857635,
      "num_tokens": 27579286.0,
      "step": 12585
    },
    {
      "entropy": 5.542362689971924,
      "epoch": 1.2495037713378325,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048517851673235196,
      "loss": 5.1301,
      "mean_token_accuracy": 0.16944528073072435,
      "num_tokens": 27590031.0,
      "step": 12590
    },
    {
      "entropy": 5.56602897644043,
      "epoch": 1.25,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004851658671551901,
      "loss": 5.1692,
      "mean_token_accuracy": 0.16481332033872603,
      "num_tokens": 27600999.0,
      "step": 12595
    },
    {
      "entropy": 5.523578882217407,
      "epoch": 1.2504962286621675,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048515321236636754,
      "loss": 5.1623,
      "mean_token_accuracy": 0.16316679865121841,
      "num_tokens": 27613394.0,
      "step": 12600
    },
    {
      "entropy": 5.541903114318847,
      "epoch": 1.250992457324335,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004851405523661982,
      "loss": 5.1342,
      "mean_token_accuracy": 0.174838787317276,
      "num_tokens": 27624819.0,
      "step": 12605
    },
    {
      "entropy": 5.527566719055176,
      "epoch": 1.2514886859865026,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048512788715499597,
      "loss": 5.1611,
      "mean_token_accuracy": 0.17427634447813034,
      "num_tokens": 27635993.0,
      "step": 12610
    },
    {
      "entropy": 5.553778076171875,
      "epoch": 1.2519849146486701,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004851152167330749,
      "loss": 5.2289,
      "mean_token_accuracy": 0.1660592347383499,
      "num_tokens": 27647298.0,
      "step": 12615
    },
    {
      "entropy": 5.5937761783599855,
      "epoch": 1.2524811433108376,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048510254110074923,
      "loss": 5.204,
      "mean_token_accuracy": 0.16549559831619262,
      "num_tokens": 27659381.0,
      "step": 12620
    },
    {
      "entropy": 5.562508487701416,
      "epoch": 1.2529773719730053,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004850898602583333,
      "loss": 5.183,
      "mean_token_accuracy": 0.16642069816589355,
      "num_tokens": 27671122.0,
      "step": 12625
    },
    {
      "entropy": 5.610251188278198,
      "epoch": 1.2534736006351728,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048507717420614153,
      "loss": 5.1836,
      "mean_token_accuracy": 0.16948157399892808,
      "num_tokens": 27683258.0,
      "step": 12630
    },
    {
      "entropy": 5.479177808761596,
      "epoch": 1.2539698292973402,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004850644829444884,
      "loss": 5.0476,
      "mean_token_accuracy": 0.17024100422859192,
      "num_tokens": 27693116.0,
      "step": 12635
    },
    {
      "entropy": 5.473917293548584,
      "epoch": 1.2544660579595077,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048505178647368887,
      "loss": 5.1367,
      "mean_token_accuracy": 0.1723328024148941,
      "num_tokens": 27704859.0,
      "step": 12640
    },
    {
      "entropy": 5.493315696716309,
      "epoch": 1.2549622866216752,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004850390847940576,
      "loss": 5.0897,
      "mean_token_accuracy": 0.17135029584169387,
      "num_tokens": 27715854.0,
      "step": 12645
    },
    {
      "entropy": 5.5634886741638185,
      "epoch": 1.2554585152838427,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004850263779059097,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17186927497386933,
      "num_tokens": 27726398.0,
      "step": 12650
    },
    {
      "entropy": 5.575637435913086,
      "epoch": 1.2559547439460104,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004850136658095601,
      "loss": 5.1663,
      "mean_token_accuracy": 0.17372089624404907,
      "num_tokens": 27737337.0,
      "step": 12655
    },
    {
      "entropy": 5.528477048873901,
      "epoch": 1.2564509726081778,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048500094850532417,
      "loss": 5.1872,
      "mean_token_accuracy": 0.17228624522686004,
      "num_tokens": 27748593.0,
      "step": 12660
    },
    {
      "entropy": 5.587150192260742,
      "epoch": 1.2569472012703453,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004849882259935172,
      "loss": 5.1519,
      "mean_token_accuracy": 0.16612059026956558,
      "num_tokens": 27760908.0,
      "step": 12665
    },
    {
      "entropy": 5.5898524761199955,
      "epoch": 1.257443429932513,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004849754982744548,
      "loss": 5.2371,
      "mean_token_accuracy": 0.16730549335479736,
      "num_tokens": 27772104.0,
      "step": 12670
    },
    {
      "entropy": 5.506442499160767,
      "epoch": 1.2579396585946805,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004849627653484524,
      "loss": 5.1718,
      "mean_token_accuracy": 0.17423984110355378,
      "num_tokens": 27781654.0,
      "step": 12675
    },
    {
      "entropy": 5.589901876449585,
      "epoch": 1.258435887256848,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004849500272158259,
      "loss": 5.1653,
      "mean_token_accuracy": 0.17162521779537201,
      "num_tokens": 27793026.0,
      "step": 12680
    },
    {
      "entropy": 5.553213787078858,
      "epoch": 1.2589321159190154,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004849372838768911,
      "loss": 5.141,
      "mean_token_accuracy": 0.17194158881902694,
      "num_tokens": 27803447.0,
      "step": 12685
    },
    {
      "entropy": 5.624784278869629,
      "epoch": 1.259428344581183,
      "grad_norm": 1.046875,
      "learning_rate": 0.000484924535331964,
      "loss": 5.1461,
      "mean_token_accuracy": 0.1697594180703163,
      "num_tokens": 27814632.0,
      "step": 12690
    },
    {
      "entropy": 5.556216812133789,
      "epoch": 1.2599245732433506,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048491178158136086,
      "loss": 5.2254,
      "mean_token_accuracy": 0.16438426673412324,
      "num_tokens": 27826582.0,
      "step": 12695
    },
    {
      "entropy": 5.561621236801147,
      "epoch": 1.260420801905518,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004848990226253978,
      "loss": 5.0793,
      "mean_token_accuracy": 0.16775094121694564,
      "num_tokens": 27838065.0,
      "step": 12700
    },
    {
      "entropy": 5.496501922607422,
      "epoch": 1.2609170305676856,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048488625846439135,
      "loss": 5.1502,
      "mean_token_accuracy": 0.1715391457080841,
      "num_tokens": 27848182.0,
      "step": 12705
    },
    {
      "entropy": 5.586357879638672,
      "epoch": 1.261413259229853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004848734890986578,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16642217040061952,
      "num_tokens": 27858524.0,
      "step": 12710
    },
    {
      "entropy": 5.546224355697632,
      "epoch": 1.2619094878920207,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004848607145285141,
      "loss": 5.0809,
      "mean_token_accuracy": 0.17713888138532638,
      "num_tokens": 27868588.0,
      "step": 12715
    },
    {
      "entropy": 5.502144622802734,
      "epoch": 1.2624057165541882,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004848479347542768,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17195945829153061,
      "num_tokens": 27878829.0,
      "step": 12720
    },
    {
      "entropy": 5.527616500854492,
      "epoch": 1.2629019452163557,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000484835149776263,
      "loss": 5.1661,
      "mean_token_accuracy": 0.1767250657081604,
      "num_tokens": 27889872.0,
      "step": 12725
    },
    {
      "entropy": 5.541048526763916,
      "epoch": 1.2633981738785232,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048482235959478953,
      "loss": 5.2285,
      "mean_token_accuracy": 0.16940861195325851,
      "num_tokens": 27900567.0,
      "step": 12730
    },
    {
      "entropy": 5.59661169052124,
      "epoch": 1.2638944025406906,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004848095642101737,
      "loss": 5.115,
      "mean_token_accuracy": 0.17305362969636917,
      "num_tokens": 27911594.0,
      "step": 12735
    },
    {
      "entropy": 5.589021110534668,
      "epoch": 1.2643906312028583,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048479676362273277,
      "loss": 5.2149,
      "mean_token_accuracy": 0.16117834001779557,
      "num_tokens": 27923694.0,
      "step": 12740
    },
    {
      "entropy": 5.581939840316773,
      "epoch": 1.2648868598650258,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048478395783278425,
      "loss": 5.2919,
      "mean_token_accuracy": 0.15176186114549636,
      "num_tokens": 27934168.0,
      "step": 12745
    },
    {
      "entropy": 5.62507586479187,
      "epoch": 1.2653830885271933,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004847711468406455,
      "loss": 5.1784,
      "mean_token_accuracy": 0.16945082247257232,
      "num_tokens": 27944401.0,
      "step": 12750
    },
    {
      "entropy": 5.526853370666504,
      "epoch": 1.265879317189361,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048475833064663443,
      "loss": 5.0741,
      "mean_token_accuracy": 0.17674310505390167,
      "num_tokens": 27956219.0,
      "step": 12755
    },
    {
      "entropy": 5.550226068496704,
      "epoch": 1.2663755458515285,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004847455092510687,
      "loss": 5.1579,
      "mean_token_accuracy": 0.17305746376514436,
      "num_tokens": 27967457.0,
      "step": 12760
    },
    {
      "entropy": 5.524404525756836,
      "epoch": 1.266871774513696,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004847326826542663,
      "loss": 5.185,
      "mean_token_accuracy": 0.16694278568029403,
      "num_tokens": 27978287.0,
      "step": 12765
    },
    {
      "entropy": 5.564602613449097,
      "epoch": 1.2673680031758634,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048471985085654537,
      "loss": 5.1989,
      "mean_token_accuracy": 0.17228431701660157,
      "num_tokens": 27989703.0,
      "step": 12770
    },
    {
      "entropy": 5.545407009124756,
      "epoch": 1.2678642318380309,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048470701385822394,
      "loss": 5.1665,
      "mean_token_accuracy": 0.17012592852115632,
      "num_tokens": 28002071.0,
      "step": 12775
    },
    {
      "entropy": 5.588917541503906,
      "epoch": 1.2683604605001986,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048469417165962047,
      "loss": 5.2263,
      "mean_token_accuracy": 0.17092974334955216,
      "num_tokens": 28014517.0,
      "step": 12780
    },
    {
      "entropy": 5.5586316108703615,
      "epoch": 1.268856689162366,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004846813242610535,
      "loss": 5.1833,
      "mean_token_accuracy": 0.17548978328704834,
      "num_tokens": 28025135.0,
      "step": 12785
    },
    {
      "entropy": 5.489023303985595,
      "epoch": 1.2693529178245335,
      "grad_norm": 1.0,
      "learning_rate": 0.0004846684716628414,
      "loss": 5.1116,
      "mean_token_accuracy": 0.1721599966287613,
      "num_tokens": 28036105.0,
      "step": 12790
    },
    {
      "entropy": 5.503118419647217,
      "epoch": 1.269849146486701,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000484655613865303,
      "loss": 5.1605,
      "mean_token_accuracy": 0.17289704233407974,
      "num_tokens": 28048825.0,
      "step": 12795
    },
    {
      "entropy": 5.5426171779632565,
      "epoch": 1.2703453751488687,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048464275086875717,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1716746285557747,
      "num_tokens": 28060408.0,
      "step": 12800
    },
    {
      "entropy": 5.543191576004029,
      "epoch": 1.2708416038110362,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004846298826735229,
      "loss": 5.1882,
      "mean_token_accuracy": 0.1657179117202759,
      "num_tokens": 28072179.0,
      "step": 12805
    },
    {
      "entropy": 5.533158779144287,
      "epoch": 1.2713378324732036,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004846170092799192,
      "loss": 5.1507,
      "mean_token_accuracy": 0.16797443926334382,
      "num_tokens": 28083825.0,
      "step": 12810
    },
    {
      "entropy": 5.733220911026001,
      "epoch": 1.2718340611353711,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048460413068826536,
      "loss": 5.3087,
      "mean_token_accuracy": 0.15962689071893693,
      "num_tokens": 28095136.0,
      "step": 12815
    },
    {
      "entropy": 5.590328788757324,
      "epoch": 1.2723302897975386,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845912468988807,
      "loss": 5.1349,
      "mean_token_accuracy": 0.16564120650291442,
      "num_tokens": 28104763.0,
      "step": 12820
    },
    {
      "entropy": 5.538482904434204,
      "epoch": 1.2728265184597063,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004845783579120849,
      "loss": 5.1957,
      "mean_token_accuracy": 0.16963932514190674,
      "num_tokens": 28115771.0,
      "step": 12825
    },
    {
      "entropy": 5.506938123703003,
      "epoch": 1.2733227471218738,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004845654637281973,
      "loss": 5.1203,
      "mean_token_accuracy": 0.1690376430749893,
      "num_tokens": 28128039.0,
      "step": 12830
    },
    {
      "entropy": 5.509294176101685,
      "epoch": 1.2738189757840412,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048455256434753774,
      "loss": 5.1257,
      "mean_token_accuracy": 0.17922435104846954,
      "num_tokens": 28139163.0,
      "step": 12835
    },
    {
      "entropy": 5.516084098815918,
      "epoch": 1.274315204446209,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004845396597704262,
      "loss": 5.168,
      "mean_token_accuracy": 0.1663430243730545,
      "num_tokens": 28150625.0,
      "step": 12840
    },
    {
      "entropy": 5.585581064224243,
      "epoch": 1.2748114331083764,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048452674999718257,
      "loss": 5.2335,
      "mean_token_accuracy": 0.1705207571387291,
      "num_tokens": 28163072.0,
      "step": 12845
    },
    {
      "entropy": 5.64183030128479,
      "epoch": 1.2753076617705439,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048451383502812705,
      "loss": 5.2653,
      "mean_token_accuracy": 0.16579011231660842,
      "num_tokens": 28174192.0,
      "step": 12850
    },
    {
      "entropy": 5.536531066894531,
      "epoch": 1.2758038904327114,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048450091486357987,
      "loss": 5.1137,
      "mean_token_accuracy": 0.1752413257956505,
      "num_tokens": 28184477.0,
      "step": 12855
    },
    {
      "entropy": 5.500038290023804,
      "epoch": 1.2763001190948788,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048448798950386147,
      "loss": 5.0469,
      "mean_token_accuracy": 0.1754678577184677,
      "num_tokens": 28194552.0,
      "step": 12860
    },
    {
      "entropy": 5.554371070861817,
      "epoch": 1.2767963477570463,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048447505894929233,
      "loss": 5.2137,
      "mean_token_accuracy": 0.1669171929359436,
      "num_tokens": 28205712.0,
      "step": 12865
    },
    {
      "entropy": 5.589458990097046,
      "epoch": 1.277292576419214,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048446212320019306,
      "loss": 5.135,
      "mean_token_accuracy": 0.17312112152576448,
      "num_tokens": 28217246.0,
      "step": 12870
    },
    {
      "entropy": 5.542981815338135,
      "epoch": 1.2777888050813815,
      "grad_norm": 0.875,
      "learning_rate": 0.0004844491822568846,
      "loss": 5.1555,
      "mean_token_accuracy": 0.16471437066793443,
      "num_tokens": 28227933.0,
      "step": 12875
    },
    {
      "entropy": 5.535181856155395,
      "epoch": 1.278285033743549,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00048443623611968764,
      "loss": 5.1864,
      "mean_token_accuracy": 0.17255932986736297,
      "num_tokens": 28239692.0,
      "step": 12880
    },
    {
      "entropy": 5.52805871963501,
      "epoch": 1.2787812624057167,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048442328478892334,
      "loss": 5.1546,
      "mean_token_accuracy": 0.1685006558895111,
      "num_tokens": 28249815.0,
      "step": 12885
    },
    {
      "entropy": 5.588373327255249,
      "epoch": 1.2792774910678841,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004844103282649128,
      "loss": 5.1781,
      "mean_token_accuracy": 0.16615121364593505,
      "num_tokens": 28260385.0,
      "step": 12890
    },
    {
      "entropy": 5.543506669998169,
      "epoch": 1.2797737197300516,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004843973665479774,
      "loss": 5.1889,
      "mean_token_accuracy": 0.1615331292152405,
      "num_tokens": 28271858.0,
      "step": 12895
    },
    {
      "entropy": 5.584776735305786,
      "epoch": 1.280269948392219,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048438439963843853,
      "loss": 5.1737,
      "mean_token_accuracy": 0.16602047234773637,
      "num_tokens": 28283988.0,
      "step": 12900
    },
    {
      "entropy": 5.564140653610229,
      "epoch": 1.2807661770543866,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048437142753661775,
      "loss": 5.2083,
      "mean_token_accuracy": 0.1701462984085083,
      "num_tokens": 28293281.0,
      "step": 12905
    },
    {
      "entropy": 5.6243995189666744,
      "epoch": 1.2812624057165543,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004843584502428367,
      "loss": 5.1618,
      "mean_token_accuracy": 0.16337970197200774,
      "num_tokens": 28304468.0,
      "step": 12910
    },
    {
      "entropy": 5.469922971725464,
      "epoch": 1.2817586343787217,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004843454677574173,
      "loss": 5.119,
      "mean_token_accuracy": 0.1762610137462616,
      "num_tokens": 28315864.0,
      "step": 12915
    },
    {
      "entropy": 5.520497465133667,
      "epoch": 1.2822548630408892,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048433248008068123,
      "loss": 5.1764,
      "mean_token_accuracy": 0.1769901067018509,
      "num_tokens": 28327212.0,
      "step": 12920
    },
    {
      "entropy": 5.599587392807007,
      "epoch": 1.282751091703057,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048431948721295085,
      "loss": 5.1907,
      "mean_token_accuracy": 0.1706276625394821,
      "num_tokens": 28339256.0,
      "step": 12925
    },
    {
      "entropy": 5.58486008644104,
      "epoch": 1.2832473203652244,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004843064891545482,
      "loss": 5.1952,
      "mean_token_accuracy": 0.16598389446735382,
      "num_tokens": 28350196.0,
      "step": 12930
    },
    {
      "entropy": 5.560580921173096,
      "epoch": 1.2837435490273919,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004842934859057956,
      "loss": 5.1717,
      "mean_token_accuracy": 0.16620030999183655,
      "num_tokens": 28361241.0,
      "step": 12935
    },
    {
      "entropy": 5.630869960784912,
      "epoch": 1.2842397776895593,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004842804774670156,
      "loss": 5.2867,
      "mean_token_accuracy": 0.1599566102027893,
      "num_tokens": 28373369.0,
      "step": 12940
    },
    {
      "entropy": 5.5808757781982425,
      "epoch": 1.2847360063517268,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048426746383853066,
      "loss": 5.126,
      "mean_token_accuracy": 0.17031968533992767,
      "num_tokens": 28384970.0,
      "step": 12945
    },
    {
      "entropy": 5.561895847320557,
      "epoch": 1.2852322350138943,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048425444502066354,
      "loss": 5.132,
      "mean_token_accuracy": 0.16839235723018647,
      "num_tokens": 28396469.0,
      "step": 12950
    },
    {
      "entropy": 5.525531530380249,
      "epoch": 1.285728463676062,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048424142101373706,
      "loss": 5.1825,
      "mean_token_accuracy": 0.16763636618852615,
      "num_tokens": 28406975.0,
      "step": 12955
    },
    {
      "entropy": 5.462316083908081,
      "epoch": 1.2862246923382294,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048422839181807426,
      "loss": 5.0651,
      "mean_token_accuracy": 0.18056480586528778,
      "num_tokens": 28418285.0,
      "step": 12960
    },
    {
      "entropy": 5.54815526008606,
      "epoch": 1.286720921000397,
      "grad_norm": 0.875,
      "learning_rate": 0.0004842153574339982,
      "loss": 5.2208,
      "mean_token_accuracy": 0.1650894656777382,
      "num_tokens": 28429769.0,
      "step": 12965
    },
    {
      "entropy": 5.536801290512085,
      "epoch": 1.2872171496625646,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048420231786183216,
      "loss": 5.1582,
      "mean_token_accuracy": 0.1743309825658798,
      "num_tokens": 28441188.0,
      "step": 12970
    },
    {
      "entropy": 5.626014471054077,
      "epoch": 1.287713378324732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004841892731018994,
      "loss": 5.2292,
      "mean_token_accuracy": 0.1629510298371315,
      "num_tokens": 28453504.0,
      "step": 12975
    },
    {
      "entropy": 5.599001359939575,
      "epoch": 1.2882096069868996,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004841762231545233,
      "loss": 5.1972,
      "mean_token_accuracy": 0.17582966536283492,
      "num_tokens": 28465930.0,
      "step": 12980
    },
    {
      "entropy": 5.528409481048584,
      "epoch": 1.288705835649067,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048416316802002776,
      "loss": 5.1183,
      "mean_token_accuracy": 0.17431016117334366,
      "num_tokens": 28477803.0,
      "step": 12985
    },
    {
      "entropy": 5.588806056976319,
      "epoch": 1.2892020643112345,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048415010769873625,
      "loss": 5.2179,
      "mean_token_accuracy": 0.1658260479569435,
      "num_tokens": 28488556.0,
      "step": 12990
    },
    {
      "entropy": 5.603724336624145,
      "epoch": 1.2896982929734022,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004841370421909728,
      "loss": 5.1209,
      "mean_token_accuracy": 0.16915047466754912,
      "num_tokens": 28499275.0,
      "step": 12995
    },
    {
      "entropy": 5.587145376205444,
      "epoch": 1.2901945216355697,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048412397149706133,
      "loss": 5.15,
      "mean_token_accuracy": 0.17171983569860458,
      "num_tokens": 28511327.0,
      "step": 13000
    },
    {
      "entropy": 5.520497512817383,
      "epoch": 1.2906907502977372,
      "grad_norm": 0.984375,
      "learning_rate": 0.000484110895617326,
      "loss": 5.1798,
      "mean_token_accuracy": 0.17440134286880493,
      "num_tokens": 28523937.0,
      "step": 13005
    },
    {
      "entropy": 5.460461568832398,
      "epoch": 1.2911869789599046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048409781455209104,
      "loss": 5.0984,
      "mean_token_accuracy": 0.16967738419771194,
      "num_tokens": 28534028.0,
      "step": 13010
    },
    {
      "entropy": 5.594855165481567,
      "epoch": 1.2916832076220723,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004840847283016809,
      "loss": 5.1875,
      "mean_token_accuracy": 0.15785751193761827,
      "num_tokens": 28545480.0,
      "step": 13015
    },
    {
      "entropy": 5.565376567840576,
      "epoch": 1.2921794362842398,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048407163686642,
      "loss": 5.1267,
      "mean_token_accuracy": 0.17909184694290162,
      "num_tokens": 28555401.0,
      "step": 13020
    },
    {
      "entropy": 5.569421434402466,
      "epoch": 1.2926756649464073,
      "grad_norm": 1.03125,
      "learning_rate": 0.000484058540246633,
      "loss": 5.1953,
      "mean_token_accuracy": 0.17224758565425874,
      "num_tokens": 28565497.0,
      "step": 13025
    },
    {
      "entropy": 5.567805814743042,
      "epoch": 1.2931718936085748,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004840454384426447,
      "loss": 5.2813,
      "mean_token_accuracy": 0.1610473707318306,
      "num_tokens": 28576186.0,
      "step": 13030
    },
    {
      "entropy": 5.609815073013306,
      "epoch": 1.2936681222707422,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004840323314547799,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1726394772529602,
      "num_tokens": 28586202.0,
      "step": 13035
    },
    {
      "entropy": 5.550849676132202,
      "epoch": 1.29416435093291,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048401921928336385,
      "loss": 5.1246,
      "mean_token_accuracy": 0.16388458013534546,
      "num_tokens": 28596486.0,
      "step": 13040
    },
    {
      "entropy": 5.56447958946228,
      "epoch": 1.2946605795950774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004840061019287215,
      "loss": 5.1658,
      "mean_token_accuracy": 0.17187954485416412,
      "num_tokens": 28607934.0,
      "step": 13045
    },
    {
      "entropy": 5.610556936264038,
      "epoch": 1.2951568082572449,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048399297939117817,
      "loss": 5.2057,
      "mean_token_accuracy": 0.15898504257202148,
      "num_tokens": 28617673.0,
      "step": 13050
    },
    {
      "entropy": 5.616237926483154,
      "epoch": 1.2956530369194126,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004839798516710593,
      "loss": 5.1855,
      "mean_token_accuracy": 0.17065990418195726,
      "num_tokens": 28629811.0,
      "step": 13055
    },
    {
      "entropy": 5.5918426513671875,
      "epoch": 1.29614926558158,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004839667187686904,
      "loss": 5.2776,
      "mean_token_accuracy": 0.1641586348414421,
      "num_tokens": 28640371.0,
      "step": 13060
    },
    {
      "entropy": 5.5759429931640625,
      "epoch": 1.2966454942437475,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004839535806843972,
      "loss": 5.1895,
      "mean_token_accuracy": 0.1715485379099846,
      "num_tokens": 28651904.0,
      "step": 13065
    },
    {
      "entropy": 5.514900350570679,
      "epoch": 1.297141722905915,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004839404374185054,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1685224026441574,
      "num_tokens": 28662241.0,
      "step": 13070
    },
    {
      "entropy": 5.519912481307983,
      "epoch": 1.2976379515680825,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000483927288971341,
      "loss": 5.1024,
      "mean_token_accuracy": 0.1728796049952507,
      "num_tokens": 28672791.0,
      "step": 13075
    },
    {
      "entropy": 5.560097694396973,
      "epoch": 1.2981341802302502,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004839141353432301,
      "loss": 5.236,
      "mean_token_accuracy": 0.1673694521188736,
      "num_tokens": 28684218.0,
      "step": 13080
    },
    {
      "entropy": 5.556421089172363,
      "epoch": 1.2986304088924177,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004839009765344987,
      "loss": 5.1614,
      "mean_token_accuracy": 0.16688167303800583,
      "num_tokens": 28694690.0,
      "step": 13085
    },
    {
      "entropy": 5.529719400405884,
      "epoch": 1.2991266375545851,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00048388781254547324,
      "loss": 5.1484,
      "mean_token_accuracy": 0.17535197883844375,
      "num_tokens": 28705890.0,
      "step": 13090
    },
    {
      "entropy": 5.5735893726348875,
      "epoch": 1.2996228662167526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048387464337648026,
      "loss": 5.1942,
      "mean_token_accuracy": 0.16832694858312608,
      "num_tokens": 28716935.0,
      "step": 13095
    },
    {
      "entropy": 5.607764768600464,
      "epoch": 1.3001190948789203,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048386146902784606,
      "loss": 5.1876,
      "mean_token_accuracy": 0.1691871091723442,
      "num_tokens": 28728053.0,
      "step": 13100
    },
    {
      "entropy": 5.504816627502441,
      "epoch": 1.3006153235410878,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048384828949989757,
      "loss": 5.1253,
      "mean_token_accuracy": 0.16837322413921357,
      "num_tokens": 28738940.0,
      "step": 13105
    },
    {
      "entropy": 5.573337554931641,
      "epoch": 1.3011115522032553,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048383510479296153,
      "loss": 5.2063,
      "mean_token_accuracy": 0.16873059868812562,
      "num_tokens": 28750259.0,
      "step": 13110
    },
    {
      "entropy": 5.597581672668457,
      "epoch": 1.3016077808654227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838219149073649,
      "loss": 5.1501,
      "mean_token_accuracy": 0.17250781655311584,
      "num_tokens": 28760119.0,
      "step": 13115
    },
    {
      "entropy": 5.603250646591187,
      "epoch": 1.3021040095275902,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004838087198434347,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1661520779132843,
      "num_tokens": 28772022.0,
      "step": 13120
    },
    {
      "entropy": 5.563337182998657,
      "epoch": 1.302600238189758,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004837955196014982,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16151117533445358,
      "num_tokens": 28782935.0,
      "step": 13125
    },
    {
      "entropy": 5.627027320861816,
      "epoch": 1.3030964668519254,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048378231418188275,
      "loss": 5.262,
      "mean_token_accuracy": 0.16959324479103088,
      "num_tokens": 28793604.0,
      "step": 13130
    },
    {
      "entropy": 5.587513875961304,
      "epoch": 1.3035926955140928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048376910358491576,
      "loss": 5.2032,
      "mean_token_accuracy": 0.16635196506977082,
      "num_tokens": 28803494.0,
      "step": 13135
    },
    {
      "entropy": 5.557760715484619,
      "epoch": 1.3040889241762605,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004837558878109249,
      "loss": 5.1792,
      "mean_token_accuracy": 0.16507562696933747,
      "num_tokens": 28816232.0,
      "step": 13140
    },
    {
      "entropy": 5.555879020690918,
      "epoch": 1.304585152838428,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004837426668602378,
      "loss": 5.14,
      "mean_token_accuracy": 0.17300387918949128,
      "num_tokens": 28827056.0,
      "step": 13145
    },
    {
      "entropy": 5.519309759140015,
      "epoch": 1.3050813815005955,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004837294407331825,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17395287305116652,
      "num_tokens": 28837217.0,
      "step": 13150
    },
    {
      "entropy": 5.498017740249634,
      "epoch": 1.305577610162763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004837162094300867,
      "loss": 5.0348,
      "mean_token_accuracy": 0.17270750254392625,
      "num_tokens": 28846838.0,
      "step": 13155
    },
    {
      "entropy": 5.566566801071167,
      "epoch": 1.3060738388249304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004837029729512787,
      "loss": 5.089,
      "mean_token_accuracy": 0.17576585561037064,
      "num_tokens": 28856269.0,
      "step": 13160
    },
    {
      "entropy": 5.564627599716187,
      "epoch": 1.306570067487098,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048368973129708667,
      "loss": 5.1249,
      "mean_token_accuracy": 0.1767018035054207,
      "num_tokens": 28867142.0,
      "step": 13165
    },
    {
      "entropy": 5.534079217910767,
      "epoch": 1.3070662961492656,
      "grad_norm": 1.125,
      "learning_rate": 0.00048367648446783894,
      "loss": 5.0954,
      "mean_token_accuracy": 0.16927044838666916,
      "num_tokens": 28877346.0,
      "step": 13170
    },
    {
      "entropy": 5.54451551437378,
      "epoch": 1.307562524811433,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004836632324638642,
      "loss": 5.0906,
      "mean_token_accuracy": 0.1706339344382286,
      "num_tokens": 28888425.0,
      "step": 13175
    },
    {
      "entropy": 5.563062429428101,
      "epoch": 1.3080587534736006,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048364997528549077,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17041877210140227,
      "num_tokens": 28899575.0,
      "step": 13180
    },
    {
      "entropy": 5.525182771682739,
      "epoch": 1.3085549821357683,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048363671293304763,
      "loss": 5.1172,
      "mean_token_accuracy": 0.17119070887565613,
      "num_tokens": 28910769.0,
      "step": 13185
    },
    {
      "entropy": 5.535555410385132,
      "epoch": 1.3090512107979357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004836234454068636,
      "loss": 5.134,
      "mean_token_accuracy": 0.17728340923786162,
      "num_tokens": 28921224.0,
      "step": 13190
    },
    {
      "entropy": 5.518017959594727,
      "epoch": 1.3095474394601032,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048361017270726764,
      "loss": 5.133,
      "mean_token_accuracy": 0.16947024762630464,
      "num_tokens": 28932585.0,
      "step": 13195
    },
    {
      "entropy": 5.5269551277160645,
      "epoch": 1.3100436681222707,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004835968948345889,
      "loss": 5.1766,
      "mean_token_accuracy": 0.16612859666347504,
      "num_tokens": 28943471.0,
      "step": 13200
    },
    {
      "entropy": 5.531363534927368,
      "epoch": 1.3105398967844382,
      "grad_norm": 0.875,
      "learning_rate": 0.00048358361178915666,
      "loss": 5.0866,
      "mean_token_accuracy": 0.1830190673470497,
      "num_tokens": 28956336.0,
      "step": 13205
    },
    {
      "entropy": 5.636476993560791,
      "epoch": 1.3110361254466059,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004835703235713003,
      "loss": 5.2116,
      "mean_token_accuracy": 0.16897871494293212,
      "num_tokens": 28966689.0,
      "step": 13210
    },
    {
      "entropy": 5.587238836288452,
      "epoch": 1.3115323541087733,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048355703018134935,
      "loss": 5.3248,
      "mean_token_accuracy": 0.1730082243680954,
      "num_tokens": 28978548.0,
      "step": 13215
    },
    {
      "entropy": 5.597854709625244,
      "epoch": 1.3120285827709408,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004835437316196335,
      "loss": 5.2262,
      "mean_token_accuracy": 0.167519274353981,
      "num_tokens": 28988899.0,
      "step": 13220
    },
    {
      "entropy": 5.531807899475098,
      "epoch": 1.3125248114331085,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004835304278864824,
      "loss": 5.1437,
      "mean_token_accuracy": 0.17729974389076233,
      "num_tokens": 28999666.0,
      "step": 13225
    },
    {
      "entropy": 5.5764772415161135,
      "epoch": 1.313021040095276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004835171189822261,
      "loss": 5.1588,
      "mean_token_accuracy": 0.16706880033016205,
      "num_tokens": 29010185.0,
      "step": 13230
    },
    {
      "entropy": 5.538579654693604,
      "epoch": 1.3135172687574435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004835038049071945,
      "loss": 5.1519,
      "mean_token_accuracy": 0.1715313822031021,
      "num_tokens": 29020761.0,
      "step": 13235
    },
    {
      "entropy": 5.643155241012574,
      "epoch": 1.314013497419611,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004834904856617178,
      "loss": 5.3085,
      "mean_token_accuracy": 0.16575097739696504,
      "num_tokens": 29031844.0,
      "step": 13240
    },
    {
      "entropy": 5.671302127838135,
      "epoch": 1.3145097260817784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048347716124612643,
      "loss": 5.2605,
      "mean_token_accuracy": 0.16600753515958785,
      "num_tokens": 29043101.0,
      "step": 13245
    },
    {
      "entropy": 5.621965312957764,
      "epoch": 1.3150059547439459,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004834638316607506,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1740931138396263,
      "num_tokens": 29054016.0,
      "step": 13250
    },
    {
      "entropy": 5.49326343536377,
      "epoch": 1.3155021834061136,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048345049690592105,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1678404688835144,
      "num_tokens": 29065127.0,
      "step": 13255
    },
    {
      "entropy": 5.5904014110565186,
      "epoch": 1.315998412068281,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048343715698196814,
      "loss": 5.1419,
      "mean_token_accuracy": 0.16814993172883988,
      "num_tokens": 29075130.0,
      "step": 13260
    },
    {
      "entropy": 5.647088146209716,
      "epoch": 1.3164946407304485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004834238118892231,
      "loss": 5.2523,
      "mean_token_accuracy": 0.16217064931988717,
      "num_tokens": 29086405.0,
      "step": 13265
    },
    {
      "entropy": 5.602945375442505,
      "epoch": 1.3169908693926162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834104616280165,
      "loss": 5.2025,
      "mean_token_accuracy": 0.16484274715185165,
      "num_tokens": 29096945.0,
      "step": 13270
    },
    {
      "entropy": 5.600357246398926,
      "epoch": 1.3174870980547837,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004833971061986796,
      "loss": 5.2754,
      "mean_token_accuracy": 0.16649991571903228,
      "num_tokens": 29108275.0,
      "step": 13275
    },
    {
      "entropy": 5.562314081192016,
      "epoch": 1.3179833267169512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048338374560154344,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1657174840569496,
      "num_tokens": 29118832.0,
      "step": 13280
    },
    {
      "entropy": 5.51747350692749,
      "epoch": 1.3184795553791187,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004833703798369395,
      "loss": 5.1386,
      "mean_token_accuracy": 0.17059405595064164,
      "num_tokens": 29130720.0,
      "step": 13285
    },
    {
      "entropy": 5.608975791931153,
      "epoch": 1.3189757840412861,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004833570089051992,
      "loss": 5.1366,
      "mean_token_accuracy": 0.17224212884902954,
      "num_tokens": 29140441.0,
      "step": 13290
    },
    {
      "entropy": 5.596837711334229,
      "epoch": 1.3194720127034538,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048334363280665393,
      "loss": 5.1917,
      "mean_token_accuracy": 0.1687237411737442,
      "num_tokens": 29151544.0,
      "step": 13295
    },
    {
      "entropy": 5.620654344558716,
      "epoch": 1.3199682413656213,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004833302515416355,
      "loss": 5.2529,
      "mean_token_accuracy": 0.1669286385178566,
      "num_tokens": 29165475.0,
      "step": 13300
    },
    {
      "entropy": 5.5311750888824465,
      "epoch": 1.3204644700277888,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048331686511047584,
      "loss": 5.1004,
      "mean_token_accuracy": 0.16908559799194336,
      "num_tokens": 29177139.0,
      "step": 13305
    },
    {
      "entropy": 5.594502353668213,
      "epoch": 1.3209606986899565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048330347351350676,
      "loss": 5.17,
      "mean_token_accuracy": 0.17286210209131242,
      "num_tokens": 29187812.0,
      "step": 13310
    },
    {
      "entropy": 5.522208976745605,
      "epoch": 1.321456927352124,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048329007675106036,
      "loss": 5.0778,
      "mean_token_accuracy": 0.172017665207386,
      "num_tokens": 29197932.0,
      "step": 13315
    },
    {
      "entropy": 5.521118974685669,
      "epoch": 1.3219531560142914,
      "grad_norm": 0.828125,
      "learning_rate": 0.00048327667482346895,
      "loss": 5.0936,
      "mean_token_accuracy": 0.17421265244483947,
      "num_tokens": 29209529.0,
      "step": 13320
    },
    {
      "entropy": 5.557190799713135,
      "epoch": 1.322449384676459,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004832632677310647,
      "loss": 5.1479,
      "mean_token_accuracy": 0.1710665762424469,
      "num_tokens": 29219027.0,
      "step": 13325
    },
    {
      "entropy": 5.564580869674683,
      "epoch": 1.3229456133386264,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048324985547418026,
      "loss": 5.1766,
      "mean_token_accuracy": 0.1619454875588417,
      "num_tokens": 29229647.0,
      "step": 13330
    },
    {
      "entropy": 5.586260890960693,
      "epoch": 1.3234418420007938,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004832364380531482,
      "loss": 5.2023,
      "mean_token_accuracy": 0.17671006321907043,
      "num_tokens": 29240263.0,
      "step": 13335
    },
    {
      "entropy": 5.613921594619751,
      "epoch": 1.3239380706629615,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00048322301546830103,
      "loss": 5.2219,
      "mean_token_accuracy": 0.16260263323783875,
      "num_tokens": 29251115.0,
      "step": 13340
    },
    {
      "entropy": 5.550369596481323,
      "epoch": 1.324434299325129,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004832095877199718,
      "loss": 5.1877,
      "mean_token_accuracy": 0.16290279626846313,
      "num_tokens": 29261888.0,
      "step": 13345
    },
    {
      "entropy": 5.511624717712403,
      "epoch": 1.3249305279872965,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048319615480849347,
      "loss": 5.0798,
      "mean_token_accuracy": 0.17703094482421874,
      "num_tokens": 29271948.0,
      "step": 13350
    },
    {
      "entropy": 5.589355087280273,
      "epoch": 1.3254267566494642,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048318271673419914,
      "loss": 5.1475,
      "mean_token_accuracy": 0.1691489115357399,
      "num_tokens": 29282449.0,
      "step": 13355
    },
    {
      "entropy": 5.574784994125366,
      "epoch": 1.3259229853116317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048316927349742207,
      "loss": 5.1673,
      "mean_token_accuracy": 0.16736889630556107,
      "num_tokens": 29294280.0,
      "step": 13360
    },
    {
      "entropy": 5.582604932785034,
      "epoch": 1.3264192139737991,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004831558250984955,
      "loss": 5.1444,
      "mean_token_accuracy": 0.17172702252864838,
      "num_tokens": 29305143.0,
      "step": 13365
    },
    {
      "entropy": 5.55707278251648,
      "epoch": 1.3269154426359666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048314237153775303,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1741911068558693,
      "num_tokens": 29314697.0,
      "step": 13370
    },
    {
      "entropy": 5.498944044113159,
      "epoch": 1.327411671298134,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004831289128155283,
      "loss": 5.1032,
      "mean_token_accuracy": 0.16522867530584334,
      "num_tokens": 29325257.0,
      "step": 13375
    },
    {
      "entropy": 5.551614475250244,
      "epoch": 1.3279078999603018,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048311544893215496,
      "loss": 5.1999,
      "mean_token_accuracy": 0.16806020736694335,
      "num_tokens": 29335449.0,
      "step": 13380
    },
    {
      "entropy": 5.559842920303344,
      "epoch": 1.3284041286224693,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000483101979887967,
      "loss": 5.1986,
      "mean_token_accuracy": 0.16706978529691696,
      "num_tokens": 29346898.0,
      "step": 13385
    },
    {
      "entropy": 5.625718688964843,
      "epoch": 1.3289003572846367,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048308850568329825,
      "loss": 5.1502,
      "mean_token_accuracy": 0.16714577823877336,
      "num_tokens": 29356118.0,
      "step": 13390
    },
    {
      "entropy": 5.571316766738891,
      "epoch": 1.3293965859468042,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048307502631848306,
      "loss": 5.232,
      "mean_token_accuracy": 0.16015283465385438,
      "num_tokens": 29367597.0,
      "step": 13395
    },
    {
      "entropy": 5.531172037124634,
      "epoch": 1.329892814608972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048306154179385546,
      "loss": 5.1004,
      "mean_token_accuracy": 0.16730550527572632,
      "num_tokens": 29377829.0,
      "step": 13400
    },
    {
      "entropy": 5.534276533126831,
      "epoch": 1.3303890432711394,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004830480521097501,
      "loss": 5.0768,
      "mean_token_accuracy": 0.17089010179042816,
      "num_tokens": 29387945.0,
      "step": 13405
    },
    {
      "entropy": 5.485321569442749,
      "epoch": 1.3308852719333069,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004830345572665012,
      "loss": 5.0416,
      "mean_token_accuracy": 0.17866674959659576,
      "num_tokens": 29398269.0,
      "step": 13410
    },
    {
      "entropy": 5.594847726821899,
      "epoch": 1.3313815005954743,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048302105726444366,
      "loss": 5.2176,
      "mean_token_accuracy": 0.1652454912662506,
      "num_tokens": 29410946.0,
      "step": 13415
    },
    {
      "entropy": 5.540194940567017,
      "epoch": 1.3318777292576418,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048300755210391205,
      "loss": 5.0634,
      "mean_token_accuracy": 0.17001270204782487,
      "num_tokens": 29421770.0,
      "step": 13420
    },
    {
      "entropy": 5.565183496475219,
      "epoch": 1.3323739579198095,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004829940417852415,
      "loss": 5.2262,
      "mean_token_accuracy": 0.1675649493932724,
      "num_tokens": 29433717.0,
      "step": 13425
    },
    {
      "entropy": 5.483129692077637,
      "epoch": 1.332870186581977,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004829805263087668,
      "loss": 5.0462,
      "mean_token_accuracy": 0.17815277278423308,
      "num_tokens": 29444382.0,
      "step": 13430
    },
    {
      "entropy": 5.57854208946228,
      "epoch": 1.3333664152441445,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048296700567482324,
      "loss": 5.1906,
      "mean_token_accuracy": 0.1626274421811104,
      "num_tokens": 29454680.0,
      "step": 13435
    },
    {
      "entropy": 5.604870986938477,
      "epoch": 1.3338626439063122,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004829534798837461,
      "loss": 5.1948,
      "mean_token_accuracy": 0.16946307569742203,
      "num_tokens": 29465065.0,
      "step": 13440
    },
    {
      "entropy": 5.596048355102539,
      "epoch": 1.3343588725684796,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004829399489358707,
      "loss": 5.1326,
      "mean_token_accuracy": 0.16983155757188798,
      "num_tokens": 29475704.0,
      "step": 13445
    },
    {
      "entropy": 5.543466806411743,
      "epoch": 1.334855101230647,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004829264128315326,
      "loss": 5.1636,
      "mean_token_accuracy": 0.16718475222587587,
      "num_tokens": 29486418.0,
      "step": 13450
    },
    {
      "entropy": 5.616748714447022,
      "epoch": 1.3353513298928146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004829128715710676,
      "loss": 5.2497,
      "mean_token_accuracy": 0.15821778029203415,
      "num_tokens": 29498939.0,
      "step": 13455
    },
    {
      "entropy": 5.550244283676148,
      "epoch": 1.335847558554982,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004828993251548114,
      "loss": 5.0782,
      "mean_token_accuracy": 0.17496539801359176,
      "num_tokens": 29508525.0,
      "step": 13460
    },
    {
      "entropy": 5.57539849281311,
      "epoch": 1.3363437872171495,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004828857735830999,
      "loss": 5.1851,
      "mean_token_accuracy": 0.17055137306451798,
      "num_tokens": 29519962.0,
      "step": 13465
    },
    {
      "entropy": 5.549408578872681,
      "epoch": 1.3368400158793172,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004828722168562691,
      "loss": 5.1438,
      "mean_token_accuracy": 0.16870589256286622,
      "num_tokens": 29529586.0,
      "step": 13470
    },
    {
      "entropy": 5.581881761550903,
      "epoch": 1.3373362445414847,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004828586549746553,
      "loss": 5.2195,
      "mean_token_accuracy": 0.1658741146326065,
      "num_tokens": 29540670.0,
      "step": 13475
    },
    {
      "entropy": 5.575357866287232,
      "epoch": 1.3378324732036522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004828450879385947,
      "loss": 5.1893,
      "mean_token_accuracy": 0.16574628055095672,
      "num_tokens": 29551484.0,
      "step": 13480
    },
    {
      "entropy": 5.559174537658691,
      "epoch": 1.3383287018658199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004828315157484239,
      "loss": 5.1242,
      "mean_token_accuracy": 0.1689518302679062,
      "num_tokens": 29561596.0,
      "step": 13485
    },
    {
      "entropy": 5.597451829910279,
      "epoch": 1.3388249305279873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004828179384044793,
      "loss": 5.1414,
      "mean_token_accuracy": 0.17164919376373292,
      "num_tokens": 29572837.0,
      "step": 13490
    },
    {
      "entropy": 5.559686613082886,
      "epoch": 1.3393211591901548,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048280435590709756,
      "loss": 5.0999,
      "mean_token_accuracy": 0.17031278163194657,
      "num_tokens": 29584095.0,
      "step": 13495
    },
    {
      "entropy": 5.554077291488648,
      "epoch": 1.3398173878523223,
      "grad_norm": 0.875,
      "learning_rate": 0.0004827907682566157,
      "loss": 5.1583,
      "mean_token_accuracy": 0.16481661945581436,
      "num_tokens": 29594903.0,
      "step": 13500
    },
    {
      "entropy": 5.554687595367431,
      "epoch": 1.3403136165144898,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004827771754533704,
      "loss": 5.1349,
      "mean_token_accuracy": 0.17024704068899155,
      "num_tokens": 29605199.0,
      "step": 13505
    },
    {
      "entropy": 5.612432432174683,
      "epoch": 1.3408098451766575,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048276357749769895,
      "loss": 5.298,
      "mean_token_accuracy": 0.1594685807824135,
      "num_tokens": 29617535.0,
      "step": 13510
    },
    {
      "entropy": 5.547749996185303,
      "epoch": 1.341306073838825,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048274997438993845,
      "loss": 5.103,
      "mean_token_accuracy": 0.1747092202305794,
      "num_tokens": 29628227.0,
      "step": 13515
    },
    {
      "entropy": 5.576234626770019,
      "epoch": 1.3418023025009924,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048273636613042626,
      "loss": 5.2168,
      "mean_token_accuracy": 0.16662221401929855,
      "num_tokens": 29639112.0,
      "step": 13520
    },
    {
      "entropy": 5.476972389221191,
      "epoch": 1.3422985311631601,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004827227527194998,
      "loss": 5.1008,
      "mean_token_accuracy": 0.16669861525297164,
      "num_tokens": 29650768.0,
      "step": 13525
    },
    {
      "entropy": 5.599837732315064,
      "epoch": 1.3427947598253276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004827091341574967,
      "loss": 5.2133,
      "mean_token_accuracy": 0.1757693409919739,
      "num_tokens": 29662388.0,
      "step": 13530
    },
    {
      "entropy": 5.550512218475342,
      "epoch": 1.343290988487495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004826955104447546,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17778134495019912,
      "num_tokens": 29672231.0,
      "step": 13535
    },
    {
      "entropy": 5.544204378128052,
      "epoch": 1.3437872171496625,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004826818815816114,
      "loss": 5.1546,
      "mean_token_accuracy": 0.17466779053211212,
      "num_tokens": 29683019.0,
      "step": 13540
    },
    {
      "entropy": 5.569805860519409,
      "epoch": 1.34428344581183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048266824756840513,
      "loss": 5.1786,
      "mean_token_accuracy": 0.17269141227006912,
      "num_tokens": 29694233.0,
      "step": 13545
    },
    {
      "entropy": 5.584534788131714,
      "epoch": 1.3447796744739975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826546084054737,
      "loss": 5.2017,
      "mean_token_accuracy": 0.1722709447145462,
      "num_tokens": 29704023.0,
      "step": 13550
    },
    {
      "entropy": 5.688522911071777,
      "epoch": 1.3452759031361652,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004826409640931555,
      "loss": 5.2948,
      "mean_token_accuracy": 0.16358633190393448,
      "num_tokens": 29715407.0,
      "step": 13555
    },
    {
      "entropy": 5.51044020652771,
      "epoch": 1.3457721317983327,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048262731463178884,
      "loss": 5.1108,
      "mean_token_accuracy": 0.1736569583415985,
      "num_tokens": 29726223.0,
      "step": 13560
    },
    {
      "entropy": 5.50967493057251,
      "epoch": 1.3462683604605001,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004826136600217121,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17427556365728378,
      "num_tokens": 29735850.0,
      "step": 13565
    },
    {
      "entropy": 5.561479997634888,
      "epoch": 1.3467645891226678,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000482600000263264,
      "loss": 5.1974,
      "mean_token_accuracy": 0.16665230989456176,
      "num_tokens": 29747381.0,
      "step": 13570
    },
    {
      "entropy": 5.623589658737183,
      "epoch": 1.3472608177848353,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004825863353567832,
      "loss": 5.2435,
      "mean_token_accuracy": 0.1681100144982338,
      "num_tokens": 29758440.0,
      "step": 13575
    },
    {
      "entropy": 5.54497799873352,
      "epoch": 1.3477570464470028,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048257266530260865,
      "loss": 5.0884,
      "mean_token_accuracy": 0.16859708428382875,
      "num_tokens": 29770606.0,
      "step": 13580
    },
    {
      "entropy": 5.511098623275757,
      "epoch": 1.3482532751091703,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004825589901010792,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1719357877969742,
      "num_tokens": 29781469.0,
      "step": 13585
    },
    {
      "entropy": 5.545900821685791,
      "epoch": 1.3487495037713377,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004825453097525341,
      "loss": 5.2265,
      "mean_token_accuracy": 0.16708861589431762,
      "num_tokens": 29792847.0,
      "step": 13590
    },
    {
      "entropy": 5.576881408691406,
      "epoch": 1.3492457324335054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048253162425731253,
      "loss": 5.1444,
      "mean_token_accuracy": 0.16709426343441008,
      "num_tokens": 29802784.0,
      "step": 13595
    },
    {
      "entropy": 5.625353622436523,
      "epoch": 1.349741961095673,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004825179336157539,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1724647432565689,
      "num_tokens": 29812849.0,
      "step": 13600
    },
    {
      "entropy": 5.583516931533813,
      "epoch": 1.3502381897578404,
      "grad_norm": 1.0,
      "learning_rate": 0.00048250423782819765,
      "loss": 5.1366,
      "mean_token_accuracy": 0.16968629509210587,
      "num_tokens": 29823495.0,
      "step": 13605
    },
    {
      "entropy": 5.563792991638183,
      "epoch": 1.350734418420008,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004824905368949834,
      "loss": 5.1752,
      "mean_token_accuracy": 0.17043911367654802,
      "num_tokens": 29835295.0,
      "step": 13610
    },
    {
      "entropy": 5.5105194568634035,
      "epoch": 1.3512306470821756,
      "grad_norm": 1.0,
      "learning_rate": 0.0004824768308164509,
      "loss": 5.1182,
      "mean_token_accuracy": 0.17353827506303787,
      "num_tokens": 29845747.0,
      "step": 13615
    },
    {
      "entropy": 5.570322036743164,
      "epoch": 1.351726875744343,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004824631195929402,
      "loss": 5.1757,
      "mean_token_accuracy": 0.17365752309560775,
      "num_tokens": 29857111.0,
      "step": 13620
    },
    {
      "entropy": 5.5386354446411135,
      "epoch": 1.3522231044065105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004824494032247911,
      "loss": 5.1614,
      "mean_token_accuracy": 0.1667430192232132,
      "num_tokens": 29866919.0,
      "step": 13625
    },
    {
      "entropy": 5.5609229564666744,
      "epoch": 1.352719333068678,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004824356817123439,
      "loss": 5.111,
      "mean_token_accuracy": 0.17309495955705642,
      "num_tokens": 29878631.0,
      "step": 13630
    },
    {
      "entropy": 5.566390705108643,
      "epoch": 1.3532155617308455,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048242195505593865,
      "loss": 5.1576,
      "mean_token_accuracy": 0.17128145694732666,
      "num_tokens": 29890809.0,
      "step": 13635
    },
    {
      "entropy": 5.612673902511597,
      "epoch": 1.3537117903930131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048240822325591596,
      "loss": 5.2266,
      "mean_token_accuracy": 0.17028311640024185,
      "num_tokens": 29901789.0,
      "step": 13640
    },
    {
      "entropy": 5.5493237495422365,
      "epoch": 1.3542080190551806,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004823944863126162,
      "loss": 5.1439,
      "mean_token_accuracy": 0.16489779949188232,
      "num_tokens": 29912576.0,
      "step": 13645
    },
    {
      "entropy": 5.606939077377319,
      "epoch": 1.354704247717348,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004823807442263801,
      "loss": 5.2313,
      "mean_token_accuracy": 0.16457829773426055,
      "num_tokens": 29924076.0,
      "step": 13650
    },
    {
      "entropy": 5.50984411239624,
      "epoch": 1.3552004763795158,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048236699699754845,
      "loss": 5.1473,
      "mean_token_accuracy": 0.16828127354383468,
      "num_tokens": 29935854.0,
      "step": 13655
    },
    {
      "entropy": 5.553005313873291,
      "epoch": 1.3556967050416833,
      "grad_norm": 0.875,
      "learning_rate": 0.000482353244626462,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1676760733127594,
      "num_tokens": 29946389.0,
      "step": 13660
    },
    {
      "entropy": 5.592410373687744,
      "epoch": 1.3561929337038507,
      "grad_norm": 1.0,
      "learning_rate": 0.00048233948711346196,
      "loss": 5.1708,
      "mean_token_accuracy": 0.16371285617351533,
      "num_tokens": 29958170.0,
      "step": 13665
    },
    {
      "entropy": 5.499467563629151,
      "epoch": 1.3566891623660182,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048232572445888945,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17025741636753083,
      "num_tokens": 29969067.0,
      "step": 13670
    },
    {
      "entropy": 5.581834363937378,
      "epoch": 1.3571853910281857,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004823119566630857,
      "loss": 5.2002,
      "mean_token_accuracy": 0.16283644288778304,
      "num_tokens": 29979727.0,
      "step": 13675
    },
    {
      "entropy": 5.602648115158081,
      "epoch": 1.3576816196903534,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048229818372639216,
      "loss": 5.1517,
      "mean_token_accuracy": 0.170918744802475,
      "num_tokens": 29989397.0,
      "step": 13680
    },
    {
      "entropy": 5.4911645412445065,
      "epoch": 1.3581778483525209,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048228440564915026,
      "loss": 5.0936,
      "mean_token_accuracy": 0.1671386882662773,
      "num_tokens": 29999415.0,
      "step": 13685
    },
    {
      "entropy": 5.532493734359742,
      "epoch": 1.3586740770146883,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048227062243170186,
      "loss": 5.1204,
      "mean_token_accuracy": 0.1768400952219963,
      "num_tokens": 30010277.0,
      "step": 13690
    },
    {
      "entropy": 5.499434757232666,
      "epoch": 1.3591703056768558,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004822568340743886,
      "loss": 5.0264,
      "mean_token_accuracy": 0.1775421231985092,
      "num_tokens": 30020684.0,
      "step": 13695
    },
    {
      "entropy": 5.624256324768067,
      "epoch": 1.3596665343390235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004822430405775525,
      "loss": 5.2511,
      "mean_token_accuracy": 0.16538104712963103,
      "num_tokens": 30030916.0,
      "step": 13700
    },
    {
      "entropy": 5.646221733093261,
      "epoch": 1.360162763001191,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004822292419415355,
      "loss": 5.1546,
      "mean_token_accuracy": 0.16920381039381027,
      "num_tokens": 30041140.0,
      "step": 13705
    },
    {
      "entropy": 5.536784601211548,
      "epoch": 1.3606589916633585,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048221543816667984,
      "loss": 5.123,
      "mean_token_accuracy": 0.17938897013664246,
      "num_tokens": 30052765.0,
      "step": 13710
    },
    {
      "entropy": 5.5144249439239506,
      "epoch": 1.361155220325526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048220162925332786,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17117499709129333,
      "num_tokens": 30062511.0,
      "step": 13715
    },
    {
      "entropy": 5.671962833404541,
      "epoch": 1.3616514489876934,
      "grad_norm": 1.203125,
      "learning_rate": 0.000482187815201822,
      "loss": 5.2244,
      "mean_token_accuracy": 0.167324960231781,
      "num_tokens": 30073404.0,
      "step": 13720
    },
    {
      "entropy": 5.517077493667602,
      "epoch": 1.362147677649861,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048217399601250464,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17155174612998964,
      "num_tokens": 30084978.0,
      "step": 13725
    },
    {
      "entropy": 5.551313638687134,
      "epoch": 1.3626439063120286,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004821601716857187,
      "loss": 5.1527,
      "mean_token_accuracy": 0.17365549355745316,
      "num_tokens": 30094498.0,
      "step": 13730
    },
    {
      "entropy": 5.549349689483643,
      "epoch": 1.363140134974196,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048214634222180686,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1728315606713295,
      "num_tokens": 30104545.0,
      "step": 13735
    },
    {
      "entropy": 5.622053432464599,
      "epoch": 1.3636363636363638,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004821325076211121,
      "loss": 5.2511,
      "mean_token_accuracy": 0.16976122558116913,
      "num_tokens": 30114863.0,
      "step": 13740
    },
    {
      "entropy": 5.63802170753479,
      "epoch": 1.3641325922985312,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048211866788397753,
      "loss": 5.1546,
      "mean_token_accuracy": 0.1657309889793396,
      "num_tokens": 30125101.0,
      "step": 13745
    },
    {
      "entropy": 5.596569490432739,
      "epoch": 1.3646288209606987,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004821048230107462,
      "loss": 5.217,
      "mean_token_accuracy": 0.16965948939323425,
      "num_tokens": 30135081.0,
      "step": 13750
    },
    {
      "entropy": 5.68363242149353,
      "epoch": 1.3651250496228662,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004820909730017616,
      "loss": 5.252,
      "mean_token_accuracy": 0.1570356622338295,
      "num_tokens": 30147169.0,
      "step": 13755
    },
    {
      "entropy": 5.566925525665283,
      "epoch": 1.3656212782850337,
      "grad_norm": 1.0,
      "learning_rate": 0.00048207711785736707,
      "loss": 5.0539,
      "mean_token_accuracy": 0.17032631486654282,
      "num_tokens": 30157523.0,
      "step": 13760
    },
    {
      "entropy": 5.5020801544189455,
      "epoch": 1.3661175069472011,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00048206325757790624,
      "loss": 5.1353,
      "mean_token_accuracy": 0.17143903374671937,
      "num_tokens": 30168313.0,
      "step": 13765
    },
    {
      "entropy": 5.484775829315185,
      "epoch": 1.3666137356093688,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048204939216372285,
      "loss": 5.0938,
      "mean_token_accuracy": 0.16789537221193312,
      "num_tokens": 30179173.0,
      "step": 13770
    },
    {
      "entropy": 5.634145450592041,
      "epoch": 1.3671099642715363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004820355216151607,
      "loss": 5.1726,
      "mean_token_accuracy": 0.16780838519334793,
      "num_tokens": 30191666.0,
      "step": 13775
    },
    {
      "entropy": 5.6299543380737305,
      "epoch": 1.3676061929337038,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00048202164593256363,
      "loss": 5.1434,
      "mean_token_accuracy": 0.17016903758049012,
      "num_tokens": 30202752.0,
      "step": 13780
    },
    {
      "entropy": 5.55454912185669,
      "epoch": 1.3681024215958715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048200776511627585,
      "loss": 5.1628,
      "mean_token_accuracy": 0.17022909075021744,
      "num_tokens": 30213127.0,
      "step": 13785
    },
    {
      "entropy": 5.605447292327881,
      "epoch": 1.368598650258039,
      "grad_norm": 0.875,
      "learning_rate": 0.0004819938791666416,
      "loss": 5.2031,
      "mean_token_accuracy": 0.16675858646631242,
      "num_tokens": 30224782.0,
      "step": 13790
    },
    {
      "entropy": 5.586220932006836,
      "epoch": 1.3690948789202064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048197998808400524,
      "loss": 5.2004,
      "mean_token_accuracy": 0.17411552369594574,
      "num_tokens": 30235129.0,
      "step": 13795
    },
    {
      "entropy": 5.598729276657105,
      "epoch": 1.369591107582374,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004819660918687111,
      "loss": 5.2182,
      "mean_token_accuracy": 0.16212333142757415,
      "num_tokens": 30247304.0,
      "step": 13800
    },
    {
      "entropy": 5.55193681716919,
      "epoch": 1.3700873362445414,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004819521905211039,
      "loss": 5.0744,
      "mean_token_accuracy": 0.1715817242860794,
      "num_tokens": 30257900.0,
      "step": 13805
    },
    {
      "entropy": 5.648249530792237,
      "epoch": 1.370583564906709,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048193828404152826,
      "loss": 5.2134,
      "mean_token_accuracy": 0.16967812925577164,
      "num_tokens": 30269741.0,
      "step": 13810
    },
    {
      "entropy": 5.546833896636963,
      "epoch": 1.3710797935688765,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004819243724303291,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16100092679262162,
      "num_tokens": 30279382.0,
      "step": 13815
    },
    {
      "entropy": 5.555518388748169,
      "epoch": 1.371576022231044,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048191045568785144,
      "loss": 5.1764,
      "mean_token_accuracy": 0.16895545721054078,
      "num_tokens": 30289837.0,
      "step": 13820
    },
    {
      "entropy": 5.531003475189209,
      "epoch": 1.3720722508932117,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048189653381444024,
      "loss": 5.137,
      "mean_token_accuracy": 0.17053737342357636,
      "num_tokens": 30302138.0,
      "step": 13825
    },
    {
      "entropy": 5.582463216781616,
      "epoch": 1.3725684795553792,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004818826068104409,
      "loss": 5.211,
      "mean_token_accuracy": 0.16949550360441207,
      "num_tokens": 30313623.0,
      "step": 13830
    },
    {
      "entropy": 5.615308523178101,
      "epoch": 1.3730647082175467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004818686746761987,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16677648574113846,
      "num_tokens": 30324533.0,
      "step": 13835
    },
    {
      "entropy": 5.615139198303223,
      "epoch": 1.3735609368797141,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004818547374120591,
      "loss": 5.2643,
      "mean_token_accuracy": 0.15903359651565552,
      "num_tokens": 30335819.0,
      "step": 13840
    },
    {
      "entropy": 5.550124597549439,
      "epoch": 1.3740571655418816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004818407950183678,
      "loss": 5.1418,
      "mean_token_accuracy": 0.1725202888250351,
      "num_tokens": 30345947.0,
      "step": 13845
    },
    {
      "entropy": 5.505348348617554,
      "epoch": 1.374553394204049,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818268474954704,
      "loss": 5.1265,
      "mean_token_accuracy": 0.16660067439079285,
      "num_tokens": 30356972.0,
      "step": 13850
    },
    {
      "entropy": 5.630555105209351,
      "epoch": 1.3750496228662168,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004818128948437129,
      "loss": 5.203,
      "mean_token_accuracy": 0.1700090691447258,
      "num_tokens": 30368414.0,
      "step": 13855
    },
    {
      "entropy": 5.645846176147461,
      "epoch": 1.3755458515283843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817989370634412,
      "loss": 5.13,
      "mean_token_accuracy": 0.1743665501475334,
      "num_tokens": 30379141.0,
      "step": 13860
    },
    {
      "entropy": 5.56193323135376,
      "epoch": 1.3760420801905517,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048178497415500153,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17879080772399902,
      "num_tokens": 30388985.0,
      "step": 13865
    },
    {
      "entropy": 5.529311609268189,
      "epoch": 1.3765383088527194,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004817710061187401,
      "loss": 5.127,
      "mean_token_accuracy": 0.1817374348640442,
      "num_tokens": 30399599.0,
      "step": 13870
    },
    {
      "entropy": 5.578681468963623,
      "epoch": 1.377034537514887,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048175703295500323,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16389575451612473,
      "num_tokens": 30411143.0,
      "step": 13875
    },
    {
      "entropy": 5.637400388717651,
      "epoch": 1.3775307661770544,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048174305466413755,
      "loss": 5.1917,
      "mean_token_accuracy": 0.17087964117527008,
      "num_tokens": 30422339.0,
      "step": 13880
    },
    {
      "entropy": 5.548777103424072,
      "epoch": 1.3780269948392219,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048172907124648947,
      "loss": 5.1203,
      "mean_token_accuracy": 0.16978117525577546,
      "num_tokens": 30432588.0,
      "step": 13885
    },
    {
      "entropy": 5.562787055969238,
      "epoch": 1.3785232235013893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004817150827024059,
      "loss": 5.1807,
      "mean_token_accuracy": 0.1717395693063736,
      "num_tokens": 30442272.0,
      "step": 13890
    },
    {
      "entropy": 5.522410535812378,
      "epoch": 1.379019452163557,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817010890322338,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1742080643773079,
      "num_tokens": 30453142.0,
      "step": 13895
    },
    {
      "entropy": 5.60747332572937,
      "epoch": 1.3795156808257245,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048168709023632006,
      "loss": 5.2019,
      "mean_token_accuracy": 0.1697382941842079,
      "num_tokens": 30465136.0,
      "step": 13900
    },
    {
      "entropy": 5.648171520233154,
      "epoch": 1.380011909487892,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048167308631501184,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17334889471530915,
      "num_tokens": 30476511.0,
      "step": 13905
    },
    {
      "entropy": 5.600936222076416,
      "epoch": 1.3805081381500597,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048165907726865643,
      "loss": 5.242,
      "mean_token_accuracy": 0.16183328479528428,
      "num_tokens": 30486547.0,
      "step": 13910
    },
    {
      "entropy": 5.58865532875061,
      "epoch": 1.3810043668122272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004816450630976012,
      "loss": 5.2012,
      "mean_token_accuracy": 0.1723966807126999,
      "num_tokens": 30496163.0,
      "step": 13915
    },
    {
      "entropy": 5.5741008758544925,
      "epoch": 1.3815005954743946,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004816310438021937,
      "loss": 5.1983,
      "mean_token_accuracy": 0.16799860894680024,
      "num_tokens": 30506979.0,
      "step": 13920
    },
    {
      "entropy": 5.608902311325073,
      "epoch": 1.381996824136562,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048161701938278156,
      "loss": 5.231,
      "mean_token_accuracy": 0.1579996109008789,
      "num_tokens": 30517492.0,
      "step": 13925
    },
    {
      "entropy": 5.512100267410278,
      "epoch": 1.3824930527987296,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048160298983971256,
      "loss": 5.1002,
      "mean_token_accuracy": 0.17216285467147827,
      "num_tokens": 30528614.0,
      "step": 13930
    },
    {
      "entropy": 5.558811902999878,
      "epoch": 1.382989281460897,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004815889551733346,
      "loss": 5.1925,
      "mean_token_accuracy": 0.16853821724653245,
      "num_tokens": 30539479.0,
      "step": 13935
    },
    {
      "entropy": 5.639448881149292,
      "epoch": 1.3834855101230648,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048157491538399573,
      "loss": 5.2204,
      "mean_token_accuracy": 0.16963192075490952,
      "num_tokens": 30549909.0,
      "step": 13940
    },
    {
      "entropy": 5.62214617729187,
      "epoch": 1.3839817387852322,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048156087047204407,
      "loss": 5.1198,
      "mean_token_accuracy": 0.1724722355604172,
      "num_tokens": 30562265.0,
      "step": 13945
    },
    {
      "entropy": 5.585242414474488,
      "epoch": 1.3844779674473997,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004815468204378279,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17177969813346863,
      "num_tokens": 30572622.0,
      "step": 13950
    },
    {
      "entropy": 5.526388788223267,
      "epoch": 1.3849741961095674,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004815327652816957,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17305349111557006,
      "num_tokens": 30583300.0,
      "step": 13955
    },
    {
      "entropy": 5.521373987197876,
      "epoch": 1.3854704247717349,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004815187050039959,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17485888451337814,
      "num_tokens": 30594460.0,
      "step": 13960
    },
    {
      "entropy": 5.5636396408081055,
      "epoch": 1.3859666534339024,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048150463960507724,
      "loss": 5.1245,
      "mean_token_accuracy": 0.17225345373153686,
      "num_tokens": 30604658.0,
      "step": 13965
    },
    {
      "entropy": 5.564675664901733,
      "epoch": 1.3864628820960698,
      "grad_norm": 0.875,
      "learning_rate": 0.00048149056908528855,
      "loss": 5.1178,
      "mean_token_accuracy": 0.1695321559906006,
      "num_tokens": 30616151.0,
      "step": 13970
    },
    {
      "entropy": 5.541741180419922,
      "epoch": 1.3869591107582373,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004814764934449786,
      "loss": 5.1713,
      "mean_token_accuracy": 0.17052412331104277,
      "num_tokens": 30627560.0,
      "step": 13975
    },
    {
      "entropy": 5.545062923431397,
      "epoch": 1.387455339420405,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048146241268449665,
      "loss": 5.1565,
      "mean_token_accuracy": 0.1733631819486618,
      "num_tokens": 30638490.0,
      "step": 13980
    },
    {
      "entropy": 5.632646131515503,
      "epoch": 1.3879515680825725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004814483268041917,
      "loss": 5.2629,
      "mean_token_accuracy": 0.16709994226694108,
      "num_tokens": 30649501.0,
      "step": 13985
    },
    {
      "entropy": 5.609997940063477,
      "epoch": 1.38844779674474,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004814342358044131,
      "loss": 5.138,
      "mean_token_accuracy": 0.17372792214155197,
      "num_tokens": 30660479.0,
      "step": 13990
    },
    {
      "entropy": 5.6739410877227785,
      "epoch": 1.3889440254069074,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048142013968551026,
      "loss": 5.3105,
      "mean_token_accuracy": 0.15868673771619796,
      "num_tokens": 30672015.0,
      "step": 13995
    },
    {
      "entropy": 5.484089279174805,
      "epoch": 1.3894402540690751,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048140603844783273,
      "loss": 5.1638,
      "mean_token_accuracy": 0.1681885227560997,
      "num_tokens": 30682612.0,
      "step": 14000
    },
    {
      "entropy": 5.516278839111328,
      "epoch": 1.3899364827312426,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004813919320917303,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17210320234298707,
      "num_tokens": 30693289.0,
      "step": 14005
    },
    {
      "entropy": 5.585805177688599,
      "epoch": 1.39043271139341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048137782061755265,
      "loss": 5.1499,
      "mean_token_accuracy": 0.17076454758644105,
      "num_tokens": 30705346.0,
      "step": 14010
    },
    {
      "entropy": 5.543377923965454,
      "epoch": 1.3909289400555775,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048136370402564974,
      "loss": 5.1481,
      "mean_token_accuracy": 0.16805513054132462,
      "num_tokens": 30715968.0,
      "step": 14015
    },
    {
      "entropy": 5.554496717453003,
      "epoch": 1.391425168717745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004813495823163716,
      "loss": 5.103,
      "mean_token_accuracy": 0.17531846165657045,
      "num_tokens": 30726225.0,
      "step": 14020
    },
    {
      "entropy": 5.550921869277954,
      "epoch": 1.3919213973799127,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004813354554900685,
      "loss": 5.0837,
      "mean_token_accuracy": 0.17345134168863297,
      "num_tokens": 30735603.0,
      "step": 14025
    },
    {
      "entropy": 5.529902935028076,
      "epoch": 1.3924176260420802,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004813213235470907,
      "loss": 5.1387,
      "mean_token_accuracy": 0.16814130246639253,
      "num_tokens": 30747414.0,
      "step": 14030
    },
    {
      "entropy": 5.603105926513672,
      "epoch": 1.3929138547042477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004813071864877887,
      "loss": 5.1928,
      "mean_token_accuracy": 0.1714324399828911,
      "num_tokens": 30758389.0,
      "step": 14035
    },
    {
      "entropy": 5.6162745475769045,
      "epoch": 1.3934100833664154,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000481293044312513,
      "loss": 5.2085,
      "mean_token_accuracy": 0.16657876819372178,
      "num_tokens": 30769484.0,
      "step": 14040
    },
    {
      "entropy": 5.53033413887024,
      "epoch": 1.3939063120285828,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048127889702161435,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17984972149133682,
      "num_tokens": 30779775.0,
      "step": 14045
    },
    {
      "entropy": 5.574795246124268,
      "epoch": 1.3944025406907503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004812647446154435,
      "loss": 5.2256,
      "mean_token_accuracy": 0.16242092698812485,
      "num_tokens": 30791456.0,
      "step": 14050
    },
    {
      "entropy": 5.539117670059204,
      "epoch": 1.3948987693529178,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004812505870943515,
      "loss": 5.0395,
      "mean_token_accuracy": 0.17584124356508254,
      "num_tokens": 30801659.0,
      "step": 14055
    },
    {
      "entropy": 5.660969066619873,
      "epoch": 1.3953949980150853,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004812364244586893,
      "loss": 5.2748,
      "mean_token_accuracy": 0.16599288284778596,
      "num_tokens": 30813389.0,
      "step": 14060
    },
    {
      "entropy": 5.530268001556396,
      "epoch": 1.395891226677253,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004812222567088082,
      "loss": 5.0519,
      "mean_token_accuracy": 0.16894870549440383,
      "num_tokens": 30823725.0,
      "step": 14065
    },
    {
      "entropy": 5.517103910446167,
      "epoch": 1.3963874553394204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004812080838450596,
      "loss": 5.0732,
      "mean_token_accuracy": 0.1763145372271538,
      "num_tokens": 30834448.0,
      "step": 14070
    },
    {
      "entropy": 5.565952587127685,
      "epoch": 1.396883684001588,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004811939058677947,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17120254039764404,
      "num_tokens": 30845891.0,
      "step": 14075
    },
    {
      "entropy": 5.532866477966309,
      "epoch": 1.3973799126637554,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004811797227773654,
      "loss": 5.0491,
      "mean_token_accuracy": 0.1779128283262253,
      "num_tokens": 30855170.0,
      "step": 14080
    },
    {
      "entropy": 5.507352066040039,
      "epoch": 1.397876141325923,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048116553457412317,
      "loss": 5.137,
      "mean_token_accuracy": 0.1686427965760231,
      "num_tokens": 30866527.0,
      "step": 14085
    },
    {
      "entropy": 5.521272659301758,
      "epoch": 1.3983723699880906,
      "grad_norm": 1.0,
      "learning_rate": 0.00048115134125841993,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1701352924108505,
      "num_tokens": 30876292.0,
      "step": 14090
    },
    {
      "entropy": 5.5562879085540775,
      "epoch": 1.398868598650258,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004811371428306076,
      "loss": 5.0812,
      "mean_token_accuracy": 0.16904907077550888,
      "num_tokens": 30887440.0,
      "step": 14095
    },
    {
      "entropy": 5.541204833984375,
      "epoch": 1.3993648273124255,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004811229392910384,
      "loss": 5.1322,
      "mean_token_accuracy": 0.17070431411266326,
      "num_tokens": 30897925.0,
      "step": 14100
    },
    {
      "entropy": 5.508321237564087,
      "epoch": 1.399861055974593,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004811087306400644,
      "loss": 5.1632,
      "mean_token_accuracy": 0.17505800873041152,
      "num_tokens": 30908413.0,
      "step": 14105
    },
    {
      "entropy": 5.581605243682861,
      "epoch": 1.4003572846367607,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048109451687803804,
      "loss": 5.1988,
      "mean_token_accuracy": 0.16573260426521302,
      "num_tokens": 30919149.0,
      "step": 14110
    },
    {
      "entropy": 5.608802556991577,
      "epoch": 1.4008535132989282,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004810802980053117,
      "loss": 5.2167,
      "mean_token_accuracy": 0.17029216885566711,
      "num_tokens": 30929883.0,
      "step": 14115
    },
    {
      "entropy": 5.552320718765259,
      "epoch": 1.4013497419610956,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048106607402223815,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17046646922826766,
      "num_tokens": 30941358.0,
      "step": 14120
    },
    {
      "entropy": 5.590188932418823,
      "epoch": 1.4018459706232633,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048105184492916995,
      "loss": 5.116,
      "mean_token_accuracy": 0.1693852186203003,
      "num_tokens": 30953289.0,
      "step": 14125
    },
    {
      "entropy": 5.544305801391602,
      "epoch": 1.4023421992854308,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048103761072645993,
      "loss": 5.0504,
      "mean_token_accuracy": 0.17555359452962876,
      "num_tokens": 30964304.0,
      "step": 14130
    },
    {
      "entropy": 5.472861814498901,
      "epoch": 1.4028384279475983,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048102337141446116,
      "loss": 5.0813,
      "mean_token_accuracy": 0.17812082022428513,
      "num_tokens": 30974840.0,
      "step": 14135
    },
    {
      "entropy": 5.52657413482666,
      "epoch": 1.4033346566097658,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048100912699352666,
      "loss": 5.1386,
      "mean_token_accuracy": 0.17427180856466293,
      "num_tokens": 30985805.0,
      "step": 14140
    },
    {
      "entropy": 5.6107367992401125,
      "epoch": 1.4038308852719332,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004809948774640098,
      "loss": 5.157,
      "mean_token_accuracy": 0.1700943887233734,
      "num_tokens": 30998174.0,
      "step": 14145
    },
    {
      "entropy": 5.626579999923706,
      "epoch": 1.4043271139341007,
      "grad_norm": 0.875,
      "learning_rate": 0.0004809806228262639,
      "loss": 5.1594,
      "mean_token_accuracy": 0.16868547201156617,
      "num_tokens": 31010050.0,
      "step": 14150
    },
    {
      "entropy": 5.666939449310303,
      "epoch": 1.4048233425962684,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048096636308064236,
      "loss": 5.282,
      "mean_token_accuracy": 0.16214755773544312,
      "num_tokens": 31023069.0,
      "step": 14155
    },
    {
      "entropy": 5.590526008605957,
      "epoch": 1.4053195712584359,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809520982274989,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1730720043182373,
      "num_tokens": 31033512.0,
      "step": 14160
    },
    {
      "entropy": 5.544665002822876,
      "epoch": 1.4058157999206033,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048093782826718706,
      "loss": 5.1253,
      "mean_token_accuracy": 0.16534166634082795,
      "num_tokens": 31045731.0,
      "step": 14165
    },
    {
      "entropy": 5.513667058944702,
      "epoch": 1.406312028582771,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048092355320006086,
      "loss": 5.0422,
      "mean_token_accuracy": 0.17603939920663833,
      "num_tokens": 31055443.0,
      "step": 14170
    },
    {
      "entropy": 5.500078010559082,
      "epoch": 1.4068082572449385,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004809092730264744,
      "loss": 5.1271,
      "mean_token_accuracy": 0.17388175278902054,
      "num_tokens": 31067575.0,
      "step": 14175
    },
    {
      "entropy": 5.589727687835693,
      "epoch": 1.407304485907106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048089498774678155,
      "loss": 5.2191,
      "mean_token_accuracy": 0.16846903562545776,
      "num_tokens": 31078198.0,
      "step": 14180
    },
    {
      "entropy": 5.536030292510986,
      "epoch": 1.4078007145692735,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048088069736133664,
      "loss": 5.1105,
      "mean_token_accuracy": 0.171412156522274,
      "num_tokens": 31089961.0,
      "step": 14185
    },
    {
      "entropy": 5.556495809555054,
      "epoch": 1.408296943231441,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004808664018704941,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17024407237768174,
      "num_tokens": 31100534.0,
      "step": 14190
    },
    {
      "entropy": 5.647650289535522,
      "epoch": 1.4087931718936086,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004808521012746085,
      "loss": 5.1617,
      "mean_token_accuracy": 0.17265866845846176,
      "num_tokens": 31112016.0,
      "step": 14195
    },
    {
      "entropy": 5.535036420822143,
      "epoch": 1.4092894005557761,
      "grad_norm": 0.875,
      "learning_rate": 0.00048083779557403424,
      "loss": 5.1795,
      "mean_token_accuracy": 0.17014693915843965,
      "num_tokens": 31123659.0,
      "step": 14200
    },
    {
      "entropy": 5.579057693481445,
      "epoch": 1.4097856292179436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004808234847691262,
      "loss": 5.1249,
      "mean_token_accuracy": 0.166135074198246,
      "num_tokens": 31133872.0,
      "step": 14205
    },
    {
      "entropy": 5.5990574836730955,
      "epoch": 1.4102818578801113,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004808091688602393,
      "loss": 5.1766,
      "mean_token_accuracy": 0.16771792098879815,
      "num_tokens": 31145834.0,
      "step": 14210
    },
    {
      "entropy": 5.585374164581299,
      "epoch": 1.4107780865422788,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00048079484784772854,
      "loss": 5.2374,
      "mean_token_accuracy": 0.16646238714456557,
      "num_tokens": 31157037.0,
      "step": 14215
    },
    {
      "entropy": 5.56239800453186,
      "epoch": 1.4112743152044462,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048078052173194886,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17500678151845933,
      "num_tokens": 31166812.0,
      "step": 14220
    },
    {
      "entropy": 5.586292123794555,
      "epoch": 1.4117705438666137,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004807661905132557,
      "loss": 5.1142,
      "mean_token_accuracy": 0.16663268506526946,
      "num_tokens": 31177162.0,
      "step": 14225
    },
    {
      "entropy": 5.587984323501587,
      "epoch": 1.4122667725287812,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004807518541920045,
      "loss": 5.1599,
      "mean_token_accuracy": 0.16959411799907684,
      "num_tokens": 31187333.0,
      "step": 14230
    },
    {
      "entropy": 5.603143835067749,
      "epoch": 1.4127630011909487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048073751276855056,
      "loss": 5.1653,
      "mean_token_accuracy": 0.16832650154829026,
      "num_tokens": 31197869.0,
      "step": 14235
    },
    {
      "entropy": 5.5740173816680905,
      "epoch": 1.4132592298531164,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004807231662432497,
      "loss": 5.1745,
      "mean_token_accuracy": 0.16534608900547026,
      "num_tokens": 31208199.0,
      "step": 14240
    },
    {
      "entropy": 5.602038812637329,
      "epoch": 1.4137554585152838,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004807088146164575,
      "loss": 5.159,
      "mean_token_accuracy": 0.16738710999488832,
      "num_tokens": 31218903.0,
      "step": 14245
    },
    {
      "entropy": 5.56422986984253,
      "epoch": 1.4142516871774513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004806944578885301,
      "loss": 5.1586,
      "mean_token_accuracy": 0.16780061274766922,
      "num_tokens": 31231146.0,
      "step": 14250
    },
    {
      "entropy": 5.580062484741211,
      "epoch": 1.414747915839619,
      "grad_norm": 1.0,
      "learning_rate": 0.00048068009605982333,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1722253903746605,
      "num_tokens": 31240980.0,
      "step": 14255
    },
    {
      "entropy": 5.533846807479859,
      "epoch": 1.4152441445017865,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048066572913069334,
      "loss": 5.1218,
      "mean_token_accuracy": 0.16866952031850815,
      "num_tokens": 31252458.0,
      "step": 14260
    },
    {
      "entropy": 5.525400400161743,
      "epoch": 1.415740373163954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004806513571014965,
      "loss": 5.123,
      "mean_token_accuracy": 0.1676151141524315,
      "num_tokens": 31262979.0,
      "step": 14265
    },
    {
      "entropy": 5.5746557235717775,
      "epoch": 1.4162366018261214,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004806369799725891,
      "loss": 5.1061,
      "mean_token_accuracy": 0.1696681261062622,
      "num_tokens": 31274811.0,
      "step": 14270
    },
    {
      "entropy": 5.55975751876831,
      "epoch": 1.416732830488289,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004806225977443277,
      "loss": 5.0737,
      "mean_token_accuracy": 0.17958507388830186,
      "num_tokens": 31285573.0,
      "step": 14275
    },
    {
      "entropy": 5.606087398529053,
      "epoch": 1.4172290591504566,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004806082104170689,
      "loss": 5.2495,
      "mean_token_accuracy": 0.15837212055921554,
      "num_tokens": 31296993.0,
      "step": 14280
    },
    {
      "entropy": 5.520363998413086,
      "epoch": 1.417725287812624,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048059381799116967,
      "loss": 5.1365,
      "mean_token_accuracy": 0.16958276629447938,
      "num_tokens": 31307217.0,
      "step": 14285
    },
    {
      "entropy": 5.512201642990112,
      "epoch": 1.4182215164747916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004805794204669867,
      "loss": 5.1655,
      "mean_token_accuracy": 0.16892223805189133,
      "num_tokens": 31317939.0,
      "step": 14290
    },
    {
      "entropy": 5.627305889129639,
      "epoch": 1.418717745136959,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048056501784487707,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16949473470449447,
      "num_tokens": 31327731.0,
      "step": 14295
    },
    {
      "entropy": 5.75149450302124,
      "epoch": 1.4192139737991267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000480550610125198,
      "loss": 5.3088,
      "mean_token_accuracy": 0.1565898984670639,
      "num_tokens": 31339388.0,
      "step": 14300
    },
    {
      "entropy": 5.513819408416748,
      "epoch": 1.4197102024612942,
      "grad_norm": 1.0,
      "learning_rate": 0.0004805361973083066,
      "loss": 5.0,
      "mean_token_accuracy": 0.1785448133945465,
      "num_tokens": 31350186.0,
      "step": 14305
    },
    {
      "entropy": 5.547780323028564,
      "epoch": 1.4202064311234617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004805217793945605,
      "loss": 5.1617,
      "mean_token_accuracy": 0.16407860219478607,
      "num_tokens": 31361219.0,
      "step": 14310
    },
    {
      "entropy": 5.558486747741699,
      "epoch": 1.4207026597856292,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004805073563843171,
      "loss": 5.1932,
      "mean_token_accuracy": 0.16570080816745758,
      "num_tokens": 31372820.0,
      "step": 14315
    },
    {
      "entropy": 5.558166027069092,
      "epoch": 1.4211988884477966,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048049292827793413,
      "loss": 5.078,
      "mean_token_accuracy": 0.1778224989771843,
      "num_tokens": 31382779.0,
      "step": 14320
    },
    {
      "entropy": 5.6600871086120605,
      "epoch": 1.4216951171099643,
      "grad_norm": 1.0,
      "learning_rate": 0.0004804784950757693,
      "loss": 5.2795,
      "mean_token_accuracy": 0.1657313734292984,
      "num_tokens": 31394872.0,
      "step": 14325
    },
    {
      "entropy": 5.60533037185669,
      "epoch": 1.4221913457721318,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048046405677818057,
      "loss": 5.1827,
      "mean_token_accuracy": 0.172329680621624,
      "num_tokens": 31405767.0,
      "step": 14330
    },
    {
      "entropy": 5.588709449768066,
      "epoch": 1.4226875744342993,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048044961338552595,
      "loss": 5.1819,
      "mean_token_accuracy": 0.16919856518507004,
      "num_tokens": 31417255.0,
      "step": 14335
    },
    {
      "entropy": 5.5642153263092045,
      "epoch": 1.423183803096467,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004804351648981636,
      "loss": 5.1313,
      "mean_token_accuracy": 0.17326364070177078,
      "num_tokens": 31428494.0,
      "step": 14340
    },
    {
      "entropy": 5.521048069000244,
      "epoch": 1.4236800317586344,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004804207113164519,
      "loss": 5.0532,
      "mean_token_accuracy": 0.17620072066783904,
      "num_tokens": 31439221.0,
      "step": 14345
    },
    {
      "entropy": 5.562240743637085,
      "epoch": 1.424176260420802,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048040625264074915,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1742454946041107,
      "num_tokens": 31450575.0,
      "step": 14350
    },
    {
      "entropy": 5.487133741378784,
      "epoch": 1.4246724890829694,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048039178887141404,
      "loss": 5.128,
      "mean_token_accuracy": 0.17392273247241974,
      "num_tokens": 31462168.0,
      "step": 14355
    },
    {
      "entropy": 5.578575563430786,
      "epoch": 1.4251687177451369,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048037732000880503,
      "loss": 5.1458,
      "mean_token_accuracy": 0.17462095916271209,
      "num_tokens": 31472619.0,
      "step": 14360
    },
    {
      "entropy": 5.61643533706665,
      "epoch": 1.4256649464073046,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048036284605328106,
      "loss": 5.2304,
      "mean_token_accuracy": 0.16764091700315475,
      "num_tokens": 31485246.0,
      "step": 14365
    },
    {
      "entropy": 5.57112717628479,
      "epoch": 1.426161175069472,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004803483670052011,
      "loss": 5.1427,
      "mean_token_accuracy": 0.16424565613269806,
      "num_tokens": 31496258.0,
      "step": 14370
    },
    {
      "entropy": 5.563767957687378,
      "epoch": 1.4266574037316395,
      "grad_norm": 0.96875,
      "learning_rate": 0.000480333882864924,
      "loss": 5.1111,
      "mean_token_accuracy": 0.1731133744120598,
      "num_tokens": 31506671.0,
      "step": 14375
    },
    {
      "entropy": 5.598369121551514,
      "epoch": 1.427153632393807,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004803193936328092,
      "loss": 5.1461,
      "mean_token_accuracy": 0.16498031616210937,
      "num_tokens": 31517001.0,
      "step": 14380
    },
    {
      "entropy": 5.56523962020874,
      "epoch": 1.4276498610559747,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004803048993092158,
      "loss": 5.1476,
      "mean_token_accuracy": 0.17295006811618804,
      "num_tokens": 31529649.0,
      "step": 14385
    },
    {
      "entropy": 5.495660495758057,
      "epoch": 1.4281460897181422,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048029039989450317,
      "loss": 5.0801,
      "mean_token_accuracy": 0.17676850855350495,
      "num_tokens": 31540083.0,
      "step": 14390
    },
    {
      "entropy": 5.5972373485565186,
      "epoch": 1.4286423183803096,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048027589538903115,
      "loss": 5.1944,
      "mean_token_accuracy": 0.17078108787536622,
      "num_tokens": 31551400.0,
      "step": 14395
    },
    {
      "entropy": 5.606319904327393,
      "epoch": 1.4291385470424771,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004802613857931592,
      "loss": 5.1884,
      "mean_token_accuracy": 0.16484396457672118,
      "num_tokens": 31562430.0,
      "step": 14400
    },
    {
      "entropy": 5.60907883644104,
      "epoch": 1.4296347757046446,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004802468711072472,
      "loss": 5.228,
      "mean_token_accuracy": 0.1685176283121109,
      "num_tokens": 31573173.0,
      "step": 14405
    },
    {
      "entropy": 5.58203387260437,
      "epoch": 1.4301310043668123,
      "grad_norm": 0.890625,
      "learning_rate": 0.000480232351331655,
      "loss": 5.162,
      "mean_token_accuracy": 0.16922738403081894,
      "num_tokens": 31584072.0,
      "step": 14410
    },
    {
      "entropy": 5.492711544036865,
      "epoch": 1.4306272330289798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048021782646674277,
      "loss": 5.1025,
      "mean_token_accuracy": 0.17219563573598862,
      "num_tokens": 31594993.0,
      "step": 14415
    },
    {
      "entropy": 5.512808418273925,
      "epoch": 1.4311234616911472,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048020329651287064,
      "loss": 5.1208,
      "mean_token_accuracy": 0.1784738391637802,
      "num_tokens": 31604284.0,
      "step": 14420
    },
    {
      "entropy": 5.549369955062867,
      "epoch": 1.431619690353315,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000480188761470399,
      "loss": 5.2125,
      "mean_token_accuracy": 0.16662057340145112,
      "num_tokens": 31616188.0,
      "step": 14425
    },
    {
      "entropy": 5.652873516082764,
      "epoch": 1.4321159190154824,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004801742213396881,
      "loss": 5.1668,
      "mean_token_accuracy": 0.17495240420103073,
      "num_tokens": 31626606.0,
      "step": 14430
    },
    {
      "entropy": 5.526907587051392,
      "epoch": 1.4326121476776499,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048015967612109875,
      "loss": 5.0718,
      "mean_token_accuracy": 0.17991526573896408,
      "num_tokens": 31638050.0,
      "step": 14435
    },
    {
      "entropy": 5.5495758056640625,
      "epoch": 1.4331083763398174,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048014512581499136,
      "loss": 5.1083,
      "mean_token_accuracy": 0.17039374262094498,
      "num_tokens": 31648451.0,
      "step": 14440
    },
    {
      "entropy": 5.557555627822876,
      "epoch": 1.4336046050019848,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000480130570421727,
      "loss": 5.1259,
      "mean_token_accuracy": 0.16736395359039308,
      "num_tokens": 31659310.0,
      "step": 14445
    },
    {
      "entropy": 5.54738564491272,
      "epoch": 1.4341008336641523,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004801160099416665,
      "loss": 5.144,
      "mean_token_accuracy": 0.16734889149665833,
      "num_tokens": 31669594.0,
      "step": 14450
    },
    {
      "entropy": 5.55546236038208,
      "epoch": 1.43459706232632,
      "grad_norm": 1.0,
      "learning_rate": 0.0004801014443751709,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1631794258952141,
      "num_tokens": 31682034.0,
      "step": 14455
    },
    {
      "entropy": 5.6243712425231935,
      "epoch": 1.4350932909884875,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004800868737226015,
      "loss": 5.1809,
      "mean_token_accuracy": 0.1695346489548683,
      "num_tokens": 31692390.0,
      "step": 14460
    },
    {
      "entropy": 5.585965967178344,
      "epoch": 1.435589519650655,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048007229798431946,
      "loss": 5.1554,
      "mean_token_accuracy": 0.17067220211029052,
      "num_tokens": 31703690.0,
      "step": 14465
    },
    {
      "entropy": 5.572841930389404,
      "epoch": 1.4360857483128227,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004800577171606864,
      "loss": 5.1491,
      "mean_token_accuracy": 0.17541060745716094,
      "num_tokens": 31714655.0,
      "step": 14470
    },
    {
      "entropy": 5.5135705947875975,
      "epoch": 1.4365819769749901,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004800431312520638,
      "loss": 5.0622,
      "mean_token_accuracy": 0.17869818955659866,
      "num_tokens": 31727022.0,
      "step": 14475
    },
    {
      "entropy": 5.578457593917847,
      "epoch": 1.4370782056371576,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004800285402588134,
      "loss": 5.1087,
      "mean_token_accuracy": 0.17700978368520737,
      "num_tokens": 31738071.0,
      "step": 14480
    },
    {
      "entropy": 5.616035652160645,
      "epoch": 1.437574434299325,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004800139441812969,
      "loss": 5.2241,
      "mean_token_accuracy": 0.16469064950942994,
      "num_tokens": 31749342.0,
      "step": 14485
    },
    {
      "entropy": 5.595319414138794,
      "epoch": 1.4380706629614926,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047999934301987644,
      "loss": 5.2043,
      "mean_token_accuracy": 0.164322130382061,
      "num_tokens": 31761218.0,
      "step": 14490
    },
    {
      "entropy": 5.584850025177002,
      "epoch": 1.4385668916236602,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000479984736774914,
      "loss": 5.2176,
      "mean_token_accuracy": 0.16821368485689164,
      "num_tokens": 31773111.0,
      "step": 14495
    },
    {
      "entropy": 5.625985908508301,
      "epoch": 1.4390631202858277,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047997012544677176,
      "loss": 5.1596,
      "mean_token_accuracy": 0.17030325531959534,
      "num_tokens": 31782948.0,
      "step": 14500
    },
    {
      "entropy": 5.617455053329468,
      "epoch": 1.4395593489479952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004799555090358121,
      "loss": 5.1492,
      "mean_token_accuracy": 0.17062585502862931,
      "num_tokens": 31793992.0,
      "step": 14505
    },
    {
      "entropy": 5.604525947570801,
      "epoch": 1.440055577610163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004799408875423975,
      "loss": 5.1623,
      "mean_token_accuracy": 0.17094217836856843,
      "num_tokens": 31804830.0,
      "step": 14510
    },
    {
      "entropy": 5.551722431182862,
      "epoch": 1.4405518062723304,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004799262609668904,
      "loss": 5.1544,
      "mean_token_accuracy": 0.17265972197055818,
      "num_tokens": 31817200.0,
      "step": 14515
    },
    {
      "entropy": 5.549322271347046,
      "epoch": 1.4410480349344978,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004799116293096537,
      "loss": 5.1886,
      "mean_token_accuracy": 0.17027641981840133,
      "num_tokens": 31828226.0,
      "step": 14520
    },
    {
      "entropy": 5.536720705032349,
      "epoch": 1.4415442635966653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004798969925710501,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17812822461128236,
      "num_tokens": 31839635.0,
      "step": 14525
    },
    {
      "entropy": 5.568207740783691,
      "epoch": 1.4420404922588328,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047988235075144254,
      "loss": 5.118,
      "mean_token_accuracy": 0.17763577848672868,
      "num_tokens": 31850333.0,
      "step": 14530
    },
    {
      "entropy": 5.529878044128418,
      "epoch": 1.4425367209210003,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004798677038511942,
      "loss": 5.0999,
      "mean_token_accuracy": 0.16968237459659577,
      "num_tokens": 31861002.0,
      "step": 14535
    },
    {
      "entropy": 5.5809080600738525,
      "epoch": 1.443032949583168,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004798530518706683,
      "loss": 5.1846,
      "mean_token_accuracy": 0.16534571945667267,
      "num_tokens": 31871713.0,
      "step": 14540
    },
    {
      "entropy": 5.592178630828857,
      "epoch": 1.4435291782453354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004798383948102281,
      "loss": 5.1086,
      "mean_token_accuracy": 0.17212580144405365,
      "num_tokens": 31882381.0,
      "step": 14545
    },
    {
      "entropy": 5.593084096908569,
      "epoch": 1.444025406907503,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004798237326702371,
      "loss": 5.1413,
      "mean_token_accuracy": 0.17558342814445496,
      "num_tokens": 31892347.0,
      "step": 14550
    },
    {
      "entropy": 5.636951637268067,
      "epoch": 1.4445216355696706,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047980906545105884,
      "loss": 5.245,
      "mean_token_accuracy": 0.16285596340894698,
      "num_tokens": 31902751.0,
      "step": 14555
    },
    {
      "entropy": 5.540946340560913,
      "epoch": 1.445017864231838,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004797943931530571,
      "loss": 5.1107,
      "mean_token_accuracy": 0.17518988996744156,
      "num_tokens": 31913294.0,
      "step": 14560
    },
    {
      "entropy": 5.556108808517456,
      "epoch": 1.4455140928940056,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004797797157765958,
      "loss": 5.2433,
      "mean_token_accuracy": 0.17099759429693223,
      "num_tokens": 31925678.0,
      "step": 14565
    },
    {
      "entropy": 5.53159909248352,
      "epoch": 1.446010321556173,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004797650333220387,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17430275082588195,
      "num_tokens": 31934578.0,
      "step": 14570
    },
    {
      "entropy": 5.522039413452148,
      "epoch": 1.4465065502183405,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047975034578975006,
      "loss": 5.0967,
      "mean_token_accuracy": 0.17289697676897048,
      "num_tokens": 31945482.0,
      "step": 14575
    },
    {
      "entropy": 5.5491708278656,
      "epoch": 1.4470027788805082,
      "grad_norm": 1.03125,
      "learning_rate": 0.000479735653180094,
      "loss": 5.2184,
      "mean_token_accuracy": 0.1736990973353386,
      "num_tokens": 31957303.0,
      "step": 14580
    },
    {
      "entropy": 5.565976619720459,
      "epoch": 1.4474990075426757,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004797209554934349,
      "loss": 5.131,
      "mean_token_accuracy": 0.17097844183444977,
      "num_tokens": 31968601.0,
      "step": 14585
    },
    {
      "entropy": 5.571213436126709,
      "epoch": 1.4479952362048432,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004797062527301373,
      "loss": 5.1625,
      "mean_token_accuracy": 0.16937379986047746,
      "num_tokens": 31978987.0,
      "step": 14590
    },
    {
      "entropy": 5.56727237701416,
      "epoch": 1.4484914648670106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004796915448905657,
      "loss": 5.1449,
      "mean_token_accuracy": 0.17063424736261368,
      "num_tokens": 31990950.0,
      "step": 14595
    },
    {
      "entropy": 5.63119649887085,
      "epoch": 1.4489876935291783,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047967683197508486,
      "loss": 5.1505,
      "mean_token_accuracy": 0.16956538408994676,
      "num_tokens": 32001795.0,
      "step": 14600
    },
    {
      "entropy": 5.58901686668396,
      "epoch": 1.4494839221913458,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047966211398405966,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16371258944272996,
      "num_tokens": 32011775.0,
      "step": 14605
    },
    {
      "entropy": 5.679873847961426,
      "epoch": 1.4499801508535133,
      "grad_norm": 1.015625,
      "learning_rate": 0.000479647390917855,
      "loss": 5.2576,
      "mean_token_accuracy": 0.15702016204595565,
      "num_tokens": 32022674.0,
      "step": 14610
    },
    {
      "entropy": 5.6167519092559814,
      "epoch": 1.4504763795156808,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000479632662776836,
      "loss": 5.2116,
      "mean_token_accuracy": 0.17228844612836838,
      "num_tokens": 32033383.0,
      "step": 14615
    },
    {
      "entropy": 5.569708299636841,
      "epoch": 1.4509726081778482,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000479617929561368,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17254067063331605,
      "num_tokens": 32044066.0,
      "step": 14620
    },
    {
      "entropy": 5.588910150527954,
      "epoch": 1.451468836840016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004796031912718162,
      "loss": 5.1946,
      "mean_token_accuracy": 0.16409365385770797,
      "num_tokens": 32054650.0,
      "step": 14625
    },
    {
      "entropy": 5.643964147567749,
      "epoch": 1.4519650655021834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047958844790854606,
      "loss": 5.1918,
      "mean_token_accuracy": 0.16113529056310655,
      "num_tokens": 32065786.0,
      "step": 14630
    },
    {
      "entropy": 5.621777534484863,
      "epoch": 1.4524612941643509,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047957369947192337,
      "loss": 5.1386,
      "mean_token_accuracy": 0.17549979239702224,
      "num_tokens": 32076326.0,
      "step": 14635
    },
    {
      "entropy": 5.540287446975708,
      "epoch": 1.4529575228265186,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004795589459623136,
      "loss": 5.077,
      "mean_token_accuracy": 0.17128066420555116,
      "num_tokens": 32086915.0,
      "step": 14640
    },
    {
      "entropy": 5.566558218002319,
      "epoch": 1.453453751488686,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047954418738008295,
      "loss": 5.0553,
      "mean_token_accuracy": 0.17387728095054628,
      "num_tokens": 32096614.0,
      "step": 14645
    },
    {
      "entropy": 5.565318870544433,
      "epoch": 1.4539499801508535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004795294237255971,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1758300930261612,
      "num_tokens": 32107721.0,
      "step": 14650
    },
    {
      "entropy": 5.537308168411255,
      "epoch": 1.454446208813021,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047951465499922234,
      "loss": 5.1551,
      "mean_token_accuracy": 0.1668999895453453,
      "num_tokens": 32116621.0,
      "step": 14655
    },
    {
      "entropy": 5.581287622451782,
      "epoch": 1.4549424374751885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004794998812013248,
      "loss": 5.1915,
      "mean_token_accuracy": 0.16888875812292098,
      "num_tokens": 32126536.0,
      "step": 14660
    },
    {
      "entropy": 5.6214762210845945,
      "epoch": 1.4554386661373562,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047948510233227084,
      "loss": 5.249,
      "mean_token_accuracy": 0.15938101559877396,
      "num_tokens": 32137059.0,
      "step": 14665
    },
    {
      "entropy": 5.552427339553833,
      "epoch": 1.4559348947995236,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004794703183924269,
      "loss": 5.1862,
      "mean_token_accuracy": 0.17184670120477677,
      "num_tokens": 32148186.0,
      "step": 14670
    },
    {
      "entropy": 5.678130054473877,
      "epoch": 1.4564311234616911,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004794555293821597,
      "loss": 5.2359,
      "mean_token_accuracy": 0.15512611716985703,
      "num_tokens": 32158244.0,
      "step": 14675
    },
    {
      "entropy": 5.5849508285522464,
      "epoch": 1.4569273521238586,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000479440735301836,
      "loss": 5.0958,
      "mean_token_accuracy": 0.17547072619199752,
      "num_tokens": 32168404.0,
      "step": 14680
    },
    {
      "entropy": 5.557720184326172,
      "epoch": 1.4574235807860263,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004794259361518226,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17637096494436263,
      "num_tokens": 32178744.0,
      "step": 14685
    },
    {
      "entropy": 5.555965757369995,
      "epoch": 1.4579198094481938,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047941113193248643,
      "loss": 5.1908,
      "mean_token_accuracy": 0.16726192384958266,
      "num_tokens": 32189948.0,
      "step": 14690
    },
    {
      "entropy": 5.59451961517334,
      "epoch": 1.4584160381103612,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004793963226441947,
      "loss": 5.1848,
      "mean_token_accuracy": 0.16054908633232118,
      "num_tokens": 32200979.0,
      "step": 14695
    },
    {
      "entropy": 5.56847186088562,
      "epoch": 1.4589122667725287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047938150828731455,
      "loss": 5.1557,
      "mean_token_accuracy": 0.17022694498300553,
      "num_tokens": 32211569.0,
      "step": 14700
    },
    {
      "entropy": 5.599681758880616,
      "epoch": 1.4594084954346962,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004793666888622134,
      "loss": 5.1519,
      "mean_token_accuracy": 0.17239338755607606,
      "num_tokens": 32222429.0,
      "step": 14705
    },
    {
      "entropy": 5.588827514648438,
      "epoch": 1.459904724096864,
      "grad_norm": 1.0,
      "learning_rate": 0.0004793518643692587,
      "loss": 5.2065,
      "mean_token_accuracy": 0.16049029678106308,
      "num_tokens": 32234332.0,
      "step": 14710
    },
    {
      "entropy": 5.662119293212891,
      "epoch": 1.4604009527590314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004793370348088182,
      "loss": 5.2383,
      "mean_token_accuracy": 0.17273847609758378,
      "num_tokens": 32244798.0,
      "step": 14715
    },
    {
      "entropy": 5.621212530136108,
      "epoch": 1.4608971814211988,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004793222001812595,
      "loss": 5.151,
      "mean_token_accuracy": 0.17294489592313766,
      "num_tokens": 32257746.0,
      "step": 14720
    },
    {
      "entropy": 5.615923738479614,
      "epoch": 1.4613934100833665,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047930736048695045,
      "loss": 5.243,
      "mean_token_accuracy": 0.16386656910181047,
      "num_tokens": 32270692.0,
      "step": 14725
    },
    {
      "entropy": 5.556577730178833,
      "epoch": 1.461889638745534,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004792925157262592,
      "loss": 5.127,
      "mean_token_accuracy": 0.17496246695518494,
      "num_tokens": 32281350.0,
      "step": 14730
    },
    {
      "entropy": 5.608032941818237,
      "epoch": 1.4623858674077015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004792776658995537,
      "loss": 5.1659,
      "mean_token_accuracy": 0.16902094930410386,
      "num_tokens": 32293182.0,
      "step": 14735
    },
    {
      "entropy": 5.575396203994751,
      "epoch": 1.462882096069869,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004792628110072023,
      "loss": 5.1352,
      "mean_token_accuracy": 0.16860028207302094,
      "num_tokens": 32304903.0,
      "step": 14740
    },
    {
      "entropy": 5.64785852432251,
      "epoch": 1.4633783247320364,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004792479510495734,
      "loss": 5.3152,
      "mean_token_accuracy": 0.167886321246624,
      "num_tokens": 32316074.0,
      "step": 14745
    },
    {
      "entropy": 5.595069122314453,
      "epoch": 1.463874553394204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047923308602703535,
      "loss": 5.1298,
      "mean_token_accuracy": 0.1832073524594307,
      "num_tokens": 32327594.0,
      "step": 14750
    },
    {
      "entropy": 5.5650605201721195,
      "epoch": 1.4643707820563716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047921821593995684,
      "loss": 5.1734,
      "mean_token_accuracy": 0.1712091624736786,
      "num_tokens": 32339224.0,
      "step": 14755
    },
    {
      "entropy": 5.615823459625244,
      "epoch": 1.464867010718539,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004792033407887066,
      "loss": 5.1476,
      "mean_token_accuracy": 0.16771445125341417,
      "num_tokens": 32349451.0,
      "step": 14760
    },
    {
      "entropy": 5.70187931060791,
      "epoch": 1.4653632393807066,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004791884605736536,
      "loss": 5.2137,
      "mean_token_accuracy": 0.16422370076179504,
      "num_tokens": 32361177.0,
      "step": 14765
    },
    {
      "entropy": 5.6085371494293215,
      "epoch": 1.4658594680428743,
      "grad_norm": 1.0,
      "learning_rate": 0.00047917357529516676,
      "loss": 5.1274,
      "mean_token_accuracy": 0.16220116913318633,
      "num_tokens": 32372031.0,
      "step": 14770
    },
    {
      "entropy": 5.542742967605591,
      "epoch": 1.4663556967050417,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004791586849536152,
      "loss": 5.05,
      "mean_token_accuracy": 0.1774493232369423,
      "num_tokens": 32383936.0,
      "step": 14775
    },
    {
      "entropy": 5.565629720687866,
      "epoch": 1.4668519253672092,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791437895493682,
      "loss": 5.2006,
      "mean_token_accuracy": 0.1712719738483429,
      "num_tokens": 32394283.0,
      "step": 14780
    },
    {
      "entropy": 5.580109786987305,
      "epoch": 1.4673481540293767,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047912888908279507,
      "loss": 5.1845,
      "mean_token_accuracy": 0.16410365402698518,
      "num_tokens": 32405549.0,
      "step": 14785
    },
    {
      "entropy": 5.5451160907745365,
      "epoch": 1.4678443826915442,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004791139835542653,
      "loss": 5.1219,
      "mean_token_accuracy": 0.16977680027484893,
      "num_tokens": 32414896.0,
      "step": 14790
    },
    {
      "entropy": 5.581904649734497,
      "epoch": 1.4683406113537119,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047909907296414864,
      "loss": 5.1812,
      "mean_token_accuracy": 0.16953709125518798,
      "num_tokens": 32426863.0,
      "step": 14795
    },
    {
      "entropy": 5.545697021484375,
      "epoch": 1.4688368400158793,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047908415731281475,
      "loss": 5.167,
      "mean_token_accuracy": 0.17042745798826217,
      "num_tokens": 32437703.0,
      "step": 14800
    },
    {
      "entropy": 5.614040517807007,
      "epoch": 1.4693330686780468,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047906923660063344,
      "loss": 5.1945,
      "mean_token_accuracy": 0.1760001540184021,
      "num_tokens": 32448517.0,
      "step": 14805
    },
    {
      "entropy": 5.57507700920105,
      "epoch": 1.4698292973402145,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000479054310827975,
      "loss": 5.1824,
      "mean_token_accuracy": 0.17051427215337753,
      "num_tokens": 32458782.0,
      "step": 14810
    },
    {
      "entropy": 5.537093448638916,
      "epoch": 1.470325526002382,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047903937999520913,
      "loss": 5.0471,
      "mean_token_accuracy": 0.175661700963974,
      "num_tokens": 32470113.0,
      "step": 14815
    },
    {
      "entropy": 5.564340877532959,
      "epoch": 1.4708217546645495,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004790244441027064,
      "loss": 5.1379,
      "mean_token_accuracy": 0.1723157837986946,
      "num_tokens": 32482455.0,
      "step": 14820
    },
    {
      "entropy": 5.620560264587402,
      "epoch": 1.471317983326717,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004790095031508371,
      "loss": 5.1833,
      "mean_token_accuracy": 0.17033948451280595,
      "num_tokens": 32493536.0,
      "step": 14825
    },
    {
      "entropy": 5.541712284088135,
      "epoch": 1.4718142119888844,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047899455713997164,
      "loss": 5.0994,
      "mean_token_accuracy": 0.1727277860045433,
      "num_tokens": 32503991.0,
      "step": 14830
    },
    {
      "entropy": 5.5453300952911375,
      "epoch": 1.4723104406510519,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047897960607048084,
      "loss": 5.1712,
      "mean_token_accuracy": 0.16686980277299882,
      "num_tokens": 32515474.0,
      "step": 14835
    },
    {
      "entropy": 5.533483552932739,
      "epoch": 1.4728066693132196,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004789646499427353,
      "loss": 5.0738,
      "mean_token_accuracy": 0.17458314597606658,
      "num_tokens": 32526751.0,
      "step": 14840
    },
    {
      "entropy": 5.641532945632934,
      "epoch": 1.473302897975387,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004789496887571059,
      "loss": 5.2478,
      "mean_token_accuracy": 0.16765937209129333,
      "num_tokens": 32538191.0,
      "step": 14845
    },
    {
      "entropy": 5.580791378021241,
      "epoch": 1.4737991266375545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047893472251396373,
      "loss": 5.1763,
      "mean_token_accuracy": 0.16882820576429367,
      "num_tokens": 32548841.0,
      "step": 14850
    },
    {
      "entropy": 5.553547143936157,
      "epoch": 1.4742953552997222,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047891975121368,
      "loss": 5.1094,
      "mean_token_accuracy": 0.1710888296365738,
      "num_tokens": 32558599.0,
      "step": 14855
    },
    {
      "entropy": 5.537466955184937,
      "epoch": 1.4747915839618897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047890477485662563,
      "loss": 5.0793,
      "mean_token_accuracy": 0.18023186624050141,
      "num_tokens": 32568641.0,
      "step": 14860
    },
    {
      "entropy": 5.582048034667968,
      "epoch": 1.4752878126240572,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047888979344317236,
      "loss": 5.1829,
      "mean_token_accuracy": 0.16714959889650344,
      "num_tokens": 32578924.0,
      "step": 14865
    },
    {
      "entropy": 5.601500177383423,
      "epoch": 1.4757840412862246,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047887480697369155,
      "loss": 5.1444,
      "mean_token_accuracy": 0.1755972385406494,
      "num_tokens": 32590233.0,
      "step": 14870
    },
    {
      "entropy": 5.569285583496094,
      "epoch": 1.4762802699483921,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047885981544855475,
      "loss": 5.1678,
      "mean_token_accuracy": 0.17029452472925186,
      "num_tokens": 32600782.0,
      "step": 14875
    },
    {
      "entropy": 5.5236656665802,
      "epoch": 1.4767764986105598,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004788448188681339,
      "loss": 5.15,
      "mean_token_accuracy": 0.17760232239961624,
      "num_tokens": 32612611.0,
      "step": 14880
    },
    {
      "entropy": 5.5672379493713375,
      "epoch": 1.4772727272727273,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004788298172328007,
      "loss": 5.2191,
      "mean_token_accuracy": 0.17034464627504348,
      "num_tokens": 32623436.0,
      "step": 14885
    },
    {
      "entropy": 5.54984450340271,
      "epoch": 1.4777689559348948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047881481054292736,
      "loss": 5.0924,
      "mean_token_accuracy": 0.17105893790721893,
      "num_tokens": 32633356.0,
      "step": 14890
    },
    {
      "entropy": 5.556531000137329,
      "epoch": 1.4782651845970625,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047879979879888584,
      "loss": 5.1299,
      "mean_token_accuracy": 0.17775177955627441,
      "num_tokens": 32643927.0,
      "step": 14895
    },
    {
      "entropy": 5.5123598098754885,
      "epoch": 1.47876141325923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047878478200104845,
      "loss": 5.1198,
      "mean_token_accuracy": 0.17775781005620955,
      "num_tokens": 32655086.0,
      "step": 14900
    },
    {
      "entropy": 5.541594696044922,
      "epoch": 1.4792576419213974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004787697601497876,
      "loss": 5.0532,
      "mean_token_accuracy": 0.17415277063846588,
      "num_tokens": 32665298.0,
      "step": 14905
    },
    {
      "entropy": 5.551559162139893,
      "epoch": 1.479753870583565,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004787547332454758,
      "loss": 5.1153,
      "mean_token_accuracy": 0.16987581104040145,
      "num_tokens": 32676717.0,
      "step": 14910
    },
    {
      "entropy": 5.632111883163452,
      "epoch": 1.4802500992457324,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047873970128848567,
      "loss": 5.182,
      "mean_token_accuracy": 0.16685492396354676,
      "num_tokens": 32686886.0,
      "step": 14915
    },
    {
      "entropy": 5.55278754234314,
      "epoch": 1.4807463279078998,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004787246642791899,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17781495302915573,
      "num_tokens": 32698631.0,
      "step": 14920
    },
    {
      "entropy": 5.547709369659424,
      "epoch": 1.4812425565700675,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047870962221796147,
      "loss": 5.1327,
      "mean_token_accuracy": 0.16709522604942323,
      "num_tokens": 32710672.0,
      "step": 14925
    },
    {
      "entropy": 5.582622289657593,
      "epoch": 1.481738785232235,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004786945751051734,
      "loss": 5.1712,
      "mean_token_accuracy": 0.1726660281419754,
      "num_tokens": 32720858.0,
      "step": 14930
    },
    {
      "entropy": 5.531434440612793,
      "epoch": 1.4822350138944025,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004786795229411988,
      "loss": 5.0938,
      "mean_token_accuracy": 0.17355458736419677,
      "num_tokens": 32731592.0,
      "step": 14935
    },
    {
      "entropy": 5.652337121963501,
      "epoch": 1.4827312425565702,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047866446572641084,
      "loss": 5.16,
      "mean_token_accuracy": 0.16156432181596755,
      "num_tokens": 32742777.0,
      "step": 14940
    },
    {
      "entropy": 5.57788577079773,
      "epoch": 1.4832274712187377,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004786494034611831,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17413914054632187,
      "num_tokens": 32753035.0,
      "step": 14945
    },
    {
      "entropy": 5.543017578125,
      "epoch": 1.4837236998809051,
      "grad_norm": 1.03125,
      "learning_rate": 0.000478634336145889,
      "loss": 5.1235,
      "mean_token_accuracy": 0.17121087461709977,
      "num_tokens": 32764318.0,
      "step": 14950
    },
    {
      "entropy": 5.569959497451782,
      "epoch": 1.4842199285430726,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047861926378090197,
      "loss": 5.1512,
      "mean_token_accuracy": 0.16542640030384065,
      "num_tokens": 32775521.0,
      "step": 14955
    },
    {
      "entropy": 5.625946474075318,
      "epoch": 1.48471615720524,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004786041863665961,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17418644279241563,
      "num_tokens": 32787101.0,
      "step": 14960
    },
    {
      "entropy": 5.575314903259278,
      "epoch": 1.4852123858674078,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004785891039033451,
      "loss": 5.1674,
      "mean_token_accuracy": 0.17160650640726088,
      "num_tokens": 32798446.0,
      "step": 14965
    },
    {
      "entropy": 5.587006759643555,
      "epoch": 1.4857086145295753,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000478574016391523,
      "loss": 5.1615,
      "mean_token_accuracy": 0.16105914860963821,
      "num_tokens": 32809636.0,
      "step": 14970
    },
    {
      "entropy": 5.5336668491363525,
      "epoch": 1.4862048431917427,
      "grad_norm": 0.875,
      "learning_rate": 0.000478558923831504,
      "loss": 5.0631,
      "mean_token_accuracy": 0.1723748430609703,
      "num_tokens": 32820132.0,
      "step": 14975
    },
    {
      "entropy": 5.590545177459717,
      "epoch": 1.4867010718539102,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047854382622366235,
      "loss": 5.1571,
      "mean_token_accuracy": 0.17179925441741944,
      "num_tokens": 32830720.0,
      "step": 14980
    },
    {
      "entropy": 5.550995302200318,
      "epoch": 1.487197300516078,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047852872356837236,
      "loss": 5.1044,
      "mean_token_accuracy": 0.1757486268877983,
      "num_tokens": 32841621.0,
      "step": 14985
    },
    {
      "entropy": 5.646034812927246,
      "epoch": 1.4876935291782454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004785136158660085,
      "loss": 5.2408,
      "mean_token_accuracy": 0.1625895619392395,
      "num_tokens": 32852074.0,
      "step": 14990
    },
    {
      "entropy": 5.656193685531616,
      "epoch": 1.4881897578404129,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047849850311694567,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1747622787952423,
      "num_tokens": 32862556.0,
      "step": 14995
    },
    {
      "entropy": 5.586263418197632,
      "epoch": 1.4886859865025803,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047848338532155833,
      "loss": 5.1157,
      "mean_token_accuracy": 0.17979874461889267,
      "num_tokens": 32874407.0,
      "step": 15000
    },
    {
      "epoch": 1.4886859865025803,
      "eval_entropy": 5.377731622893954,
      "eval_loss": 5.247989177703857,
      "eval_mean_token_accuracy": 0.17137185383183595,
      "eval_num_tokens": 32874407.0,
      "eval_runtime": 20.5622,
      "eval_samples_per_second": 1637.424,
      "eval_steps_per_second": 204.696,
      "step": 15000
    },
    {
      "entropy": 5.584698486328125,
      "epoch": 1.4891822151647478,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047846826248022155,
      "loss": 5.1565,
      "mean_token_accuracy": 0.16499363034963607,
      "num_tokens": 32884986.0,
      "step": 15005
    },
    {
      "entropy": 5.647080707550049,
      "epoch": 1.4896784438269155,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00047845313459331034,
      "loss": 5.2097,
      "mean_token_accuracy": 0.1694749504327774,
      "num_tokens": 32896016.0,
      "step": 15010
    },
    {
      "entropy": 5.566056203842163,
      "epoch": 1.490174672489083,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047843800166119974,
      "loss": 5.1091,
      "mean_token_accuracy": 0.175386418402195,
      "num_tokens": 32905648.0,
      "step": 15015
    },
    {
      "entropy": 5.621082448959351,
      "epoch": 1.4906709011512504,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000478422863684265,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17434526234865189,
      "num_tokens": 32917103.0,
      "step": 15020
    },
    {
      "entropy": 5.620783567428589,
      "epoch": 1.4911671298134181,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047840772066288167,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17454504370689392,
      "num_tokens": 32928741.0,
      "step": 15025
    },
    {
      "entropy": 5.574712800979614,
      "epoch": 1.4916633584755856,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004783925725974252,
      "loss": 5.1717,
      "mean_token_accuracy": 0.16855937242507935,
      "num_tokens": 32939292.0,
      "step": 15030
    },
    {
      "entropy": 5.621930074691773,
      "epoch": 1.492159587137753,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004783774194882712,
      "loss": 5.1638,
      "mean_token_accuracy": 0.17142217010259628,
      "num_tokens": 32949940.0,
      "step": 15035
    },
    {
      "entropy": 5.553967046737671,
      "epoch": 1.4926558157999206,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004783622613357954,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17190817445516587,
      "num_tokens": 32961222.0,
      "step": 15040
    },
    {
      "entropy": 5.5207123279571535,
      "epoch": 1.493152044462088,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004783470981403737,
      "loss": 5.1163,
      "mean_token_accuracy": 0.17732470184564592,
      "num_tokens": 32972095.0,
      "step": 15045
    },
    {
      "entropy": 5.615609884262085,
      "epoch": 1.4936482731242555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004783319299023821,
      "loss": 5.1513,
      "mean_token_accuracy": 0.17321223616600037,
      "num_tokens": 32981834.0,
      "step": 15050
    },
    {
      "entropy": 5.686815690994263,
      "epoch": 1.4941445017864232,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004783167566221969,
      "loss": 5.246,
      "mean_token_accuracy": 0.16528500616550446,
      "num_tokens": 32994347.0,
      "step": 15055
    },
    {
      "entropy": 5.654731273651123,
      "epoch": 1.4946407304485907,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047830157830019423,
      "loss": 5.2007,
      "mean_token_accuracy": 0.16660542339086531,
      "num_tokens": 33004780.0,
      "step": 15060
    },
    {
      "entropy": 5.638878583908081,
      "epoch": 1.4951369591107582,
      "grad_norm": 1.0,
      "learning_rate": 0.0004782863949367505,
      "loss": 5.1622,
      "mean_token_accuracy": 0.16810104995965958,
      "num_tokens": 33015697.0,
      "step": 15065
    },
    {
      "entropy": 5.534800386428833,
      "epoch": 1.4956331877729259,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004782712065322421,
      "loss": 5.1316,
      "mean_token_accuracy": 0.169401878118515,
      "num_tokens": 33026563.0,
      "step": 15070
    },
    {
      "entropy": 5.623071575164795,
      "epoch": 1.4961294164350933,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004782560130870459,
      "loss": 5.1564,
      "mean_token_accuracy": 0.16503316760063172,
      "num_tokens": 33037290.0,
      "step": 15075
    },
    {
      "entropy": 5.614276933670044,
      "epoch": 1.4966256450972608,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047824081460153846,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17045436352491378,
      "num_tokens": 33047199.0,
      "step": 15080
    },
    {
      "entropy": 5.534988641738892,
      "epoch": 1.4971218737594283,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004782256110760969,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17150217145681382,
      "num_tokens": 33057253.0,
      "step": 15085
    },
    {
      "entropy": 5.613107347488404,
      "epoch": 1.4976181024215958,
      "grad_norm": 1.0625,
      "learning_rate": 0.000478210402511098,
      "loss": 5.2516,
      "mean_token_accuracy": 0.16386423110961915,
      "num_tokens": 33068228.0,
      "step": 15090
    },
    {
      "entropy": 5.648596858978271,
      "epoch": 1.4981143310837635,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004781951889069189,
      "loss": 5.2075,
      "mean_token_accuracy": 0.1712303191423416,
      "num_tokens": 33079285.0,
      "step": 15095
    },
    {
      "entropy": 5.5636553287506105,
      "epoch": 1.498610559745931,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047817997026393714,
      "loss": 5.0726,
      "mean_token_accuracy": 0.17549518793821334,
      "num_tokens": 33091253.0,
      "step": 15100
    },
    {
      "entropy": 5.5365032196044925,
      "epoch": 1.4991067884080984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047816474658252985,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17224933356046676,
      "num_tokens": 33102492.0,
      "step": 15105
    },
    {
      "entropy": 5.534586954116821,
      "epoch": 1.499603017070266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047814951786307467,
      "loss": 5.1187,
      "mean_token_accuracy": 0.1717909649014473,
      "num_tokens": 33115038.0,
      "step": 15110
    },
    {
      "entropy": 5.6050327777862545,
      "epoch": 1.5000992457324336,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004781342841059492,
      "loss": 5.0973,
      "mean_token_accuracy": 0.17550544142723085,
      "num_tokens": 33124470.0,
      "step": 15115
    },
    {
      "entropy": 5.561952066421509,
      "epoch": 1.500595474394601,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004781190453115311,
      "loss": 5.1111,
      "mean_token_accuracy": 0.17133009433746338,
      "num_tokens": 33135994.0,
      "step": 15120
    },
    {
      "entropy": 5.5293896198272705,
      "epoch": 1.5010917030567685,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047810380148019847,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17059921622276306,
      "num_tokens": 33145783.0,
      "step": 15125
    },
    {
      "entropy": 5.5610555648803714,
      "epoch": 1.501587931718936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047808855261232915,
      "loss": 5.2274,
      "mean_token_accuracy": 0.17403534203767776,
      "num_tokens": 33156404.0,
      "step": 15130
    },
    {
      "entropy": 5.598877334594727,
      "epoch": 1.5020841603811035,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047807329870830134,
      "loss": 5.2146,
      "mean_token_accuracy": 0.17088855803012848,
      "num_tokens": 33168226.0,
      "step": 15135
    },
    {
      "entropy": 5.573947620391846,
      "epoch": 1.5025803890432712,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004780580397684933,
      "loss": 5.1315,
      "mean_token_accuracy": 0.1691361203789711,
      "num_tokens": 33178265.0,
      "step": 15140
    },
    {
      "entropy": 5.540897464752197,
      "epoch": 1.5030766177054387,
      "grad_norm": 0.875,
      "learning_rate": 0.0004780427757932835,
      "loss": 5.0995,
      "mean_token_accuracy": 0.17740151137113572,
      "num_tokens": 33188992.0,
      "step": 15145
    },
    {
      "entropy": 5.55078911781311,
      "epoch": 1.5035728463676064,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004780275067830503,
      "loss": 5.1754,
      "mean_token_accuracy": 0.16840895712375642,
      "num_tokens": 33199424.0,
      "step": 15150
    },
    {
      "entropy": 5.55648946762085,
      "epoch": 1.5040690750297738,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004780122327381725,
      "loss": 5.0851,
      "mean_token_accuracy": 0.1815241426229477,
      "num_tokens": 33210900.0,
      "step": 15155
    },
    {
      "entropy": 5.633826732635498,
      "epoch": 1.5045653036919413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047799695365902865,
      "loss": 5.1798,
      "mean_token_accuracy": 0.1665154829621315,
      "num_tokens": 33220796.0,
      "step": 15160
    },
    {
      "entropy": 5.561729001998901,
      "epoch": 1.5050615323541088,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004779816695459979,
      "loss": 5.1446,
      "mean_token_accuracy": 0.1707197204232216,
      "num_tokens": 33233454.0,
      "step": 15165
    },
    {
      "entropy": 5.587725973129272,
      "epoch": 1.5055577610162763,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004779663803994591,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17555942982435227,
      "num_tokens": 33243574.0,
      "step": 15170
    },
    {
      "entropy": 5.5678260803222654,
      "epoch": 1.5060539896784437,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004779510862197914,
      "loss": 5.1194,
      "mean_token_accuracy": 0.1709432676434517,
      "num_tokens": 33256298.0,
      "step": 15175
    },
    {
      "entropy": 5.6033087253570555,
      "epoch": 1.5065502183406112,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004779357870073741,
      "loss": 5.2302,
      "mean_token_accuracy": 0.16859439760446548,
      "num_tokens": 33268089.0,
      "step": 15180
    },
    {
      "entropy": 5.581967210769653,
      "epoch": 1.507046447002779,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047792048276258653,
      "loss": 5.1401,
      "mean_token_accuracy": 0.16609764248132705,
      "num_tokens": 33279955.0,
      "step": 15185
    },
    {
      "entropy": 5.577269411087036,
      "epoch": 1.5075426756649464,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004779051734858083,
      "loss": 5.1229,
      "mean_token_accuracy": 0.17021737098693848,
      "num_tokens": 33291470.0,
      "step": 15190
    },
    {
      "entropy": 5.61254916191101,
      "epoch": 1.508038904327114,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047788985917741894,
      "loss": 5.1653,
      "mean_token_accuracy": 0.17523350864648818,
      "num_tokens": 33304169.0,
      "step": 15195
    },
    {
      "entropy": 5.504221105575562,
      "epoch": 1.5085351329892815,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047787453983779826,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18250815123319625,
      "num_tokens": 33314436.0,
      "step": 15200
    },
    {
      "entropy": 5.556545782089233,
      "epoch": 1.509031361651449,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004778592154673262,
      "loss": 5.2239,
      "mean_token_accuracy": 0.16504380851984024,
      "num_tokens": 33325361.0,
      "step": 15205
    },
    {
      "entropy": 5.483824634552002,
      "epoch": 1.5095275903136165,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047784388606638263,
      "loss": 5.095,
      "mean_token_accuracy": 0.167599880695343,
      "num_tokens": 33336644.0,
      "step": 15210
    },
    {
      "entropy": 5.560460376739502,
      "epoch": 1.510023818975784,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004778285516353478,
      "loss": 5.1464,
      "mean_token_accuracy": 0.1703789636492729,
      "num_tokens": 33348418.0,
      "step": 15215
    },
    {
      "entropy": 5.577379512786865,
      "epoch": 1.5105200476379514,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047781321217460205,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17845593392848969,
      "num_tokens": 33360615.0,
      "step": 15220
    },
    {
      "entropy": 5.566539764404297,
      "epoch": 1.5110162763001191,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047779786768452554,
      "loss": 5.1054,
      "mean_token_accuracy": 0.16962546110153198,
      "num_tokens": 33372030.0,
      "step": 15225
    },
    {
      "entropy": 5.5845033645629885,
      "epoch": 1.5115125049622866,
      "grad_norm": 1.0,
      "learning_rate": 0.00047778251816549895,
      "loss": 5.1116,
      "mean_token_accuracy": 0.17124814242124559,
      "num_tokens": 33382834.0,
      "step": 15230
    },
    {
      "entropy": 5.558841562271118,
      "epoch": 1.512008733624454,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047776716361790284,
      "loss": 5.1274,
      "mean_token_accuracy": 0.17803507745265962,
      "num_tokens": 33393250.0,
      "step": 15235
    },
    {
      "entropy": 5.552697277069091,
      "epoch": 1.5125049622866218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047775180404211804,
      "loss": 5.0779,
      "mean_token_accuracy": 0.18078757971525192,
      "num_tokens": 33402778.0,
      "step": 15240
    },
    {
      "entropy": 5.623908948898316,
      "epoch": 1.5130011909487893,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004777364394385254,
      "loss": 5.1594,
      "mean_token_accuracy": 0.16896474361419678,
      "num_tokens": 33413565.0,
      "step": 15245
    },
    {
      "entropy": 5.610683536529541,
      "epoch": 1.5134974196109567,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047772106980750593,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17064012736082076,
      "num_tokens": 33424108.0,
      "step": 15250
    },
    {
      "entropy": 5.55219464302063,
      "epoch": 1.5139936482731242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004777056951494407,
      "loss": 5.1937,
      "mean_token_accuracy": 0.1672116920351982,
      "num_tokens": 33435683.0,
      "step": 15255
    },
    {
      "entropy": 5.544473505020141,
      "epoch": 1.5144898769352917,
      "grad_norm": 1.078125,
      "learning_rate": 0.000477690315464711,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17356909513473512,
      "num_tokens": 33446267.0,
      "step": 15260
    },
    {
      "entropy": 5.567493343353272,
      "epoch": 1.5149861055974592,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047767493075369827,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18663946390151978,
      "num_tokens": 33457901.0,
      "step": 15265
    },
    {
      "entropy": 5.6366698265075685,
      "epoch": 1.5154823342596269,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000477659541016784,
      "loss": 5.2372,
      "mean_token_accuracy": 0.16959605664014815,
      "num_tokens": 33469215.0,
      "step": 15270
    },
    {
      "entropy": 5.589060497283936,
      "epoch": 1.5159785629217943,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047764414625434983,
      "loss": 5.1029,
      "mean_token_accuracy": 0.17361409813165665,
      "num_tokens": 33480384.0,
      "step": 15275
    },
    {
      "entropy": 5.590121459960938,
      "epoch": 1.516474791583962,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004776287464667774,
      "loss": 5.1156,
      "mean_token_accuracy": 0.17584359794855117,
      "num_tokens": 33491404.0,
      "step": 15280
    },
    {
      "entropy": 5.635841131210327,
      "epoch": 1.5169710202461295,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047761334165444876,
      "loss": 5.2289,
      "mean_token_accuracy": 0.16371001154184342,
      "num_tokens": 33501888.0,
      "step": 15285
    },
    {
      "entropy": 5.581897592544555,
      "epoch": 1.517467248908297,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047759793181774586,
      "loss": 5.1784,
      "mean_token_accuracy": 0.17198490500450134,
      "num_tokens": 33513274.0,
      "step": 15290
    },
    {
      "entropy": 5.586254739761353,
      "epoch": 1.5179634775704645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047758251695705073,
      "loss": 5.1629,
      "mean_token_accuracy": 0.16762983798980713,
      "num_tokens": 33524188.0,
      "step": 15295
    },
    {
      "entropy": 5.6427366733551025,
      "epoch": 1.518459706232632,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047756709707274576,
      "loss": 5.2021,
      "mean_token_accuracy": 0.17214544713497162,
      "num_tokens": 33535207.0,
      "step": 15300
    },
    {
      "entropy": 5.632576942443848,
      "epoch": 1.5189559348947994,
      "grad_norm": 1.0,
      "learning_rate": 0.0004775516721652133,
      "loss": 5.1534,
      "mean_token_accuracy": 0.16575083136558533,
      "num_tokens": 33546154.0,
      "step": 15305
    },
    {
      "entropy": 5.5438182830810545,
      "epoch": 1.519452163556967,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004775362422348358,
      "loss": 5.0843,
      "mean_token_accuracy": 0.17624764889478683,
      "num_tokens": 33555365.0,
      "step": 15310
    },
    {
      "entropy": 5.567187738418579,
      "epoch": 1.5199483922191346,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047752080728199586,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17298535704612733,
      "num_tokens": 33566316.0,
      "step": 15315
    },
    {
      "entropy": 5.558042907714844,
      "epoch": 1.520444620881302,
      "grad_norm": 1.0,
      "learning_rate": 0.0004775053673070764,
      "loss": 5.0724,
      "mean_token_accuracy": 0.1824422374367714,
      "num_tokens": 33576549.0,
      "step": 15320
    },
    {
      "entropy": 5.5379798889160154,
      "epoch": 1.5209408495434698,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004774899223104601,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17658571749925614,
      "num_tokens": 33586681.0,
      "step": 15325
    },
    {
      "entropy": 5.655026245117187,
      "epoch": 1.5214370782056372,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004774744722925301,
      "loss": 5.2676,
      "mean_token_accuracy": 0.16330951154232026,
      "num_tokens": 33599201.0,
      "step": 15330
    },
    {
      "entropy": 5.615880250930786,
      "epoch": 1.5219333068678047,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004774590172536695,
      "loss": 5.1323,
      "mean_token_accuracy": 0.17253618091344833,
      "num_tokens": 33610217.0,
      "step": 15335
    },
    {
      "entropy": 5.579677581787109,
      "epoch": 1.5224295355299722,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047744355719426146,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1725304663181305,
      "num_tokens": 33620586.0,
      "step": 15340
    },
    {
      "entropy": 5.578193426132202,
      "epoch": 1.5229257641921397,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047742809211468953,
      "loss": 5.1314,
      "mean_token_accuracy": 0.17260937243700028,
      "num_tokens": 33630712.0,
      "step": 15345
    },
    {
      "entropy": 5.564428758621216,
      "epoch": 1.5234219928543071,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000477412622015337,
      "loss": 5.0998,
      "mean_token_accuracy": 0.1742965266108513,
      "num_tokens": 33640367.0,
      "step": 15350
    },
    {
      "entropy": 5.604958724975586,
      "epoch": 1.5239182215164748,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004773971468965876,
      "loss": 5.1872,
      "mean_token_accuracy": 0.16792533695697784,
      "num_tokens": 33651816.0,
      "step": 15355
    },
    {
      "entropy": 5.585981559753418,
      "epoch": 1.5244144501786423,
      "grad_norm": 1.0,
      "learning_rate": 0.00047738166675882516,
      "loss": 5.0631,
      "mean_token_accuracy": 0.17838085442781448,
      "num_tokens": 33663661.0,
      "step": 15360
    },
    {
      "entropy": 5.506061124801636,
      "epoch": 1.52491067884081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004773661816024335,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18224163502454757,
      "num_tokens": 33673670.0,
      "step": 15365
    },
    {
      "entropy": 5.545253038406372,
      "epoch": 1.5254069075029775,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004773506914277965,
      "loss": 5.2148,
      "mean_token_accuracy": 0.16352831125259398,
      "num_tokens": 33685434.0,
      "step": 15370
    },
    {
      "entropy": 5.592655611038208,
      "epoch": 1.525903136165145,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047733519623529853,
      "loss": 5.1676,
      "mean_token_accuracy": 0.1715659037232399,
      "num_tokens": 33696751.0,
      "step": 15375
    },
    {
      "entropy": 5.594798517227173,
      "epoch": 1.5263993648273124,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004773196960253236,
      "loss": 5.1033,
      "mean_token_accuracy": 0.17445114403963088,
      "num_tokens": 33708141.0,
      "step": 15380
    },
    {
      "entropy": 5.49371747970581,
      "epoch": 1.52689559348948,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004773041907982561,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18360562175512313,
      "num_tokens": 33719219.0,
      "step": 15385
    },
    {
      "entropy": 5.587513828277588,
      "epoch": 1.5273918221516474,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047728868055448064,
      "loss": 5.1476,
      "mean_token_accuracy": 0.17035889476537705,
      "num_tokens": 33730493.0,
      "step": 15390
    },
    {
      "entropy": 5.620696020126343,
      "epoch": 1.5278880508138148,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004772731652943818,
      "loss": 5.1829,
      "mean_token_accuracy": 0.16647169142961502,
      "num_tokens": 33740273.0,
      "step": 15395
    },
    {
      "entropy": 5.591207218170166,
      "epoch": 1.5283842794759825,
      "grad_norm": 1.0,
      "learning_rate": 0.0004772576450183444,
      "loss": 5.1551,
      "mean_token_accuracy": 0.17279521524906158,
      "num_tokens": 33752162.0,
      "step": 15400
    },
    {
      "entropy": 5.571941614151001,
      "epoch": 1.52888050813815,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047724211972675315,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17392804920673371,
      "num_tokens": 33763139.0,
      "step": 15405
    },
    {
      "entropy": 5.559553575515747,
      "epoch": 1.5293767368003177,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004772265894199931,
      "loss": 5.1444,
      "mean_token_accuracy": 0.17270752042531967,
      "num_tokens": 33774641.0,
      "step": 15410
    },
    {
      "entropy": 5.585414075851441,
      "epoch": 1.5298729654624852,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004772110540984494,
      "loss": 5.1208,
      "mean_token_accuracy": 0.18351941108703612,
      "num_tokens": 33785644.0,
      "step": 15415
    },
    {
      "entropy": 5.665813636779785,
      "epoch": 1.5303691941246527,
      "grad_norm": 0.875,
      "learning_rate": 0.0004771955137625073,
      "loss": 5.2628,
      "mean_token_accuracy": 0.16227585226297378,
      "num_tokens": 33797880.0,
      "step": 15420
    },
    {
      "entropy": 5.5254316329956055,
      "epoch": 1.5308654227868201,
      "grad_norm": 1.0,
      "learning_rate": 0.00047717996841255224,
      "loss": 5.0774,
      "mean_token_accuracy": 0.1714808911085129,
      "num_tokens": 33808782.0,
      "step": 15425
    },
    {
      "entropy": 5.5354416847229,
      "epoch": 1.5313616514489876,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047716441804896955,
      "loss": 5.1186,
      "mean_token_accuracy": 0.17557396590709687,
      "num_tokens": 33818483.0,
      "step": 15430
    },
    {
      "entropy": 5.569600200653076,
      "epoch": 1.531857880111155,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004771488626721449,
      "loss": 5.1207,
      "mean_token_accuracy": 0.16898479014635087,
      "num_tokens": 33828590.0,
      "step": 15435
    },
    {
      "entropy": 5.637904500961303,
      "epoch": 1.5323541087733228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047713330228246406,
      "loss": 5.2006,
      "mean_token_accuracy": 0.16827566027641297,
      "num_tokens": 33838978.0,
      "step": 15440
    },
    {
      "entropy": 5.596557903289795,
      "epoch": 1.5328503374354903,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004771177368803129,
      "loss": 5.1364,
      "mean_token_accuracy": 0.17314620763063432,
      "num_tokens": 33851244.0,
      "step": 15445
    },
    {
      "entropy": 5.55531873703003,
      "epoch": 1.533346566097658,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047710216646607733,
      "loss": 5.0259,
      "mean_token_accuracy": 0.18145370334386826,
      "num_tokens": 33861666.0,
      "step": 15450
    },
    {
      "entropy": 5.5860998153686525,
      "epoch": 1.5338427947598254,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047708659104014363,
      "loss": 5.1076,
      "mean_token_accuracy": 0.17519719451665877,
      "num_tokens": 33872416.0,
      "step": 15455
    },
    {
      "entropy": 5.537006473541259,
      "epoch": 1.534339023421993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004770710106028978,
      "loss": 5.101,
      "mean_token_accuracy": 0.17427923232316972,
      "num_tokens": 33881808.0,
      "step": 15460
    },
    {
      "entropy": 5.606580972671509,
      "epoch": 1.5348352520841604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047705542515472647,
      "loss": 5.1844,
      "mean_token_accuracy": 0.17471001148223878,
      "num_tokens": 33892925.0,
      "step": 15465
    },
    {
      "entropy": 5.599288558959961,
      "epoch": 1.5353314807463279,
      "grad_norm": 1.25,
      "learning_rate": 0.0004770398346960159,
      "loss": 5.126,
      "mean_token_accuracy": 0.16890227794647217,
      "num_tokens": 33903702.0,
      "step": 15470
    },
    {
      "entropy": 5.624025678634643,
      "epoch": 1.5358277094084953,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047702423922715284,
      "loss": 5.2026,
      "mean_token_accuracy": 0.17136026471853255,
      "num_tokens": 33917233.0,
      "step": 15475
    },
    {
      "entropy": 5.611566257476807,
      "epoch": 1.5363239380706628,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004770086387485239,
      "loss": 5.167,
      "mean_token_accuracy": 0.17637471854686737,
      "num_tokens": 33928362.0,
      "step": 15480
    },
    {
      "entropy": 5.654768991470337,
      "epoch": 1.5368201667328305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000476993033260516,
      "loss": 5.1882,
      "mean_token_accuracy": 0.16746950149536133,
      "num_tokens": 33938624.0,
      "step": 15485
    },
    {
      "entropy": 5.523258352279663,
      "epoch": 1.537316395394998,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004769774227635162,
      "loss": 5.086,
      "mean_token_accuracy": 0.1726633295416832,
      "num_tokens": 33950058.0,
      "step": 15490
    },
    {
      "entropy": 5.593592262268066,
      "epoch": 1.5378126240571657,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769618072579115,
      "loss": 5.1947,
      "mean_token_accuracy": 0.1671859622001648,
      "num_tokens": 33962404.0,
      "step": 15495
    },
    {
      "entropy": 5.6118971824646,
      "epoch": 1.5383088527193332,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00047694618674408917,
      "loss": 5.1299,
      "mean_token_accuracy": 0.1708168387413025,
      "num_tokens": 33973095.0,
      "step": 15500
    },
    {
      "entropy": 5.562897157669068,
      "epoch": 1.5388050813815006,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047693056122243655,
      "loss": 5.151,
      "mean_token_accuracy": 0.1700301229953766,
      "num_tokens": 33984152.0,
      "step": 15505
    },
    {
      "entropy": 5.6114137172698975,
      "epoch": 1.539301310043668,
      "grad_norm": 1.0,
      "learning_rate": 0.00047691493069334124,
      "loss": 5.1646,
      "mean_token_accuracy": 0.16597212105989456,
      "num_tokens": 33994854.0,
      "step": 15510
    },
    {
      "entropy": 5.617474937438965,
      "epoch": 1.5397975387058356,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047689929515719066,
      "loss": 5.169,
      "mean_token_accuracy": 0.17413103431463242,
      "num_tokens": 34006573.0,
      "step": 15515
    },
    {
      "entropy": 5.595598983764648,
      "epoch": 1.540293767368003,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047688365461437263,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1719505399465561,
      "num_tokens": 34016537.0,
      "step": 15520
    },
    {
      "entropy": 5.59567699432373,
      "epoch": 1.5407899960301707,
      "grad_norm": 1.0,
      "learning_rate": 0.00047686800906527493,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17549248337745665,
      "num_tokens": 34026825.0,
      "step": 15525
    },
    {
      "entropy": 5.63997631072998,
      "epoch": 1.5412862246923382,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004768523585102856,
      "loss": 5.1206,
      "mean_token_accuracy": 0.1743641495704651,
      "num_tokens": 34036786.0,
      "step": 15530
    },
    {
      "entropy": 5.599544477462769,
      "epoch": 1.541782453354506,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004768367029497928,
      "loss": 5.119,
      "mean_token_accuracy": 0.1772516116499901,
      "num_tokens": 34046784.0,
      "step": 15535
    },
    {
      "entropy": 5.608798885345459,
      "epoch": 1.5422786820166734,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047682104238418464,
      "loss": 5.1755,
      "mean_token_accuracy": 0.17145233750343322,
      "num_tokens": 34058250.0,
      "step": 15540
    },
    {
      "entropy": 5.528745698928833,
      "epoch": 1.5427749106788409,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047680537681384956,
      "loss": 5.1116,
      "mean_token_accuracy": 0.17416155487298965,
      "num_tokens": 34069969.0,
      "step": 15545
    },
    {
      "entropy": 5.6232696056365965,
      "epoch": 1.5432711393410083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004767897062391759,
      "loss": 5.2216,
      "mean_token_accuracy": 0.1682211622595787,
      "num_tokens": 34080817.0,
      "step": 15550
    },
    {
      "entropy": 5.589405488967896,
      "epoch": 1.5437673680031758,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004767740306605524,
      "loss": 5.1484,
      "mean_token_accuracy": 0.16989017874002457,
      "num_tokens": 34091138.0,
      "step": 15555
    },
    {
      "entropy": 5.5701953887939455,
      "epoch": 1.5442635966653433,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004767583500783677,
      "loss": 5.0701,
      "mean_token_accuracy": 0.1778403714299202,
      "num_tokens": 34102035.0,
      "step": 15560
    },
    {
      "entropy": 5.548117351531983,
      "epoch": 1.5447598253275108,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047674266449301074,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17230828553438188,
      "num_tokens": 34112533.0,
      "step": 15565
    },
    {
      "entropy": 5.5683173656463625,
      "epoch": 1.5452560539896785,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047672697390487034,
      "loss": 5.1479,
      "mean_token_accuracy": 0.17172082364559174,
      "num_tokens": 34124196.0,
      "step": 15570
    },
    {
      "entropy": 5.6053179740905765,
      "epoch": 1.545752282651846,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004767112783143357,
      "loss": 5.119,
      "mean_token_accuracy": 0.1758653923869133,
      "num_tokens": 34135635.0,
      "step": 15575
    },
    {
      "entropy": 5.551077556610108,
      "epoch": 1.5462485113140136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000476695577721796,
      "loss": 5.049,
      "mean_token_accuracy": 0.18202011287212372,
      "num_tokens": 34146202.0,
      "step": 15580
    },
    {
      "entropy": 5.57732253074646,
      "epoch": 1.5467447399761811,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047667987212764065,
      "loss": 5.0965,
      "mean_token_accuracy": 0.17440434396266938,
      "num_tokens": 34157354.0,
      "step": 15585
    },
    {
      "entropy": 5.580416631698609,
      "epoch": 1.5472409686383486,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047666416153225894,
      "loss": 5.1569,
      "mean_token_accuracy": 0.1673568829894066,
      "num_tokens": 34168563.0,
      "step": 15590
    },
    {
      "entropy": 5.579088449478149,
      "epoch": 1.547737197300516,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004766484459360406,
      "loss": 5.097,
      "mean_token_accuracy": 0.1780419871211052,
      "num_tokens": 34178604.0,
      "step": 15595
    },
    {
      "entropy": 5.580549812316894,
      "epoch": 1.5482334259626835,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047663272533937537,
      "loss": 5.1303,
      "mean_token_accuracy": 0.17384752929210662,
      "num_tokens": 34190321.0,
      "step": 15600
    },
    {
      "entropy": 5.528491401672364,
      "epoch": 1.548729654624851,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047661699974265294,
      "loss": 5.1511,
      "mean_token_accuracy": 0.16967853754758835,
      "num_tokens": 34201519.0,
      "step": 15605
    },
    {
      "entropy": 5.539703464508056,
      "epoch": 1.5492258832870187,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00047660126914626344,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17056724280118943,
      "num_tokens": 34212710.0,
      "step": 15610
    },
    {
      "entropy": 5.613179874420166,
      "epoch": 1.5497221119491862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004765855335505969,
      "loss": 5.2207,
      "mean_token_accuracy": 0.168785397708416,
      "num_tokens": 34223187.0,
      "step": 15615
    },
    {
      "entropy": 5.576096773147583,
      "epoch": 1.5502183406113537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004765697929560434,
      "loss": 5.161,
      "mean_token_accuracy": 0.17535919100046157,
      "num_tokens": 34234530.0,
      "step": 15620
    },
    {
      "entropy": 5.622468757629394,
      "epoch": 1.5507145692735214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047655404736299354,
      "loss": 5.1728,
      "mean_token_accuracy": 0.16912703067064286,
      "num_tokens": 34245003.0,
      "step": 15625
    },
    {
      "entropy": 5.565361881256104,
      "epoch": 1.5512107979356888,
      "grad_norm": 1.0,
      "learning_rate": 0.00047653829677183745,
      "loss": 5.1295,
      "mean_token_accuracy": 0.1782390981912613,
      "num_tokens": 34255630.0,
      "step": 15630
    },
    {
      "entropy": 5.565342283248901,
      "epoch": 1.5517070265978563,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047652254118296596,
      "loss": 5.1203,
      "mean_token_accuracy": 0.17997909039258958,
      "num_tokens": 34266716.0,
      "step": 15635
    },
    {
      "entropy": 5.597660160064697,
      "epoch": 1.5522032552600238,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004765067805967697,
      "loss": 5.1508,
      "mean_token_accuracy": 0.16779181361198425,
      "num_tokens": 34278905.0,
      "step": 15640
    },
    {
      "entropy": 5.562551975250244,
      "epoch": 1.5526994839221913,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004764910150136395,
      "loss": 5.103,
      "mean_token_accuracy": 0.17104328125715257,
      "num_tokens": 34289688.0,
      "step": 15645
    },
    {
      "entropy": 5.57739372253418,
      "epoch": 1.5531957125843587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047647524443396625,
      "loss": 5.1379,
      "mean_token_accuracy": 0.17226581871509553,
      "num_tokens": 34300964.0,
      "step": 15650
    },
    {
      "entropy": 5.56873459815979,
      "epoch": 1.5536919412465264,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047645946885814105,
      "loss": 5.1324,
      "mean_token_accuracy": 0.17225696593523027,
      "num_tokens": 34311435.0,
      "step": 15655
    },
    {
      "entropy": 5.5915374755859375,
      "epoch": 1.554188169908694,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004764436882865552,
      "loss": 5.1221,
      "mean_token_accuracy": 0.17058883905410765,
      "num_tokens": 34321742.0,
      "step": 15660
    },
    {
      "entropy": 5.636772108078003,
      "epoch": 1.5546843985708616,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004764279027195999,
      "loss": 5.2018,
      "mean_token_accuracy": 0.16625822484493255,
      "num_tokens": 34333796.0,
      "step": 15665
    },
    {
      "entropy": 5.691735553741455,
      "epoch": 1.555180627233029,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047641211215766673,
      "loss": 5.2524,
      "mean_token_accuracy": 0.1696785420179367,
      "num_tokens": 34344521.0,
      "step": 15670
    },
    {
      "entropy": 5.539317321777344,
      "epoch": 1.5556768558951966,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004763963166011471,
      "loss": 5.1332,
      "mean_token_accuracy": 0.1740424394607544,
      "num_tokens": 34355629.0,
      "step": 15675
    },
    {
      "entropy": 5.551272249221801,
      "epoch": 1.556173084557364,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004763805160504328,
      "loss": 5.1862,
      "mean_token_accuracy": 0.17267680764198304,
      "num_tokens": 34366935.0,
      "step": 15680
    },
    {
      "entropy": 5.563038921356201,
      "epoch": 1.5566693132195315,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004763647105059157,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17033598572015762,
      "num_tokens": 34378355.0,
      "step": 15685
    },
    {
      "entropy": 5.590302419662476,
      "epoch": 1.557165541881699,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004763488999679877,
      "loss": 5.0457,
      "mean_token_accuracy": 0.17383378595113755,
      "num_tokens": 34388811.0,
      "step": 15690
    },
    {
      "entropy": 5.65074200630188,
      "epoch": 1.5576617705438665,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00047633308443704076,
      "loss": 5.1604,
      "mean_token_accuracy": 0.1695440486073494,
      "num_tokens": 34399905.0,
      "step": 15695
    },
    {
      "entropy": 5.532579565048218,
      "epoch": 1.5581579992060341,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047631726391346717,
      "loss": 5.0505,
      "mean_token_accuracy": 0.1731049045920372,
      "num_tokens": 34410338.0,
      "step": 15700
    },
    {
      "entropy": 5.639637088775634,
      "epoch": 1.5586542278682016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004763014383976593,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1679335430264473,
      "num_tokens": 34421204.0,
      "step": 15705
    },
    {
      "entropy": 5.499878787994385,
      "epoch": 1.5591504565303693,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047628560789000945,
      "loss": 5.04,
      "mean_token_accuracy": 0.18360144644975662,
      "num_tokens": 34431002.0,
      "step": 15710
    },
    {
      "entropy": 5.525462627410889,
      "epoch": 1.5596466851925368,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047626977239091027,
      "loss": 5.0824,
      "mean_token_accuracy": 0.17336479127407073,
      "num_tokens": 34442446.0,
      "step": 15715
    },
    {
      "entropy": 5.55399522781372,
      "epoch": 1.5601429138547043,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004762539319007545,
      "loss": 5.0409,
      "mean_token_accuracy": 0.1787635251879692,
      "num_tokens": 34453590.0,
      "step": 15720
    },
    {
      "entropy": 5.629614925384521,
      "epoch": 1.5606391425168717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004762380864199348,
      "loss": 5.1762,
      "mean_token_accuracy": 0.17132293581962585,
      "num_tokens": 34465614.0,
      "step": 15725
    },
    {
      "entropy": 5.667820978164673,
      "epoch": 1.5611353711790392,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004762222359488443,
      "loss": 5.2805,
      "mean_token_accuracy": 0.16419804692268372,
      "num_tokens": 34476616.0,
      "step": 15730
    },
    {
      "entropy": 5.617419624328614,
      "epoch": 1.5616315998412067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004762063804878758,
      "loss": 5.0988,
      "mean_token_accuracy": 0.16436351984739303,
      "num_tokens": 34486260.0,
      "step": 15735
    },
    {
      "entropy": 5.585049962997436,
      "epoch": 1.5621278285033744,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047619052003742277,
      "loss": 5.1546,
      "mean_token_accuracy": 0.17054353058338165,
      "num_tokens": 34496392.0,
      "step": 15740
    },
    {
      "entropy": 5.577547216415406,
      "epoch": 1.5626240571655419,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004761746545978782,
      "loss": 5.1364,
      "mean_token_accuracy": 0.17610348016023636,
      "num_tokens": 34507218.0,
      "step": 15745
    },
    {
      "entropy": 5.621696329116821,
      "epoch": 1.5631202858277096,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004761587841696358,
      "loss": 5.1894,
      "mean_token_accuracy": 0.1702187776565552,
      "num_tokens": 34518888.0,
      "step": 15750
    },
    {
      "entropy": 5.594956541061402,
      "epoch": 1.563616514489877,
      "grad_norm": 1.0,
      "learning_rate": 0.000476142908753089,
      "loss": 5.1065,
      "mean_token_accuracy": 0.17332418262958527,
      "num_tokens": 34529865.0,
      "step": 15755
    },
    {
      "entropy": 5.633364009857178,
      "epoch": 1.5641127431520445,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004761270283486314,
      "loss": 5.1852,
      "mean_token_accuracy": 0.17102617025375366,
      "num_tokens": 34542541.0,
      "step": 15760
    },
    {
      "entropy": 5.6301970958709715,
      "epoch": 1.564608971814212,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047611114295665697,
      "loss": 5.1833,
      "mean_token_accuracy": 0.17069462090730667,
      "num_tokens": 34553056.0,
      "step": 15765
    },
    {
      "entropy": 5.6599273681640625,
      "epoch": 1.5651052004763795,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047609525257755967,
      "loss": 5.1743,
      "mean_token_accuracy": 0.16856639981269836,
      "num_tokens": 34563344.0,
      "step": 15770
    },
    {
      "entropy": 5.616159963607788,
      "epoch": 1.565601429138547,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004760793572117333,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1805104658007622,
      "num_tokens": 34574320.0,
      "step": 15775
    },
    {
      "entropy": 5.5500490188598635,
      "epoch": 1.5660976578007144,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004760634568595721,
      "loss": 5.127,
      "mean_token_accuracy": 0.16711512506008147,
      "num_tokens": 34584728.0,
      "step": 15780
    },
    {
      "entropy": 5.6234029769897464,
      "epoch": 1.5665938864628821,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004760475515214704,
      "loss": 5.2372,
      "mean_token_accuracy": 0.17107450515031813,
      "num_tokens": 34596395.0,
      "step": 15785
    },
    {
      "entropy": 5.5590709209442135,
      "epoch": 1.5670901151250496,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004760316411978228,
      "loss": 5.1346,
      "mean_token_accuracy": 0.17077884078025818,
      "num_tokens": 34607950.0,
      "step": 15790
    },
    {
      "entropy": 5.573279571533203,
      "epoch": 1.5675863437872173,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047601572588902363,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17734950482845308,
      "num_tokens": 34617640.0,
      "step": 15795
    },
    {
      "entropy": 5.587791872024536,
      "epoch": 1.5680825724493848,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047599980559546757,
      "loss": 5.1829,
      "mean_token_accuracy": 0.16359065771102904,
      "num_tokens": 34629475.0,
      "step": 15800
    },
    {
      "entropy": 5.586582517623901,
      "epoch": 1.5685788011115522,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004759838803175495,
      "loss": 5.1144,
      "mean_token_accuracy": 0.1697483777999878,
      "num_tokens": 34641423.0,
      "step": 15805
    },
    {
      "entropy": 5.708580160140992,
      "epoch": 1.5690750297737197,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047596795005566423,
      "loss": 5.1514,
      "mean_token_accuracy": 0.17095966041088104,
      "num_tokens": 34651894.0,
      "step": 15810
    },
    {
      "entropy": 5.615035629272461,
      "epoch": 1.5695712584358872,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047595201481020685,
      "loss": 5.1733,
      "mean_token_accuracy": 0.16662030220031737,
      "num_tokens": 34664177.0,
      "step": 15815
    },
    {
      "entropy": 5.550269889831543,
      "epoch": 1.5700674870980547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047593607458157257,
      "loss": 5.139,
      "mean_token_accuracy": 0.17155104726552964,
      "num_tokens": 34674850.0,
      "step": 15820
    },
    {
      "entropy": 5.596954727172852,
      "epoch": 1.5705637157602224,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047592012937015647,
      "loss": 5.1699,
      "mean_token_accuracy": 0.1756455898284912,
      "num_tokens": 34685719.0,
      "step": 15825
    },
    {
      "entropy": 5.651635646820068,
      "epoch": 1.5710599444223898,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047590417917635423,
      "loss": 5.1654,
      "mean_token_accuracy": 0.16552997529506683,
      "num_tokens": 34697261.0,
      "step": 15830
    },
    {
      "entropy": 5.6315412521362305,
      "epoch": 1.5715561730845575,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004758882240005613,
      "loss": 5.1596,
      "mean_token_accuracy": 0.17224330455064774,
      "num_tokens": 34708196.0,
      "step": 15835
    },
    {
      "entropy": 5.6407839298248295,
      "epoch": 1.572052401746725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047587226384317323,
      "loss": 5.1807,
      "mean_token_accuracy": 0.167033351957798,
      "num_tokens": 34718855.0,
      "step": 15840
    },
    {
      "entropy": 5.592995595932007,
      "epoch": 1.5725486304088925,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00047585629870458584,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17824140787124634,
      "num_tokens": 34731137.0,
      "step": 15845
    },
    {
      "entropy": 5.621906423568726,
      "epoch": 1.57304485907106,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004758403285851951,
      "loss": 5.0725,
      "mean_token_accuracy": 0.17906162589788438,
      "num_tokens": 34741628.0,
      "step": 15850
    },
    {
      "entropy": 5.5560705184936525,
      "epoch": 1.5735410877332274,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00047582435348539695,
      "loss": 5.1509,
      "mean_token_accuracy": 0.17379579097032546,
      "num_tokens": 34752984.0,
      "step": 15855
    },
    {
      "entropy": 5.591576719284058,
      "epoch": 1.574037316395395,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004758083734055875,
      "loss": 5.213,
      "mean_token_accuracy": 0.16995763927698135,
      "num_tokens": 34764363.0,
      "step": 15860
    },
    {
      "entropy": 5.6082422733306885,
      "epoch": 1.5745335450575624,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004757923883461632,
      "loss": 5.1491,
      "mean_token_accuracy": 0.16764347702264787,
      "num_tokens": 34776289.0,
      "step": 15865
    },
    {
      "entropy": 5.652585458755493,
      "epoch": 1.57502977371973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047577639830752034,
      "loss": 5.1371,
      "mean_token_accuracy": 0.16802215129137038,
      "num_tokens": 34786760.0,
      "step": 15870
    },
    {
      "entropy": 5.6062681674957275,
      "epoch": 1.5755260023818976,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047576040329005536,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17375023066997528,
      "num_tokens": 34797737.0,
      "step": 15875
    },
    {
      "entropy": 5.572108364105224,
      "epoch": 1.5760222310440652,
      "grad_norm": 1.0,
      "learning_rate": 0.000475744403294165,
      "loss": 5.1606,
      "mean_token_accuracy": 0.1765473112463951,
      "num_tokens": 34809650.0,
      "step": 15880
    },
    {
      "entropy": 5.69287691116333,
      "epoch": 1.5765184597062327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047572839832024606,
      "loss": 5.179,
      "mean_token_accuracy": 0.16144877672195435,
      "num_tokens": 34820171.0,
      "step": 15885
    },
    {
      "entropy": 5.576609182357788,
      "epoch": 1.5770146883684002,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047571238836869537,
      "loss": 5.0427,
      "mean_token_accuracy": 0.17854354679584503,
      "num_tokens": 34831286.0,
      "step": 15890
    },
    {
      "entropy": 5.489605236053467,
      "epoch": 1.5775109170305677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004756963734399099,
      "loss": 5.0603,
      "mean_token_accuracy": 0.17938444167375564,
      "num_tokens": 34841767.0,
      "step": 15895
    },
    {
      "entropy": 5.555397605895996,
      "epoch": 1.5780071456927351,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047568035353428683,
      "loss": 5.1667,
      "mean_token_accuracy": 0.1709470883011818,
      "num_tokens": 34852644.0,
      "step": 15900
    },
    {
      "entropy": 5.646950197219849,
      "epoch": 1.5785033743549026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047566432865222343,
      "loss": 5.1654,
      "mean_token_accuracy": 0.1748142123222351,
      "num_tokens": 34863605.0,
      "step": 15905
    },
    {
      "entropy": 5.571821451187134,
      "epoch": 1.5789996030170703,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047564829879411706,
      "loss": 5.1082,
      "mean_token_accuracy": 0.17918685972690582,
      "num_tokens": 34874594.0,
      "step": 15910
    },
    {
      "entropy": 5.618953704833984,
      "epoch": 1.5794958316792378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004756322639603653,
      "loss": 5.207,
      "mean_token_accuracy": 0.16953258961439133,
      "num_tokens": 34884627.0,
      "step": 15915
    },
    {
      "entropy": 5.6971354484558105,
      "epoch": 1.5799920603414053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047561622415136565,
      "loss": 5.1597,
      "mean_token_accuracy": 0.1786976858973503,
      "num_tokens": 34895679.0,
      "step": 15920
    },
    {
      "entropy": 5.581293535232544,
      "epoch": 1.580488289003573,
      "grad_norm": 1.0,
      "learning_rate": 0.00047560017936751597,
      "loss": 5.1095,
      "mean_token_accuracy": 0.17618269920349122,
      "num_tokens": 34906811.0,
      "step": 15925
    },
    {
      "entropy": 5.604450845718384,
      "epoch": 1.5809845176657404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755841296092141,
      "loss": 5.1389,
      "mean_token_accuracy": 0.17078711092472076,
      "num_tokens": 34918655.0,
      "step": 15930
    },
    {
      "entropy": 5.578640031814575,
      "epoch": 1.581480746327908,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047556807487685797,
      "loss": 5.1288,
      "mean_token_accuracy": 0.168772329390049,
      "num_tokens": 34931034.0,
      "step": 15935
    },
    {
      "entropy": 5.608658504486084,
      "epoch": 1.5819769749900754,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004755520151708458,
      "loss": 5.0752,
      "mean_token_accuracy": 0.17211978286504745,
      "num_tokens": 34939707.0,
      "step": 15940
    },
    {
      "entropy": 5.575910663604736,
      "epoch": 1.5824732036522429,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755359504915758,
      "loss": 5.1797,
      "mean_token_accuracy": 0.1668235182762146,
      "num_tokens": 34949237.0,
      "step": 15945
    },
    {
      "entropy": 5.634340524673462,
      "epoch": 1.5829694323144103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004755198808394464,
      "loss": 5.2105,
      "mean_token_accuracy": 0.17056822925806045,
      "num_tokens": 34960380.0,
      "step": 15950
    },
    {
      "entropy": 5.637466382980347,
      "epoch": 1.583465660976578,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475503806214856,
      "loss": 5.1622,
      "mean_token_accuracy": 0.17007072120904923,
      "num_tokens": 34971699.0,
      "step": 15955
    },
    {
      "entropy": 5.598207950592041,
      "epoch": 1.5839618896387455,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004754877266182033,
      "loss": 5.1305,
      "mean_token_accuracy": 0.1774409905076027,
      "num_tokens": 34982333.0,
      "step": 15960
    },
    {
      "entropy": 5.598082065582275,
      "epoch": 1.5844581183009132,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000475471642049887,
      "loss": 5.1692,
      "mean_token_accuracy": 0.17295277565717698,
      "num_tokens": 34993277.0,
      "step": 15965
    },
    {
      "entropy": 5.590831661224366,
      "epoch": 1.5849543469630807,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004754555525103058,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17658014446496964,
      "num_tokens": 35004252.0,
      "step": 15970
    },
    {
      "entropy": 5.538004398345947,
      "epoch": 1.5854505756252482,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000475439457999859,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17992551922798156,
      "num_tokens": 35016974.0,
      "step": 15975
    },
    {
      "entropy": 5.626382684707641,
      "epoch": 1.5859468042874156,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047542335851894553,
      "loss": 5.1852,
      "mean_token_accuracy": 0.1697611927986145,
      "num_tokens": 35029017.0,
      "step": 15980
    },
    {
      "entropy": 5.659984493255616,
      "epoch": 1.586443032949583,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047540725406796454,
      "loss": 5.198,
      "mean_token_accuracy": 0.16977013945579528,
      "num_tokens": 35040239.0,
      "step": 15985
    },
    {
      "entropy": 5.6302196979522705,
      "epoch": 1.5869392616117506,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004753911446473156,
      "loss": 5.1263,
      "mean_token_accuracy": 0.1689956232905388,
      "num_tokens": 35051426.0,
      "step": 15990
    },
    {
      "entropy": 5.51170539855957,
      "epoch": 1.587435490273918,
      "grad_norm": 1.015625,
      "learning_rate": 0.000475375030257398,
      "loss": 5.093,
      "mean_token_accuracy": 0.1704554006457329,
      "num_tokens": 35061777.0,
      "step": 15995
    },
    {
      "entropy": 5.5646562576293945,
      "epoch": 1.5879317189360858,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004753589108986114,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1871209591627121,
      "num_tokens": 35071822.0,
      "step": 16000
    },
    {
      "entropy": 5.617754507064819,
      "epoch": 1.5884279475982532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047534278657135565,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17154599875211715,
      "num_tokens": 35081900.0,
      "step": 16005
    },
    {
      "entropy": 5.637001609802246,
      "epoch": 1.588924176260421,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004753266572760304,
      "loss": 5.1932,
      "mean_token_accuracy": 0.1671551376581192,
      "num_tokens": 35092866.0,
      "step": 16010
    },
    {
      "entropy": 5.761134004592895,
      "epoch": 1.5894204049225884,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047531052301303577,
      "loss": 5.2868,
      "mean_token_accuracy": 0.17267492562532424,
      "num_tokens": 35103425.0,
      "step": 16015
    },
    {
      "entropy": 5.541793537139893,
      "epoch": 1.5899166335847559,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004752943837827718,
      "loss": 5.0603,
      "mean_token_accuracy": 0.17095179557800294,
      "num_tokens": 35114620.0,
      "step": 16020
    },
    {
      "entropy": 5.646689367294312,
      "epoch": 1.5904128622469234,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004752782395856387,
      "loss": 5.1654,
      "mean_token_accuracy": 0.16901233941316604,
      "num_tokens": 35125199.0,
      "step": 16025
    },
    {
      "entropy": 5.582147121429443,
      "epoch": 1.5909090909090908,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004752620904220368,
      "loss": 5.1381,
      "mean_token_accuracy": 0.1741214230656624,
      "num_tokens": 35135892.0,
      "step": 16030
    },
    {
      "entropy": 5.610662126541138,
      "epoch": 1.5914053195712583,
      "grad_norm": 1.0,
      "learning_rate": 0.0004752459362923666,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17109443694353105,
      "num_tokens": 35146186.0,
      "step": 16035
    },
    {
      "entropy": 5.579604625701904,
      "epoch": 1.591901548233426,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004752297771970287,
      "loss": 5.1961,
      "mean_token_accuracy": 0.17025927901268006,
      "num_tokens": 35157261.0,
      "step": 16040
    },
    {
      "entropy": 5.6247802734375,
      "epoch": 1.5923977768955935,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047521361313642373,
      "loss": 5.1496,
      "mean_token_accuracy": 0.1663690149784088,
      "num_tokens": 35168508.0,
      "step": 16045
    },
    {
      "entropy": 5.5624223232269285,
      "epoch": 1.5928940055577612,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004751974441109526,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17138774394989015,
      "num_tokens": 35179444.0,
      "step": 16050
    },
    {
      "entropy": 5.544235467910767,
      "epoch": 1.5933902342199286,
      "grad_norm": 1.0,
      "learning_rate": 0.00047518127012101624,
      "loss": 5.1075,
      "mean_token_accuracy": 0.17422276139259338,
      "num_tokens": 35189517.0,
      "step": 16055
    },
    {
      "entropy": 5.580736064910889,
      "epoch": 1.5938864628820961,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004751650911670157,
      "loss": 5.1443,
      "mean_token_accuracy": 0.16847994476556777,
      "num_tokens": 35200191.0,
      "step": 16060
    },
    {
      "entropy": 5.613438463211059,
      "epoch": 1.5943826915442636,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047514890724935236,
      "loss": 5.1567,
      "mean_token_accuracy": 0.16769902855157853,
      "num_tokens": 35211260.0,
      "step": 16065
    },
    {
      "entropy": 5.51862940788269,
      "epoch": 1.594878920206431,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004751327183684272,
      "loss": 5.1398,
      "mean_token_accuracy": 0.1683967664837837,
      "num_tokens": 35222408.0,
      "step": 16070
    },
    {
      "entropy": 5.631817102432251,
      "epoch": 1.5953751488685985,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047511652452464196,
      "loss": 5.1968,
      "mean_token_accuracy": 0.16965654790401458,
      "num_tokens": 35232760.0,
      "step": 16075
    },
    {
      "entropy": 5.630973196029663,
      "epoch": 1.595871377530766,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047510032571839816,
      "loss": 5.1503,
      "mean_token_accuracy": 0.17005118429660798,
      "num_tokens": 35243849.0,
      "step": 16080
    },
    {
      "entropy": 5.648681354522705,
      "epoch": 1.5963676061929337,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004750841219500975,
      "loss": 5.1692,
      "mean_token_accuracy": 0.1689106360077858,
      "num_tokens": 35254962.0,
      "step": 16085
    },
    {
      "entropy": 5.6104662895202635,
      "epoch": 1.5968638348551012,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004750679132201416,
      "loss": 5.2282,
      "mean_token_accuracy": 0.16736824810504913,
      "num_tokens": 35265895.0,
      "step": 16090
    },
    {
      "entropy": 5.700689125061035,
      "epoch": 1.597360063517269,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047505169952893276,
      "loss": 5.2316,
      "mean_token_accuracy": 0.16893060505390167,
      "num_tokens": 35277543.0,
      "step": 16095
    },
    {
      "entropy": 5.606198453903199,
      "epoch": 1.5978562921794364,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047503548087687267,
      "loss": 5.1247,
      "mean_token_accuracy": 0.1757278248667717,
      "num_tokens": 35289130.0,
      "step": 16100
    },
    {
      "entropy": 5.564643812179566,
      "epoch": 1.5983525208416038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047501925726436376,
      "loss": 5.0671,
      "mean_token_accuracy": 0.16926935613155364,
      "num_tokens": 35298414.0,
      "step": 16105
    },
    {
      "entropy": 5.563416862487793,
      "epoch": 1.5988487495037713,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004750030286918082,
      "loss": 5.1406,
      "mean_token_accuracy": 0.17215073257684707,
      "num_tokens": 35310629.0,
      "step": 16110
    },
    {
      "entropy": 5.636842584609985,
      "epoch": 1.5993449781659388,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004749867951596086,
      "loss": 5.1701,
      "mean_token_accuracy": 0.17372881770133972,
      "num_tokens": 35321810.0,
      "step": 16115
    },
    {
      "entropy": 5.6139716625213625,
      "epoch": 1.5998412068281063,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047497055666816736,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17949010580778121,
      "num_tokens": 35333517.0,
      "step": 16120
    },
    {
      "entropy": 5.551135540008545,
      "epoch": 1.600337435490274,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004749543132178873,
      "loss": 5.0421,
      "mean_token_accuracy": 0.176265712082386,
      "num_tokens": 35344170.0,
      "step": 16125
    },
    {
      "entropy": 5.622438764572143,
      "epoch": 1.6008336641524414,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000474938064809171,
      "loss": 5.207,
      "mean_token_accuracy": 0.1682305157184601,
      "num_tokens": 35354963.0,
      "step": 16130
    },
    {
      "entropy": 5.628072452545166,
      "epoch": 1.6013298928146091,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004749218114424216,
      "loss": 5.1749,
      "mean_token_accuracy": 0.17679030150175096,
      "num_tokens": 35365716.0,
      "step": 16135
    },
    {
      "entropy": 5.557168292999267,
      "epoch": 1.6018261214767766,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047490555311804206,
      "loss": 5.1098,
      "mean_token_accuracy": 0.1740991771221161,
      "num_tokens": 35376113.0,
      "step": 16140
    },
    {
      "entropy": 5.53159465789795,
      "epoch": 1.602322350138944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047488928983643565,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17015409618616104,
      "num_tokens": 35386178.0,
      "step": 16145
    },
    {
      "entropy": 5.533686113357544,
      "epoch": 1.6028185788011116,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047487302159800545,
      "loss": 5.008,
      "mean_token_accuracy": 0.17713668048381806,
      "num_tokens": 35396874.0,
      "step": 16150
    },
    {
      "entropy": 5.600658273696899,
      "epoch": 1.603314807463279,
      "grad_norm": 0.890625,
      "learning_rate": 0.000474856748403155,
      "loss": 5.2203,
      "mean_token_accuracy": 0.16721452325582503,
      "num_tokens": 35409405.0,
      "step": 16155
    },
    {
      "entropy": 5.639662075042724,
      "epoch": 1.6038110361254465,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004748404702522879,
      "loss": 5.1607,
      "mean_token_accuracy": 0.17237268388271332,
      "num_tokens": 35421449.0,
      "step": 16160
    },
    {
      "entropy": 5.596003246307373,
      "epoch": 1.604307264787614,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047482418714580774,
      "loss": 5.2161,
      "mean_token_accuracy": 0.17112949788570403,
      "num_tokens": 35434612.0,
      "step": 16165
    },
    {
      "entropy": 5.58344464302063,
      "epoch": 1.6048034934497817,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004748078990841183,
      "loss": 5.101,
      "mean_token_accuracy": 0.17633241713047026,
      "num_tokens": 35446016.0,
      "step": 16170
    },
    {
      "entropy": 5.643645286560059,
      "epoch": 1.6052997221119492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004747916060676235,
      "loss": 5.1506,
      "mean_token_accuracy": 0.16969554871320724,
      "num_tokens": 35457405.0,
      "step": 16175
    },
    {
      "entropy": 5.560308837890625,
      "epoch": 1.6057959507741169,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047477530809672747,
      "loss": 5.2065,
      "mean_token_accuracy": 0.16585485637187958,
      "num_tokens": 35468361.0,
      "step": 16180
    },
    {
      "entropy": 5.571908617019654,
      "epoch": 1.6062921794362843,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004747590051718342,
      "loss": 5.088,
      "mean_token_accuracy": 0.17132386714220046,
      "num_tokens": 35480092.0,
      "step": 16185
    },
    {
      "entropy": 5.628490209579468,
      "epoch": 1.6067884080984518,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047474269729334806,
      "loss": 5.2177,
      "mean_token_accuracy": 0.16508890241384505,
      "num_tokens": 35492131.0,
      "step": 16190
    },
    {
      "entropy": 5.62976975440979,
      "epoch": 1.6072846367606193,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047472638446167346,
      "loss": 5.1552,
      "mean_token_accuracy": 0.17143058627843857,
      "num_tokens": 35503600.0,
      "step": 16195
    },
    {
      "entropy": 5.648710298538208,
      "epoch": 1.6077808654227868,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047471006667721484,
      "loss": 5.2168,
      "mean_token_accuracy": 0.16859940588474273,
      "num_tokens": 35514716.0,
      "step": 16200
    },
    {
      "entropy": 5.614390516281128,
      "epoch": 1.6082770940849542,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000474693743940377,
      "loss": 5.1266,
      "mean_token_accuracy": 0.17894127815961838,
      "num_tokens": 35525688.0,
      "step": 16205
    },
    {
      "entropy": 5.600659370422363,
      "epoch": 1.608773322747122,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047467741625156453,
      "loss": 5.137,
      "mean_token_accuracy": 0.1746787831187248,
      "num_tokens": 35536391.0,
      "step": 16210
    },
    {
      "entropy": 5.564913129806518,
      "epoch": 1.6092695514092894,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047466108361118235,
      "loss": 5.0799,
      "mean_token_accuracy": 0.1779962033033371,
      "num_tokens": 35547016.0,
      "step": 16215
    },
    {
      "entropy": 5.600070381164551,
      "epoch": 1.6097657800714569,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047464474601963564,
      "loss": 5.1501,
      "mean_token_accuracy": 0.1721704989671707,
      "num_tokens": 35558402.0,
      "step": 16220
    },
    {
      "entropy": 5.610692930221558,
      "epoch": 1.6102620087336246,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004746284034773294,
      "loss": 5.1494,
      "mean_token_accuracy": 0.16781592220067978,
      "num_tokens": 35569779.0,
      "step": 16225
    },
    {
      "entropy": 5.609346580505371,
      "epoch": 1.610758237395792,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004746120559846689,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17149755954742432,
      "num_tokens": 35580044.0,
      "step": 16230
    },
    {
      "entropy": 5.637996339797974,
      "epoch": 1.6112544660579595,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047459570354205954,
      "loss": 5.2436,
      "mean_token_accuracy": 0.16991163790225983,
      "num_tokens": 35591216.0,
      "step": 16235
    },
    {
      "entropy": 5.552469205856323,
      "epoch": 1.611750694720127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047457934614990684,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17355382442474365,
      "num_tokens": 35602376.0,
      "step": 16240
    },
    {
      "entropy": 5.594023370742798,
      "epoch": 1.6122469233822945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004745629838086164,
      "loss": 5.0361,
      "mean_token_accuracy": 0.17899127602577208,
      "num_tokens": 35612184.0,
      "step": 16245
    },
    {
      "entropy": 5.599906873703003,
      "epoch": 1.612743152044462,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047454661651859394,
      "loss": 5.2094,
      "mean_token_accuracy": 0.17085298299789428,
      "num_tokens": 35624864.0,
      "step": 16250
    },
    {
      "entropy": 5.590623807907105,
      "epoch": 1.6132393807066296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004745302442802454,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17309827208518982,
      "num_tokens": 35636492.0,
      "step": 16255
    },
    {
      "entropy": 5.568262815475464,
      "epoch": 1.6137356093687971,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004745138670939768,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17463687658309937,
      "num_tokens": 35646831.0,
      "step": 16260
    },
    {
      "entropy": 5.675525236129761,
      "epoch": 1.6142318380309648,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004744974849601941,
      "loss": 5.1887,
      "mean_token_accuracy": 0.17240960597991944,
      "num_tokens": 35657703.0,
      "step": 16265
    },
    {
      "entropy": 5.522848892211914,
      "epoch": 1.6147280666931323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004744810978793038,
      "loss": 5.0041,
      "mean_token_accuracy": 0.175139719247818,
      "num_tokens": 35668334.0,
      "step": 16270
    },
    {
      "entropy": 5.57087812423706,
      "epoch": 1.6152242953552998,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047446470585171206,
      "loss": 5.1755,
      "mean_token_accuracy": 0.17126649916172026,
      "num_tokens": 35679796.0,
      "step": 16275
    },
    {
      "entropy": 5.546227073669433,
      "epoch": 1.6157205240174672,
      "grad_norm": 1.0,
      "learning_rate": 0.00047444830887782537,
      "loss": 5.0621,
      "mean_token_accuracy": 0.173496749997139,
      "num_tokens": 35690825.0,
      "step": 16280
    },
    {
      "entropy": 5.679032182693481,
      "epoch": 1.6162167526796347,
      "grad_norm": 1.0,
      "learning_rate": 0.0004744319069580504,
      "loss": 5.1638,
      "mean_token_accuracy": 0.16990370750427247,
      "num_tokens": 35701824.0,
      "step": 16285
    },
    {
      "entropy": 5.66945447921753,
      "epoch": 1.6167129813418022,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047441550009279395,
      "loss": 5.1056,
      "mean_token_accuracy": 0.1781490594148636,
      "num_tokens": 35713250.0,
      "step": 16290
    },
    {
      "entropy": 5.571934986114502,
      "epoch": 1.6172092100039697,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004743990882824627,
      "loss": 5.1548,
      "mean_token_accuracy": 0.17356335371732712,
      "num_tokens": 35725049.0,
      "step": 16295
    },
    {
      "entropy": 5.5343903541564945,
      "epoch": 1.6177054386661374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047438267152746376,
      "loss": 5.1125,
      "mean_token_accuracy": 0.17064690291881562,
      "num_tokens": 35734676.0,
      "step": 16300
    },
    {
      "entropy": 5.638377523422241,
      "epoch": 1.6182016673283048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047436624982820423,
      "loss": 5.1105,
      "mean_token_accuracy": 0.16956356018781663,
      "num_tokens": 35743953.0,
      "step": 16305
    },
    {
      "entropy": 5.583350944519043,
      "epoch": 1.6186978959904725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743498231850913,
      "loss": 5.0482,
      "mean_token_accuracy": 0.1791108250617981,
      "num_tokens": 35753149.0,
      "step": 16310
    },
    {
      "entropy": 5.549193954467773,
      "epoch": 1.61919412465264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004743333915985323,
      "loss": 5.1287,
      "mean_token_accuracy": 0.168913771212101,
      "num_tokens": 35763783.0,
      "step": 16315
    },
    {
      "entropy": 5.549267387390136,
      "epoch": 1.6196903533148075,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047431695506893476,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18040068000555037,
      "num_tokens": 35774663.0,
      "step": 16320
    },
    {
      "entropy": 5.629247617721558,
      "epoch": 1.620186581976975,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047430051359670616,
      "loss": 5.2172,
      "mean_token_accuracy": 0.16416287869215013,
      "num_tokens": 35786990.0,
      "step": 16325
    },
    {
      "entropy": 5.690722513198852,
      "epoch": 1.6206828106391424,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004742840671822542,
      "loss": 5.2191,
      "mean_token_accuracy": 0.1734220042824745,
      "num_tokens": 35798818.0,
      "step": 16330
    },
    {
      "entropy": 5.641177415847778,
      "epoch": 1.62117903930131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000474267615825987,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16891805082559586,
      "num_tokens": 35808813.0,
      "step": 16335
    },
    {
      "entropy": 5.58909158706665,
      "epoch": 1.6216752679634776,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047425115952831213,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1726187363266945,
      "num_tokens": 35818664.0,
      "step": 16340
    },
    {
      "entropy": 5.534975147247314,
      "epoch": 1.622171496625645,
      "grad_norm": 1.0,
      "learning_rate": 0.000474234698289638,
      "loss": 5.0762,
      "mean_token_accuracy": 0.18005962967872619,
      "num_tokens": 35828975.0,
      "step": 16345
    },
    {
      "entropy": 5.539949083328247,
      "epoch": 1.6226677252878128,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004742182321103725,
      "loss": 5.1088,
      "mean_token_accuracy": 0.1736476242542267,
      "num_tokens": 35839437.0,
      "step": 16350
    },
    {
      "entropy": 5.578946256637574,
      "epoch": 1.6231639539499803,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004742017609909242,
      "loss": 5.0706,
      "mean_token_accuracy": 0.17905042916536332,
      "num_tokens": 35850413.0,
      "step": 16355
    },
    {
      "entropy": 5.654530477523804,
      "epoch": 1.6236601826121477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047418528493170157,
      "loss": 5.1402,
      "mean_token_accuracy": 0.16812205761671067,
      "num_tokens": 35862745.0,
      "step": 16360
    },
    {
      "entropy": 5.653662014007568,
      "epoch": 1.6241564112743152,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004741688039331129,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1710583448410034,
      "num_tokens": 35873934.0,
      "step": 16365
    },
    {
      "entropy": 5.556177473068237,
      "epoch": 1.6246526399364827,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047415231799556725,
      "loss": 5.0461,
      "mean_token_accuracy": 0.18180738538503646,
      "num_tokens": 35884279.0,
      "step": 16370
    },
    {
      "entropy": 5.55698595046997,
      "epoch": 1.6251488685986502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047413582711947305,
      "loss": 5.1009,
      "mean_token_accuracy": 0.17988390028476714,
      "num_tokens": 35894144.0,
      "step": 16375
    },
    {
      "entropy": 5.560732460021972,
      "epoch": 1.6256450972608176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047411933130523954,
      "loss": 5.1017,
      "mean_token_accuracy": 0.17227184921503066,
      "num_tokens": 35903641.0,
      "step": 16380
    },
    {
      "entropy": 5.546038818359375,
      "epoch": 1.6261413259229853,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004741028305532757,
      "loss": 5.0684,
      "mean_token_accuracy": 0.17541877925395966,
      "num_tokens": 35913634.0,
      "step": 16385
    },
    {
      "entropy": 5.5720408916473385,
      "epoch": 1.6266375545851528,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004740863248639907,
      "loss": 5.1114,
      "mean_token_accuracy": 0.17902370542287827,
      "num_tokens": 35923106.0,
      "step": 16390
    },
    {
      "entropy": 5.59875864982605,
      "epoch": 1.6271337832473205,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047406981423779376,
      "loss": 5.1095,
      "mean_token_accuracy": 0.17180464416742325,
      "num_tokens": 35933709.0,
      "step": 16395
    },
    {
      "entropy": 5.578087043762207,
      "epoch": 1.627630011909488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047405329867509447,
      "loss": 5.1306,
      "mean_token_accuracy": 0.16590722948312758,
      "num_tokens": 35944734.0,
      "step": 16400
    },
    {
      "entropy": 5.579199361801147,
      "epoch": 1.6281262405716554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047403677817630226,
      "loss": 5.0913,
      "mean_token_accuracy": 0.1737969696521759,
      "num_tokens": 35956513.0,
      "step": 16405
    },
    {
      "entropy": 5.547866153717041,
      "epoch": 1.628622469233823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004740202527418269,
      "loss": 5.108,
      "mean_token_accuracy": 0.17377176135778427,
      "num_tokens": 35967096.0,
      "step": 16410
    },
    {
      "entropy": 5.603216695785522,
      "epoch": 1.6291186978959904,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000474003722372078,
      "loss": 5.1803,
      "mean_token_accuracy": 0.16923546344041823,
      "num_tokens": 35979010.0,
      "step": 16415
    },
    {
      "entropy": 5.545507907867432,
      "epoch": 1.6296149265581579,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047398718706746556,
      "loss": 5.0408,
      "mean_token_accuracy": 0.17602762579917908,
      "num_tokens": 35989253.0,
      "step": 16420
    },
    {
      "entropy": 5.5565286636352536,
      "epoch": 1.6301111552203256,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004739706468283998,
      "loss": 5.0619,
      "mean_token_accuracy": 0.17655352205038072,
      "num_tokens": 35999659.0,
      "step": 16425
    },
    {
      "entropy": 5.5558753490448,
      "epoch": 1.630607383882493,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047395410165529075,
      "loss": 5.1238,
      "mean_token_accuracy": 0.1683488056063652,
      "num_tokens": 36010295.0,
      "step": 16430
    },
    {
      "entropy": 5.597892332077026,
      "epoch": 1.6311036125446607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047393755154854855,
      "loss": 5.138,
      "mean_token_accuracy": 0.17247526496648788,
      "num_tokens": 36021575.0,
      "step": 16435
    },
    {
      "entropy": 5.625938749313354,
      "epoch": 1.6315998412068282,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004739209965085839,
      "loss": 5.1341,
      "mean_token_accuracy": 0.1665845736861229,
      "num_tokens": 36032764.0,
      "step": 16440
    },
    {
      "entropy": 5.586221647262573,
      "epoch": 1.6320960698689957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000473904436535807,
      "loss": 5.0934,
      "mean_token_accuracy": 0.18002111315727234,
      "num_tokens": 36041864.0,
      "step": 16445
    },
    {
      "entropy": 5.519574069976807,
      "epoch": 1.6325922985311632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004738878716306288,
      "loss": 5.0814,
      "mean_token_accuracy": 0.1742129921913147,
      "num_tokens": 36051321.0,
      "step": 16450
    },
    {
      "entropy": 5.5300546169281,
      "epoch": 1.6330885271933306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047387130179345996,
      "loss": 5.0527,
      "mean_token_accuracy": 0.1750680163502693,
      "num_tokens": 36061448.0,
      "step": 16455
    },
    {
      "entropy": 5.607499027252198,
      "epoch": 1.6335847558554981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047385472702471133,
      "loss": 5.1817,
      "mean_token_accuracy": 0.17706872522830963,
      "num_tokens": 36071832.0,
      "step": 16460
    },
    {
      "entropy": 5.558157825469971,
      "epoch": 1.6340809845176656,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000473838147324794,
      "loss": 5.1078,
      "mean_token_accuracy": 0.17032204568386078,
      "num_tokens": 36082103.0,
      "step": 16465
    },
    {
      "entropy": 5.603205633163452,
      "epoch": 1.6345772131798333,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000473821562694119,
      "loss": 5.137,
      "mean_token_accuracy": 0.1706890344619751,
      "num_tokens": 36092063.0,
      "step": 16470
    },
    {
      "entropy": 5.609879684448242,
      "epoch": 1.6350734418420008,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004738049731330977,
      "loss": 5.122,
      "mean_token_accuracy": 0.16356065720319748,
      "num_tokens": 36102573.0,
      "step": 16475
    },
    {
      "entropy": 5.655090093612671,
      "epoch": 1.6355696705041685,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004737883786421414,
      "loss": 5.1064,
      "mean_token_accuracy": 0.17099539488554,
      "num_tokens": 36112677.0,
      "step": 16480
    },
    {
      "entropy": 5.608657789230347,
      "epoch": 1.636065899166336,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047377177922166175,
      "loss": 5.1809,
      "mean_token_accuracy": 0.16671934574842454,
      "num_tokens": 36123337.0,
      "step": 16485
    },
    {
      "entropy": 5.626515436172485,
      "epoch": 1.6365621278285034,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047375517487207024,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1715997576713562,
      "num_tokens": 36134503.0,
      "step": 16490
    },
    {
      "entropy": 5.682374429702759,
      "epoch": 1.6370583564906709,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047373856559377876,
      "loss": 5.2215,
      "mean_token_accuracy": 0.17025102823972701,
      "num_tokens": 36145766.0,
      "step": 16495
    },
    {
      "entropy": 5.61710557937622,
      "epoch": 1.6375545851528384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004737219513871989,
      "loss": 5.1487,
      "mean_token_accuracy": 0.17260605096817017,
      "num_tokens": 36156584.0,
      "step": 16500
    },
    {
      "entropy": 5.596297264099121,
      "epoch": 1.6380508138150058,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047370533225274305,
      "loss": 5.1393,
      "mean_token_accuracy": 0.174840547144413,
      "num_tokens": 36166999.0,
      "step": 16505
    },
    {
      "entropy": 5.595086097717285,
      "epoch": 1.6385470424771735,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004736887081908231,
      "loss": 5.113,
      "mean_token_accuracy": 0.17617233097553253,
      "num_tokens": 36178724.0,
      "step": 16510
    },
    {
      "entropy": 5.671789741516113,
      "epoch": 1.639043271139341,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004736720792018513,
      "loss": 5.2062,
      "mean_token_accuracy": 0.1702543392777443,
      "num_tokens": 36188912.0,
      "step": 16515
    },
    {
      "entropy": 5.587574672698975,
      "epoch": 1.6395394998015085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004736554452862399,
      "loss": 5.102,
      "mean_token_accuracy": 0.1721709594130516,
      "num_tokens": 36198525.0,
      "step": 16520
    },
    {
      "entropy": 5.55933403968811,
      "epoch": 1.6400357284636762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047363880644440173,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17557340264320373,
      "num_tokens": 36209384.0,
      "step": 16525
    },
    {
      "entropy": 5.634116220474243,
      "epoch": 1.6405319571258437,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047362216267674906,
      "loss": 5.1678,
      "mean_token_accuracy": 0.1715904325246811,
      "num_tokens": 36220597.0,
      "step": 16530
    },
    {
      "entropy": 5.63750991821289,
      "epoch": 1.6410281857880111,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004736055139836948,
      "loss": 5.2067,
      "mean_token_accuracy": 0.17045432329177856,
      "num_tokens": 36230317.0,
      "step": 16535
    },
    {
      "entropy": 5.699209308624267,
      "epoch": 1.6415244144501786,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047358886036565176,
      "loss": 5.2978,
      "mean_token_accuracy": 0.16692895889282228,
      "num_tokens": 36241645.0,
      "step": 16540
    },
    {
      "entropy": 5.653608083724976,
      "epoch": 1.642020643112346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047357220182303285,
      "loss": 5.2155,
      "mean_token_accuracy": 0.1680998310446739,
      "num_tokens": 36253098.0,
      "step": 16545
    },
    {
      "entropy": 5.5630144596099855,
      "epoch": 1.6425168717745136,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004735555383562513,
      "loss": 5.1099,
      "mean_token_accuracy": 0.17187026739120484,
      "num_tokens": 36264417.0,
      "step": 16550
    },
    {
      "entropy": 5.668043947219848,
      "epoch": 1.6430131004366813,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00047353886996572014,
      "loss": 5.2047,
      "mean_token_accuracy": 0.16684735268354417,
      "num_tokens": 36276589.0,
      "step": 16555
    },
    {
      "entropy": 5.590042924880981,
      "epoch": 1.6435093290988487,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047352219665185293,
      "loss": 5.1154,
      "mean_token_accuracy": 0.17048840820789338,
      "num_tokens": 36287308.0,
      "step": 16560
    },
    {
      "entropy": 5.5755480289459225,
      "epoch": 1.6440055577610164,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004735055184150629,
      "loss": 5.1302,
      "mean_token_accuracy": 0.17566781789064406,
      "num_tokens": 36298966.0,
      "step": 16565
    },
    {
      "entropy": 5.583449840545654,
      "epoch": 1.644501786423184,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047348883525576385,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17239987403154372,
      "num_tokens": 36310498.0,
      "step": 16570
    },
    {
      "entropy": 5.633149242401123,
      "epoch": 1.6449980150853514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004734721471743694,
      "loss": 5.1748,
      "mean_token_accuracy": 0.1761878401041031,
      "num_tokens": 36321589.0,
      "step": 16575
    },
    {
      "entropy": 5.570087242126465,
      "epoch": 1.6454942437475188,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004734554541712933,
      "loss": 5.1031,
      "mean_token_accuracy": 0.16844381093978883,
      "num_tokens": 36333041.0,
      "step": 16580
    },
    {
      "entropy": 5.622600460052491,
      "epoch": 1.6459904724096863,
      "grad_norm": 1.0,
      "learning_rate": 0.0004734387562469496,
      "loss": 5.1818,
      "mean_token_accuracy": 0.1709860682487488,
      "num_tokens": 36344225.0,
      "step": 16585
    },
    {
      "entropy": 5.617960739135742,
      "epoch": 1.6464867010718538,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047342205340175235,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1778441399335861,
      "num_tokens": 36355524.0,
      "step": 16590
    },
    {
      "entropy": 5.651898765563965,
      "epoch": 1.6469829297340213,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047340534563611574,
      "loss": 5.147,
      "mean_token_accuracy": 0.17341646254062654,
      "num_tokens": 36366495.0,
      "step": 16595
    },
    {
      "entropy": 5.607076263427734,
      "epoch": 1.647479158396189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047338863295045414,
      "loss": 5.1381,
      "mean_token_accuracy": 0.17128214985132217,
      "num_tokens": 36378316.0,
      "step": 16600
    },
    {
      "entropy": 5.651200628280639,
      "epoch": 1.6479753870583564,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047337191534518186,
      "loss": 5.2283,
      "mean_token_accuracy": 0.16330428272485734,
      "num_tokens": 36390500.0,
      "step": 16605
    },
    {
      "entropy": 5.622259044647217,
      "epoch": 1.6484716157205241,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004733551928207135,
      "loss": 5.1605,
      "mean_token_accuracy": 0.17350108027458191,
      "num_tokens": 36403831.0,
      "step": 16610
    },
    {
      "entropy": 5.610906028747559,
      "epoch": 1.6489678443826916,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004733384653774639,
      "loss": 5.1205,
      "mean_token_accuracy": 0.169162355363369,
      "num_tokens": 36414896.0,
      "step": 16615
    },
    {
      "entropy": 5.577035140991211,
      "epoch": 1.649464073044859,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047332173301584764,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17698683440685273,
      "num_tokens": 36426260.0,
      "step": 16620
    },
    {
      "entropy": 5.505938959121704,
      "epoch": 1.6499603017070266,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047330499573627975,
      "loss": 5.0232,
      "mean_token_accuracy": 0.18077400773763658,
      "num_tokens": 36437125.0,
      "step": 16625
    },
    {
      "entropy": 5.569142389297485,
      "epoch": 1.650456530369194,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047328825353917526,
      "loss": 5.0673,
      "mean_token_accuracy": 0.17768927067518234,
      "num_tokens": 36447507.0,
      "step": 16630
    },
    {
      "entropy": 5.676078748703003,
      "epoch": 1.6509527590313615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047327150642494944,
      "loss": 5.2344,
      "mean_token_accuracy": 0.17096995413303376,
      "num_tokens": 36459134.0,
      "step": 16635
    },
    {
      "entropy": 5.64393253326416,
      "epoch": 1.6514489876935292,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004732547543940174,
      "loss": 5.1392,
      "mean_token_accuracy": 0.16744721680879593,
      "num_tokens": 36469594.0,
      "step": 16640
    },
    {
      "entropy": 5.596857261657715,
      "epoch": 1.6519452163556967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004732379974467947,
      "loss": 5.1451,
      "mean_token_accuracy": 0.17411863207817077,
      "num_tokens": 36480477.0,
      "step": 16645
    },
    {
      "entropy": 5.561244964599609,
      "epoch": 1.6524414450178644,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004732212355836968,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17627407759428024,
      "num_tokens": 36491304.0,
      "step": 16650
    },
    {
      "entropy": 5.660111999511718,
      "epoch": 1.6529376736800319,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004732044688051395,
      "loss": 5.2077,
      "mean_token_accuracy": 0.16461412012577056,
      "num_tokens": 36502030.0,
      "step": 16655
    },
    {
      "entropy": 5.631765604019165,
      "epoch": 1.6534339023421993,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047318769711153833,
      "loss": 5.2348,
      "mean_token_accuracy": 0.16811648458242417,
      "num_tokens": 36514405.0,
      "step": 16660
    },
    {
      "entropy": 5.659211730957031,
      "epoch": 1.6539301310043668,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004731709205033094,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17175668179988862,
      "num_tokens": 36524626.0,
      "step": 16665
    },
    {
      "entropy": 5.629488134384156,
      "epoch": 1.6544263596665343,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047315413898086863,
      "loss": 5.1038,
      "mean_token_accuracy": 0.1784445211291313,
      "num_tokens": 36535340.0,
      "step": 16670
    },
    {
      "entropy": 5.647333717346191,
      "epoch": 1.6549225883287018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004731373525446322,
      "loss": 5.1158,
      "mean_token_accuracy": 0.1741843342781067,
      "num_tokens": 36547189.0,
      "step": 16675
    },
    {
      "entropy": 5.613614273071289,
      "epoch": 1.6554188169908692,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004731205611950164,
      "loss": 5.1925,
      "mean_token_accuracy": 0.17074184715747834,
      "num_tokens": 36558177.0,
      "step": 16680
    },
    {
      "entropy": 5.58017315864563,
      "epoch": 1.655915045653037,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047310376493243754,
      "loss": 5.0079,
      "mean_token_accuracy": 0.17769603431224823,
      "num_tokens": 36569559.0,
      "step": 16685
    },
    {
      "entropy": 5.603822660446167,
      "epoch": 1.6564112743152044,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004730869637573122,
      "loss": 5.0883,
      "mean_token_accuracy": 0.17639841884374619,
      "num_tokens": 36579617.0,
      "step": 16690
    },
    {
      "entropy": 5.577926540374756,
      "epoch": 1.656907502977372,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00047307015767005705,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1760268360376358,
      "num_tokens": 36590427.0,
      "step": 16695
    },
    {
      "entropy": 5.596349096298217,
      "epoch": 1.6574037316395396,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004730533466710888,
      "loss": 5.0826,
      "mean_token_accuracy": 0.17677031457424164,
      "num_tokens": 36602313.0,
      "step": 16700
    },
    {
      "entropy": 5.661788082122802,
      "epoch": 1.657899960301707,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004730365307608242,
      "loss": 5.1809,
      "mean_token_accuracy": 0.16944610476493835,
      "num_tokens": 36615191.0,
      "step": 16705
    },
    {
      "entropy": 5.574722766876221,
      "epoch": 1.6583961889638745,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047301970993968054,
      "loss": 5.091,
      "mean_token_accuracy": 0.18071553409099578,
      "num_tokens": 36624692.0,
      "step": 16710
    },
    {
      "entropy": 5.612470197677612,
      "epoch": 1.658892417626042,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047300288420807463,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1663797155022621,
      "num_tokens": 36635961.0,
      "step": 16715
    },
    {
      "entropy": 5.543233203887939,
      "epoch": 1.6593886462882095,
      "grad_norm": 1.0,
      "learning_rate": 0.00047298605356642387,
      "loss": 5.1238,
      "mean_token_accuracy": 0.17286624014377594,
      "num_tokens": 36646710.0,
      "step": 16720
    },
    {
      "entropy": 5.622425651550293,
      "epoch": 1.6598848749503772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047296921801514565,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1696367844939232,
      "num_tokens": 36657168.0,
      "step": 16725
    },
    {
      "entropy": 5.65752272605896,
      "epoch": 1.6603811036125447,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004729523775546574,
      "loss": 5.1978,
      "mean_token_accuracy": 0.16970750838518142,
      "num_tokens": 36668428.0,
      "step": 16730
    },
    {
      "entropy": 5.621710109710693,
      "epoch": 1.6608773322747123,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004729355321853767,
      "loss": 5.122,
      "mean_token_accuracy": 0.17651771008968353,
      "num_tokens": 36679361.0,
      "step": 16735
    },
    {
      "entropy": 5.702149057388306,
      "epoch": 1.6613735609368798,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004729186819077214,
      "loss": 5.199,
      "mean_token_accuracy": 0.16845204979181289,
      "num_tokens": 36689480.0,
      "step": 16740
    },
    {
      "entropy": 5.680377960205078,
      "epoch": 1.6618697895990473,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004729018267221092,
      "loss": 5.2329,
      "mean_token_accuracy": 0.1672963961958885,
      "num_tokens": 36700420.0,
      "step": 16745
    },
    {
      "entropy": 5.633373546600342,
      "epoch": 1.6623660182612148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004728849666289581,
      "loss": 5.1848,
      "mean_token_accuracy": 0.17319359332323075,
      "num_tokens": 36710106.0,
      "step": 16750
    },
    {
      "entropy": 5.6008484840393065,
      "epoch": 1.6628622469233822,
      "grad_norm": 1.0,
      "learning_rate": 0.00047286810162868625,
      "loss": 5.0522,
      "mean_token_accuracy": 0.1816716492176056,
      "num_tokens": 36719630.0,
      "step": 16755
    },
    {
      "entropy": 5.542885160446167,
      "epoch": 1.6633584755855497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004728512317217119,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18086077272891998,
      "num_tokens": 36729379.0,
      "step": 16760
    },
    {
      "entropy": 5.569194364547729,
      "epoch": 1.6638547042477172,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047283435690845336,
      "loss": 5.0776,
      "mean_token_accuracy": 0.18063281625509262,
      "num_tokens": 36740103.0,
      "step": 16765
    },
    {
      "entropy": 5.6080433368682865,
      "epoch": 1.664350932909885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047281747718932906,
      "loss": 5.1341,
      "mean_token_accuracy": 0.17510365396738053,
      "num_tokens": 36750928.0,
      "step": 16770
    },
    {
      "entropy": 5.591271448135376,
      "epoch": 1.6648471615720524,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004728005925647575,
      "loss": 5.1267,
      "mean_token_accuracy": 0.17279553562402725,
      "num_tokens": 36761415.0,
      "step": 16775
    },
    {
      "entropy": 5.558135318756103,
      "epoch": 1.66534339023422,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047278370303515744,
      "loss": 5.171,
      "mean_token_accuracy": 0.17234934419393538,
      "num_tokens": 36772560.0,
      "step": 16780
    },
    {
      "entropy": 5.624247646331787,
      "epoch": 1.6658396188963875,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004727668086009478,
      "loss": 5.1572,
      "mean_token_accuracy": 0.17446713596582414,
      "num_tokens": 36783597.0,
      "step": 16785
    },
    {
      "entropy": 5.702061128616333,
      "epoch": 1.666335847558555,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047274990926254753,
      "loss": 5.2373,
      "mean_token_accuracy": 0.16710264831781388,
      "num_tokens": 36794219.0,
      "step": 16790
    },
    {
      "entropy": 5.54986081123352,
      "epoch": 1.6668320762207225,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004727330050203755,
      "loss": 5.077,
      "mean_token_accuracy": 0.17584026902914046,
      "num_tokens": 36804156.0,
      "step": 16795
    },
    {
      "entropy": 5.538628149032593,
      "epoch": 1.66732830488289,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004727160958748511,
      "loss": 5.0951,
      "mean_token_accuracy": 0.17700162529945374,
      "num_tokens": 36814414.0,
      "step": 16800
    },
    {
      "entropy": 5.59779577255249,
      "epoch": 1.6678245335450574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047269918182639355,
      "loss": 5.0721,
      "mean_token_accuracy": 0.16848777681589128,
      "num_tokens": 36826173.0,
      "step": 16805
    },
    {
      "entropy": 5.629490756988526,
      "epoch": 1.6683207622072251,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004726822628754222,
      "loss": 5.1648,
      "mean_token_accuracy": 0.1704873964190483,
      "num_tokens": 36837726.0,
      "step": 16810
    },
    {
      "entropy": 5.6095871925354,
      "epoch": 1.6688169908693926,
      "grad_norm": 1.0,
      "learning_rate": 0.00047266533902235683,
      "loss": 5.1224,
      "mean_token_accuracy": 0.1667923539876938,
      "num_tokens": 36848605.0,
      "step": 16815
    },
    {
      "entropy": 5.626727724075318,
      "epoch": 1.66931321953156,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004726484102676169,
      "loss": 5.1399,
      "mean_token_accuracy": 0.1761493593454361,
      "num_tokens": 36858730.0,
      "step": 16820
    },
    {
      "entropy": 5.5997237205505375,
      "epoch": 1.6698094481937278,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004726314766116223,
      "loss": 5.015,
      "mean_token_accuracy": 0.18054923117160798,
      "num_tokens": 36868239.0,
      "step": 16825
    },
    {
      "entropy": 5.562602567672729,
      "epoch": 1.6703056768558953,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000472614538054793,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17564985454082488,
      "num_tokens": 36878077.0,
      "step": 16830
    },
    {
      "entropy": 5.56316647529602,
      "epoch": 1.6708019055180627,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047259759459754893,
      "loss": 5.1344,
      "mean_token_accuracy": 0.1646384671330452,
      "num_tokens": 36888874.0,
      "step": 16835
    },
    {
      "entropy": 5.61406717300415,
      "epoch": 1.6712981341802302,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004725806462403103,
      "loss": 5.1369,
      "mean_token_accuracy": 0.17481710612773896,
      "num_tokens": 36898584.0,
      "step": 16840
    },
    {
      "entropy": 5.647138786315918,
      "epoch": 1.6717943628423977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047256369298349736,
      "loss": 5.1291,
      "mean_token_accuracy": 0.1711537942290306,
      "num_tokens": 36908662.0,
      "step": 16845
    },
    {
      "entropy": 5.570994281768799,
      "epoch": 1.6722905915045652,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047254673482753065,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17797506004571914,
      "num_tokens": 36919482.0,
      "step": 16850
    },
    {
      "entropy": 5.562100410461426,
      "epoch": 1.6727868201667329,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047252977177283045,
      "loss": 5.1107,
      "mean_token_accuracy": 0.17097083032131194,
      "num_tokens": 36930470.0,
      "step": 16855
    },
    {
      "entropy": 5.594316101074218,
      "epoch": 1.6732830488289003,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047251280381981764,
      "loss": 5.0988,
      "mean_token_accuracy": 0.18057047873735427,
      "num_tokens": 36942837.0,
      "step": 16860
    },
    {
      "entropy": 5.66077733039856,
      "epoch": 1.673779277491068,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004724958309689129,
      "loss": 5.169,
      "mean_token_accuracy": 0.16790551394224168,
      "num_tokens": 36952994.0,
      "step": 16865
    },
    {
      "entropy": 5.619121646881103,
      "epoch": 1.6742755061532355,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047247885322053705,
      "loss": 5.1462,
      "mean_token_accuracy": 0.1700314089655876,
      "num_tokens": 36963973.0,
      "step": 16870
    },
    {
      "entropy": 5.528723478317261,
      "epoch": 1.674771734815403,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00047246187057511127,
      "loss": 5.0867,
      "mean_token_accuracy": 0.17749009728431703,
      "num_tokens": 36975897.0,
      "step": 16875
    },
    {
      "entropy": 5.587921857833862,
      "epoch": 1.6752679634775705,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004724448830330565,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17387866824865342,
      "num_tokens": 36986481.0,
      "step": 16880
    },
    {
      "entropy": 5.63957839012146,
      "epoch": 1.675764192139738,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047242789059479413,
      "loss": 5.1898,
      "mean_token_accuracy": 0.16843899488449096,
      "num_tokens": 36997863.0,
      "step": 16885
    },
    {
      "entropy": 5.652133893966675,
      "epoch": 1.6762604208019054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047241089326074547,
      "loss": 5.1345,
      "mean_token_accuracy": 0.16859585493803025,
      "num_tokens": 37009294.0,
      "step": 16890
    },
    {
      "entropy": 5.6472259044647215,
      "epoch": 1.6767566494640729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004723938910313321,
      "loss": 5.1289,
      "mean_token_accuracy": 0.17697568237781525,
      "num_tokens": 37019748.0,
      "step": 16895
    },
    {
      "entropy": 5.658229160308838,
      "epoch": 1.6772528781262406,
      "grad_norm": 1.0,
      "learning_rate": 0.00047237688390697555,
      "loss": 5.1765,
      "mean_token_accuracy": 0.17665166333317756,
      "num_tokens": 37030903.0,
      "step": 16900
    },
    {
      "entropy": 5.610582160949707,
      "epoch": 1.677749106788408,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004723598718880975,
      "loss": 5.1048,
      "mean_token_accuracy": 0.17559353560209273,
      "num_tokens": 37042551.0,
      "step": 16905
    },
    {
      "entropy": 5.6218335151672365,
      "epoch": 1.6782453354505757,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047234285497512005,
      "loss": 5.1628,
      "mean_token_accuracy": 0.16890343725681306,
      "num_tokens": 37053151.0,
      "step": 16910
    },
    {
      "entropy": 5.664807510375977,
      "epoch": 1.6787415641127432,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00047232583316846494,
      "loss": 5.1714,
      "mean_token_accuracy": 0.1659543365240097,
      "num_tokens": 37064421.0,
      "step": 16915
    },
    {
      "entropy": 5.637000274658203,
      "epoch": 1.6792377927749107,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004723088064685544,
      "loss": 5.117,
      "mean_token_accuracy": 0.18055375069379806,
      "num_tokens": 37074758.0,
      "step": 16920
    },
    {
      "entropy": 5.595488977432251,
      "epoch": 1.6797340214370782,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047229177487581055,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17691901624202727,
      "num_tokens": 37084898.0,
      "step": 16925
    },
    {
      "entropy": 5.646871423721313,
      "epoch": 1.6802302500992456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047227473839065587,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16822021156549455,
      "num_tokens": 37094366.0,
      "step": 16930
    },
    {
      "entropy": 5.737091398239135,
      "epoch": 1.6807264787614131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047225769701351284,
      "loss": 5.2332,
      "mean_token_accuracy": 0.16562853306531905,
      "num_tokens": 37104965.0,
      "step": 16935
    },
    {
      "entropy": 5.582026863098145,
      "epoch": 1.6812227074235808,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004722406507448038,
      "loss": 5.0915,
      "mean_token_accuracy": 0.1741964712738991,
      "num_tokens": 37115622.0,
      "step": 16940
    },
    {
      "entropy": 5.663673353195191,
      "epoch": 1.6817189360857483,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047222359958495175,
      "loss": 5.1719,
      "mean_token_accuracy": 0.16799180656671525,
      "num_tokens": 37127234.0,
      "step": 16945
    },
    {
      "entropy": 5.645137071609497,
      "epoch": 1.682215164747916,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004722065435343794,
      "loss": 5.0622,
      "mean_token_accuracy": 0.17556205689907073,
      "num_tokens": 37136533.0,
      "step": 16950
    },
    {
      "entropy": 5.622964715957641,
      "epoch": 1.6827113934100835,
      "grad_norm": 1.0,
      "learning_rate": 0.0004721894825935097,
      "loss": 5.1507,
      "mean_token_accuracy": 0.17153382003307344,
      "num_tokens": 37147821.0,
      "step": 16955
    },
    {
      "entropy": 5.555866718292236,
      "epoch": 1.683207622072251,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004721724167627657,
      "loss": 5.1435,
      "mean_token_accuracy": 0.17195263803005217,
      "num_tokens": 37158929.0,
      "step": 16960
    },
    {
      "entropy": 5.659469175338745,
      "epoch": 1.6837038507344184,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004721553460425708,
      "loss": 5.1473,
      "mean_token_accuracy": 0.17934174984693527,
      "num_tokens": 37169817.0,
      "step": 16965
    },
    {
      "entropy": 5.680769729614258,
      "epoch": 1.684200079396586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000472138270433348,
      "loss": 5.269,
      "mean_token_accuracy": 0.16132428348064423,
      "num_tokens": 37181009.0,
      "step": 16970
    },
    {
      "entropy": 5.5956672668457035,
      "epoch": 1.6846963080587534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047212118993552097,
      "loss": 5.0997,
      "mean_token_accuracy": 0.17441382706165315,
      "num_tokens": 37192110.0,
      "step": 16975
    },
    {
      "entropy": 5.593505096435547,
      "epoch": 1.6851925367209208,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004721041045495131,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17718729227781296,
      "num_tokens": 37203426.0,
      "step": 16980
    },
    {
      "entropy": 5.626435708999634,
      "epoch": 1.6856887653830885,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047208701427574817,
      "loss": 5.2208,
      "mean_token_accuracy": 0.16637452840805053,
      "num_tokens": 37215025.0,
      "step": 16985
    },
    {
      "entropy": 5.686810731887817,
      "epoch": 1.686184994045256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047206991911465,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1786564141511917,
      "num_tokens": 37225653.0,
      "step": 16990
    },
    {
      "entropy": 5.635810232162475,
      "epoch": 1.6866812227074237,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004720528190666425,
      "loss": 5.1272,
      "mean_token_accuracy": 0.1684884876012802,
      "num_tokens": 37237282.0,
      "step": 16995
    },
    {
      "entropy": 5.535738754272461,
      "epoch": 1.6871774513695912,
      "grad_norm": 1.0,
      "learning_rate": 0.00047203571413214967,
      "loss": 5.0719,
      "mean_token_accuracy": 0.1842014968395233,
      "num_tokens": 37248476.0,
      "step": 17000
    },
    {
      "entropy": 5.578322410583496,
      "epoch": 1.6876736800317587,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004720186043115957,
      "loss": 5.0879,
      "mean_token_accuracy": 0.171076700091362,
      "num_tokens": 37259354.0,
      "step": 17005
    },
    {
      "entropy": 5.611423969268799,
      "epoch": 1.6881699086939261,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004720014896054048,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17046214640140533,
      "num_tokens": 37269895.0,
      "step": 17010
    },
    {
      "entropy": 5.598416137695312,
      "epoch": 1.6886661373560936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004719843700140017,
      "loss": 5.1044,
      "mean_token_accuracy": 0.1826440289616585,
      "num_tokens": 37281201.0,
      "step": 17015
    },
    {
      "entropy": 5.592598342895508,
      "epoch": 1.689162366018261,
      "grad_norm": 1.125,
      "learning_rate": 0.0004719672455378105,
      "loss": 5.1306,
      "mean_token_accuracy": 0.16848499774932862,
      "num_tokens": 37293362.0,
      "step": 17020
    },
    {
      "entropy": 5.672120523452759,
      "epoch": 1.6896585946804288,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000471950116177256,
      "loss": 5.1589,
      "mean_token_accuracy": 0.1673130676150322,
      "num_tokens": 37304291.0,
      "step": 17025
    },
    {
      "entropy": 5.620660018920899,
      "epoch": 1.6901548233425963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004719329819327631,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17119102776050568,
      "num_tokens": 37315608.0,
      "step": 17030
    },
    {
      "entropy": 5.594802045822144,
      "epoch": 1.690651052004764,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004719158428047566,
      "loss": 5.0802,
      "mean_token_accuracy": 0.17064009755849838,
      "num_tokens": 37327559.0,
      "step": 17035
    },
    {
      "entropy": 5.588363075256348,
      "epoch": 1.6911472806669314,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004718986987936615,
      "loss": 5.1235,
      "mean_token_accuracy": 0.17742763906717302,
      "num_tokens": 37338101.0,
      "step": 17040
    },
    {
      "entropy": 5.631590366363525,
      "epoch": 1.691643509329099,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047188154989990296,
      "loss": 5.2116,
      "mean_token_accuracy": 0.1704496070742607,
      "num_tokens": 37349124.0,
      "step": 17045
    },
    {
      "entropy": 5.5856695652008055,
      "epoch": 1.6921397379912664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004718643961239062,
      "loss": 5.0656,
      "mean_token_accuracy": 0.17955634295940398,
      "num_tokens": 37360061.0,
      "step": 17050
    },
    {
      "entropy": 5.611816883087158,
      "epoch": 1.6926359666534339,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047184723746609666,
      "loss": 5.1525,
      "mean_token_accuracy": 0.17059623450040817,
      "num_tokens": 37371210.0,
      "step": 17055
    },
    {
      "entropy": 5.5947874069213865,
      "epoch": 1.6931321953156013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047183007392689977,
      "loss": 5.0833,
      "mean_token_accuracy": 0.16953487396240235,
      "num_tokens": 37382547.0,
      "step": 17060
    },
    {
      "entropy": 5.620542860031128,
      "epoch": 1.6936284239777688,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047181290550674115,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17517104148864746,
      "num_tokens": 37392311.0,
      "step": 17065
    },
    {
      "entropy": 5.588214588165283,
      "epoch": 1.6941246526399365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004717957322060466,
      "loss": 5.0893,
      "mean_token_accuracy": 0.1779579222202301,
      "num_tokens": 37402626.0,
      "step": 17070
    },
    {
      "entropy": 5.589341259002685,
      "epoch": 1.694620881302104,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004717785540252419,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17097070962190627,
      "num_tokens": 37412100.0,
      "step": 17075
    },
    {
      "entropy": 5.536940193176269,
      "epoch": 1.6951171099642717,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004717613709647532,
      "loss": 5.0113,
      "mean_token_accuracy": 0.17956485450267792,
      "num_tokens": 37421735.0,
      "step": 17080
    },
    {
      "entropy": 5.677174186706543,
      "epoch": 1.6956133386264391,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047174418302500635,
      "loss": 5.2048,
      "mean_token_accuracy": 0.1626304492354393,
      "num_tokens": 37432179.0,
      "step": 17085
    },
    {
      "entropy": 5.613541841506958,
      "epoch": 1.6961095672886066,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004717269902064278,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1828702211380005,
      "num_tokens": 37443276.0,
      "step": 17090
    },
    {
      "entropy": 5.622309255599975,
      "epoch": 1.696605795950774,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004717097925094438,
      "loss": 5.1947,
      "mean_token_accuracy": 0.16982924193143845,
      "num_tokens": 37454537.0,
      "step": 17095
    },
    {
      "entropy": 5.610753107070923,
      "epoch": 1.6971020246129416,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716925899344808,
      "loss": 5.1015,
      "mean_token_accuracy": 0.17051625102758408,
      "num_tokens": 37465493.0,
      "step": 17100
    },
    {
      "entropy": 5.563348770141602,
      "epoch": 1.697598253275109,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004716753824819654,
      "loss": 5.0953,
      "mean_token_accuracy": 0.17190059572458266,
      "num_tokens": 37475511.0,
      "step": 17105
    },
    {
      "entropy": 5.540134763717651,
      "epoch": 1.6980944819372767,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004716581701523243,
      "loss": 5.0148,
      "mean_token_accuracy": 0.1720792457461357,
      "num_tokens": 37487046.0,
      "step": 17110
    },
    {
      "entropy": 5.592929697036743,
      "epoch": 1.6985907105994442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047164095294598433,
      "loss": 5.1654,
      "mean_token_accuracy": 0.17208407521247865,
      "num_tokens": 37498398.0,
      "step": 17115
    },
    {
      "entropy": 5.638840866088867,
      "epoch": 1.699086939261612,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047162373086337243,
      "loss": 5.1652,
      "mean_token_accuracy": 0.16845489293336868,
      "num_tokens": 37509397.0,
      "step": 17120
    },
    {
      "entropy": 5.631436061859131,
      "epoch": 1.6995831679237794,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047160650390491574,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18060020059347154,
      "num_tokens": 37519284.0,
      "step": 17125
    },
    {
      "entropy": 5.610557126998901,
      "epoch": 1.7000793965859469,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004715892720710414,
      "loss": 5.1432,
      "mean_token_accuracy": 0.17462328672409058,
      "num_tokens": 37529699.0,
      "step": 17130
    },
    {
      "entropy": 5.663481855392456,
      "epoch": 1.7005756252481143,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004715720353621767,
      "loss": 5.1881,
      "mean_token_accuracy": 0.16586233377456666,
      "num_tokens": 37541756.0,
      "step": 17135
    },
    {
      "entropy": 5.581813669204712,
      "epoch": 1.7010718539102818,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047155479377874917,
      "loss": 5.0871,
      "mean_token_accuracy": 0.17903389930725097,
      "num_tokens": 37552693.0,
      "step": 17140
    },
    {
      "entropy": 5.576038408279419,
      "epoch": 1.7015680825724493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047153754732118616,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17487225979566573,
      "num_tokens": 37563826.0,
      "step": 17145
    },
    {
      "entropy": 5.589671802520752,
      "epoch": 1.7020643112346168,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004715202959899156,
      "loss": 5.1121,
      "mean_token_accuracy": 0.17631376683712005,
      "num_tokens": 37576645.0,
      "step": 17150
    },
    {
      "entropy": 5.47814679145813,
      "epoch": 1.7025605398967845,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004715030397853651,
      "loss": 4.9849,
      "mean_token_accuracy": 0.1817948877811432,
      "num_tokens": 37587467.0,
      "step": 17155
    },
    {
      "entropy": 5.635016059875488,
      "epoch": 1.703056768558952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047148577870796264,
      "loss": 5.1644,
      "mean_token_accuracy": 0.16771882474422456,
      "num_tokens": 37598475.0,
      "step": 17160
    },
    {
      "entropy": 5.687562131881714,
      "epoch": 1.7035529972211196,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004714685127581363,
      "loss": 5.2516,
      "mean_token_accuracy": 0.16771269738674163,
      "num_tokens": 37610260.0,
      "step": 17165
    },
    {
      "entropy": 5.6327817916870115,
      "epoch": 1.704049225883287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047145124193631415,
      "loss": 5.1875,
      "mean_token_accuracy": 0.16802210360765457,
      "num_tokens": 37621129.0,
      "step": 17170
    },
    {
      "entropy": 5.599156522750855,
      "epoch": 1.7045454545454546,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047143396624292454,
      "loss": 5.0814,
      "mean_token_accuracy": 0.1730399325489998,
      "num_tokens": 37633130.0,
      "step": 17175
    },
    {
      "entropy": 5.611918926239014,
      "epoch": 1.705041683207622,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047141668567839584,
      "loss": 5.2027,
      "mean_token_accuracy": 0.17446798980236053,
      "num_tokens": 37644247.0,
      "step": 17180
    },
    {
      "entropy": 5.64642825126648,
      "epoch": 1.7055379118697895,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004713994002431566,
      "loss": 5.1127,
      "mean_token_accuracy": 0.17664121240377426,
      "num_tokens": 37655602.0,
      "step": 17185
    },
    {
      "entropy": 5.568411254882813,
      "epoch": 1.706034140531957,
      "grad_norm": 0.875,
      "learning_rate": 0.00047138210993763534,
      "loss": 5.0445,
      "mean_token_accuracy": 0.17921496629714967,
      "num_tokens": 37667150.0,
      "step": 17190
    },
    {
      "entropy": 5.574297618865967,
      "epoch": 1.7065303691941247,
      "grad_norm": 0.890625,
      "learning_rate": 0.000471364814762261,
      "loss": 5.0614,
      "mean_token_accuracy": 0.1797898605465889,
      "num_tokens": 37677626.0,
      "step": 17195
    },
    {
      "entropy": 5.664226865768432,
      "epoch": 1.7070265978562922,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004713475147174624,
      "loss": 5.2018,
      "mean_token_accuracy": 0.17003358751535416,
      "num_tokens": 37689050.0,
      "step": 17200
    },
    {
      "entropy": 5.66594820022583,
      "epoch": 1.7075228265184597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004713302098036685,
      "loss": 5.091,
      "mean_token_accuracy": 0.18070708513259887,
      "num_tokens": 37699190.0,
      "step": 17205
    },
    {
      "entropy": 5.543094778060913,
      "epoch": 1.7080190551806274,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004713129000213085,
      "loss": 5.0776,
      "mean_token_accuracy": 0.17788154035806655,
      "num_tokens": 37710465.0,
      "step": 17210
    },
    {
      "entropy": 5.5975257396698,
      "epoch": 1.7085152838427948,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004712955853708115,
      "loss": 5.1967,
      "mean_token_accuracy": 0.17106454372406005,
      "num_tokens": 37722426.0,
      "step": 17215
    },
    {
      "entropy": 5.568806457519531,
      "epoch": 1.7090115125049623,
      "grad_norm": 1.0,
      "learning_rate": 0.000471278265852607,
      "loss": 5.0755,
      "mean_token_accuracy": 0.17916608601808548,
      "num_tokens": 37733104.0,
      "step": 17220
    },
    {
      "entropy": 5.551222896575927,
      "epoch": 1.7095077411671298,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004712609414671245,
      "loss": 5.0146,
      "mean_token_accuracy": 0.1794799894094467,
      "num_tokens": 37742747.0,
      "step": 17225
    },
    {
      "entropy": 5.5846398830413815,
      "epoch": 1.7100039698292973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004712436122147935,
      "loss": 5.0479,
      "mean_token_accuracy": 0.17757536023855208,
      "num_tokens": 37753338.0,
      "step": 17230
    },
    {
      "entropy": 5.5456665515899655,
      "epoch": 1.7105001984914647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004712262780960438,
      "loss": 5.0659,
      "mean_token_accuracy": 0.17865682542324066,
      "num_tokens": 37763122.0,
      "step": 17235
    },
    {
      "entropy": 5.564337491989136,
      "epoch": 1.7109964271536324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047120893911130525,
      "loss": 5.0727,
      "mean_token_accuracy": 0.16953400820493697,
      "num_tokens": 37774677.0,
      "step": 17240
    },
    {
      "entropy": 5.602055883407592,
      "epoch": 1.7114926558158,
      "grad_norm": 0.890625,
      "learning_rate": 0.00047119159526100773,
      "loss": 5.0729,
      "mean_token_accuracy": 0.16994416415691377,
      "num_tokens": 37786186.0,
      "step": 17245
    },
    {
      "entropy": 5.539776754379273,
      "epoch": 1.7119888844779676,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047117424654558146,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17685806155204772,
      "num_tokens": 37796837.0,
      "step": 17250
    },
    {
      "entropy": 5.6028564453125,
      "epoch": 1.712485113140135,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004711568929654566,
      "loss": 5.1636,
      "mean_token_accuracy": 0.1698177322745323,
      "num_tokens": 37808813.0,
      "step": 17255
    },
    {
      "entropy": 5.6151529312133786,
      "epoch": 1.7129813418023025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004711395345210635,
      "loss": 5.0306,
      "mean_token_accuracy": 0.1766620397567749,
      "num_tokens": 37819770.0,
      "step": 17260
    },
    {
      "entropy": 5.616785669326783,
      "epoch": 1.71347757046447,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004711221712128325,
      "loss": 5.1117,
      "mean_token_accuracy": 0.17415693253278733,
      "num_tokens": 37831831.0,
      "step": 17265
    },
    {
      "entropy": 5.566071891784668,
      "epoch": 1.7139737991266375,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004711048030411943,
      "loss": 5.1596,
      "mean_token_accuracy": 0.17029980570077896,
      "num_tokens": 37843092.0,
      "step": 17270
    },
    {
      "entropy": 5.590189361572266,
      "epoch": 1.714470027788805,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047108743000657963,
      "loss": 5.1367,
      "mean_token_accuracy": 0.17724588066339492,
      "num_tokens": 37853192.0,
      "step": 17275
    },
    {
      "entropy": 5.660770416259766,
      "epoch": 1.7149662564509724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004710700521094192,
      "loss": 5.152,
      "mean_token_accuracy": 0.17128089517354966,
      "num_tokens": 37863510.0,
      "step": 17280
    },
    {
      "entropy": 5.589938974380493,
      "epoch": 1.7154624851131401,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00047105266935014385,
      "loss": 5.0566,
      "mean_token_accuracy": 0.18139216750860215,
      "num_tokens": 37876156.0,
      "step": 17285
    },
    {
      "entropy": 5.5575121402740475,
      "epoch": 1.7159587137753076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047103528172918487,
      "loss": 5.1032,
      "mean_token_accuracy": 0.17726526409387589,
      "num_tokens": 37887153.0,
      "step": 17290
    },
    {
      "entropy": 5.613091564178466,
      "epoch": 1.7164549424374753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004710178892469733,
      "loss": 5.0994,
      "mean_token_accuracy": 0.1767705872654915,
      "num_tokens": 37898820.0,
      "step": 17295
    },
    {
      "entropy": 5.607368659973145,
      "epoch": 1.7169511710996428,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047100049190394044,
      "loss": 5.0835,
      "mean_token_accuracy": 0.17401736974716187,
      "num_tokens": 37910257.0,
      "step": 17300
    },
    {
      "entropy": 5.587104082107544,
      "epoch": 1.7174473997618103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004709830897005178,
      "loss": 5.1818,
      "mean_token_accuracy": 0.17060097306966782,
      "num_tokens": 37921574.0,
      "step": 17305
    },
    {
      "entropy": 5.670722341537475,
      "epoch": 1.7179436284239777,
      "grad_norm": 1.0,
      "learning_rate": 0.0004709656826371367,
      "loss": 5.1477,
      "mean_token_accuracy": 0.1725548893213272,
      "num_tokens": 37932741.0,
      "step": 17310
    },
    {
      "entropy": 5.6021661281585695,
      "epoch": 1.7184398570861452,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004709482707142291,
      "loss": 5.1288,
      "mean_token_accuracy": 0.17645669728517532,
      "num_tokens": 37944592.0,
      "step": 17315
    },
    {
      "entropy": 5.589171838760376,
      "epoch": 1.7189360857483127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047093085393222646,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18001656979322433,
      "num_tokens": 37954983.0,
      "step": 17320
    },
    {
      "entropy": 5.611456060409546,
      "epoch": 1.7194323144104804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047091343229156094,
      "loss": 5.1567,
      "mean_token_accuracy": 0.17272499054670334,
      "num_tokens": 37964491.0,
      "step": 17325
    },
    {
      "entropy": 5.584730625152588,
      "epoch": 1.7199285430726479,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004708960057926644,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1762470245361328,
      "num_tokens": 37976140.0,
      "step": 17330
    },
    {
      "entropy": 5.579538202285766,
      "epoch": 1.7204247717348156,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047087857443596905,
      "loss": 5.101,
      "mean_token_accuracy": 0.18025367110967636,
      "num_tokens": 37987776.0,
      "step": 17335
    },
    {
      "entropy": 5.604762887954712,
      "epoch": 1.720921000396983,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047086113822190716,
      "loss": 5.1443,
      "mean_token_accuracy": 0.1730215758085251,
      "num_tokens": 37998088.0,
      "step": 17340
    },
    {
      "entropy": 5.580350255966186,
      "epoch": 1.7214172290591505,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00047084369715091103,
      "loss": 5.1309,
      "mean_token_accuracy": 0.17242063134908675,
      "num_tokens": 38009802.0,
      "step": 17345
    },
    {
      "entropy": 5.635538816452026,
      "epoch": 1.721913457721318,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004708262512234133,
      "loss": 5.1682,
      "mean_token_accuracy": 0.1798313468694687,
      "num_tokens": 38020797.0,
      "step": 17350
    },
    {
      "entropy": 5.654097318649292,
      "epoch": 1.7224096863834855,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004708088004398464,
      "loss": 5.1681,
      "mean_token_accuracy": 0.17000040262937546,
      "num_tokens": 38031816.0,
      "step": 17355
    },
    {
      "entropy": 5.615639925003052,
      "epoch": 1.722905915045653,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004707913448006433,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16562496721744538,
      "num_tokens": 38043953.0,
      "step": 17360
    },
    {
      "entropy": 5.552724123001099,
      "epoch": 1.7234021437078204,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004707738843062367,
      "loss": 5.023,
      "mean_token_accuracy": 0.18501942604780197,
      "num_tokens": 38053929.0,
      "step": 17365
    },
    {
      "entropy": 5.660346651077271,
      "epoch": 1.723898372369988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004707564189570596,
      "loss": 5.2607,
      "mean_token_accuracy": 0.17156601846218109,
      "num_tokens": 38067091.0,
      "step": 17370
    },
    {
      "entropy": 5.649452877044678,
      "epoch": 1.7243946010321556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004707389487535451,
      "loss": 5.0945,
      "mean_token_accuracy": 0.17151003777980806,
      "num_tokens": 38079002.0,
      "step": 17375
    },
    {
      "entropy": 5.640258646011352,
      "epoch": 1.7248908296943233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004707214736961265,
      "loss": 5.1165,
      "mean_token_accuracy": 0.17813628762960435,
      "num_tokens": 38089526.0,
      "step": 17380
    },
    {
      "entropy": 5.630862188339234,
      "epoch": 1.7253870583564908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004707039937852371,
      "loss": 5.1065,
      "mean_token_accuracy": 0.1795389786362648,
      "num_tokens": 38099902.0,
      "step": 17385
    },
    {
      "entropy": 5.573202276229859,
      "epoch": 1.7258832870186582,
      "grad_norm": 0.859375,
      "learning_rate": 0.00047068650902131034,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1832755535840988,
      "num_tokens": 38111725.0,
      "step": 17390
    },
    {
      "entropy": 5.602475500106811,
      "epoch": 1.7263795156808257,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004706690194047798,
      "loss": 5.0394,
      "mean_token_accuracy": 0.17570024579763413,
      "num_tokens": 38123522.0,
      "step": 17395
    },
    {
      "entropy": 5.590220642089844,
      "epoch": 1.7268757443429932,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004706515249360792,
      "loss": 5.1796,
      "mean_token_accuracy": 0.16989659368991852,
      "num_tokens": 38133968.0,
      "step": 17400
    },
    {
      "entropy": 5.581753540039062,
      "epoch": 1.7273719730051607,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004706340256156425,
      "loss": 5.2099,
      "mean_token_accuracy": 0.16452309340238572,
      "num_tokens": 38145288.0,
      "step": 17405
    },
    {
      "entropy": 5.650903606414795,
      "epoch": 1.7278682016673284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004706165214439034,
      "loss": 5.098,
      "mean_token_accuracy": 0.17576660215854645,
      "num_tokens": 38156141.0,
      "step": 17410
    },
    {
      "entropy": 5.641202735900879,
      "epoch": 1.7283644303294958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004705990124212961,
      "loss": 5.0572,
      "mean_token_accuracy": 0.18241175264120102,
      "num_tokens": 38167211.0,
      "step": 17415
    },
    {
      "entropy": 5.57245774269104,
      "epoch": 1.7288606589916635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004705814985482548,
      "loss": 5.1442,
      "mean_token_accuracy": 0.17542638629674911,
      "num_tokens": 38177360.0,
      "step": 17420
    },
    {
      "entropy": 5.580187797546387,
      "epoch": 1.729356887653831,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047056397982521366,
      "loss": 5.0457,
      "mean_token_accuracy": 0.1783628985285759,
      "num_tokens": 38187857.0,
      "step": 17425
    },
    {
      "entropy": 5.676581239700317,
      "epoch": 1.7298531163159985,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004705464562526074,
      "loss": 5.1719,
      "mean_token_accuracy": 0.16175856441259384,
      "num_tokens": 38198161.0,
      "step": 17430
    },
    {
      "entropy": 5.625281429290771,
      "epoch": 1.730349344978166,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004705289278308703,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17638128399848937,
      "num_tokens": 38208612.0,
      "step": 17435
    },
    {
      "entropy": 5.606398630142212,
      "epoch": 1.7308455736403334,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047051139456043707,
      "loss": 5.127,
      "mean_token_accuracy": 0.17148384749889373,
      "num_tokens": 38220153.0,
      "step": 17440
    },
    {
      "entropy": 5.570207595825195,
      "epoch": 1.731341802302501,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004704938564417427,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17802739441394805,
      "num_tokens": 38230790.0,
      "step": 17445
    },
    {
      "entropy": 5.61423888206482,
      "epoch": 1.7318380309646684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047047631347522175,
      "loss": 5.0499,
      "mean_token_accuracy": 0.17884979993104935,
      "num_tokens": 38242585.0,
      "step": 17450
    },
    {
      "entropy": 5.629331588745117,
      "epoch": 1.732334259626836,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047045876566130957,
      "loss": 5.1057,
      "mean_token_accuracy": 0.1774953544139862,
      "num_tokens": 38253316.0,
      "step": 17455
    },
    {
      "entropy": 5.569477462768555,
      "epoch": 1.7328304882890035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004704412130004411,
      "loss": 5.0834,
      "mean_token_accuracy": 0.18003611713647844,
      "num_tokens": 38263230.0,
      "step": 17460
    },
    {
      "entropy": 5.652558612823486,
      "epoch": 1.7333267169511712,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047042365549305174,
      "loss": 5.1426,
      "mean_token_accuracy": 0.1726859614253044,
      "num_tokens": 38273416.0,
      "step": 17465
    },
    {
      "entropy": 5.560487604141235,
      "epoch": 1.7338229456133387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004704060931395768,
      "loss": 5.0706,
      "mean_token_accuracy": 0.18012719303369523,
      "num_tokens": 38285764.0,
      "step": 17470
    },
    {
      "entropy": 5.632775735855103,
      "epoch": 1.7343191742755062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047038852594045186,
      "loss": 5.1864,
      "mean_token_accuracy": 0.17095785439014435,
      "num_tokens": 38296899.0,
      "step": 17475
    },
    {
      "entropy": 5.6769287109375,
      "epoch": 1.7348154029376737,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047037095389611245,
      "loss": 5.1746,
      "mean_token_accuracy": 0.18056826591491698,
      "num_tokens": 38309050.0,
      "step": 17480
    },
    {
      "entropy": 5.577844858169556,
      "epoch": 1.7353116315998411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004703533770069944,
      "loss": 5.1057,
      "mean_token_accuracy": 0.17621064633131028,
      "num_tokens": 38320831.0,
      "step": 17485
    },
    {
      "entropy": 5.555945777893067,
      "epoch": 1.7358078602620086,
      "grad_norm": 0.90625,
      "learning_rate": 0.00047033579527353355,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17267663180828094,
      "num_tokens": 38332096.0,
      "step": 17490
    },
    {
      "entropy": 5.643559265136719,
      "epoch": 1.7363040889241763,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004703182086961659,
      "loss": 5.1569,
      "mean_token_accuracy": 0.171490116417408,
      "num_tokens": 38343110.0,
      "step": 17495
    },
    {
      "entropy": 5.671960496902466,
      "epoch": 1.7368003175863438,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004703006172753274,
      "loss": 5.1343,
      "mean_token_accuracy": 0.17679170668125152,
      "num_tokens": 38353971.0,
      "step": 17500
    },
    {
      "entropy": 5.568271160125732,
      "epoch": 1.7372965462485113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047028302101145467,
      "loss": 5.0682,
      "mean_token_accuracy": 0.18154161870479585,
      "num_tokens": 38365114.0,
      "step": 17505
    },
    {
      "entropy": 5.619519281387329,
      "epoch": 1.737792774910679,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004702654199049836,
      "loss": 5.0978,
      "mean_token_accuracy": 0.1687868781387806,
      "num_tokens": 38375784.0,
      "step": 17510
    },
    {
      "entropy": 5.595320081710815,
      "epoch": 1.7382890035728464,
      "grad_norm": 1.046875,
      "learning_rate": 0.000470247813956351,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1751424953341484,
      "num_tokens": 38386053.0,
      "step": 17515
    },
    {
      "entropy": 5.542771816253662,
      "epoch": 1.738785232235014,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047023020316599334,
      "loss": 5.0868,
      "mean_token_accuracy": 0.17648202627897264,
      "num_tokens": 38396545.0,
      "step": 17520
    },
    {
      "entropy": 5.58643274307251,
      "epoch": 1.7392814608971814,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047021258753434725,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1759536936879158,
      "num_tokens": 38406344.0,
      "step": 17525
    },
    {
      "entropy": 5.627340078353882,
      "epoch": 1.7397776895593489,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047019496706184966,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18307138830423356,
      "num_tokens": 38416388.0,
      "step": 17530
    },
    {
      "entropy": 5.6158421516418455,
      "epoch": 1.7402739182215163,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004701773417489375,
      "loss": 5.1882,
      "mean_token_accuracy": 0.1646512195467949,
      "num_tokens": 38427608.0,
      "step": 17535
    },
    {
      "entropy": 5.6113951206207275,
      "epoch": 1.740770146883684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047015971159604787,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17145409435033798,
      "num_tokens": 38437921.0,
      "step": 17540
    },
    {
      "entropy": 5.607096529006958,
      "epoch": 1.7412663755458515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000470142076603618,
      "loss": 5.1463,
      "mean_token_accuracy": 0.17007818669080735,
      "num_tokens": 38448767.0,
      "step": 17545
    },
    {
      "entropy": 5.639474153518677,
      "epoch": 1.7417626042080192,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047012443677208496,
      "loss": 5.223,
      "mean_token_accuracy": 0.16885150820016862,
      "num_tokens": 38460052.0,
      "step": 17550
    },
    {
      "entropy": 5.600141716003418,
      "epoch": 1.7422588328701867,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004701067921018864,
      "loss": 5.0704,
      "mean_token_accuracy": 0.17633064091205597,
      "num_tokens": 38470177.0,
      "step": 17555
    },
    {
      "entropy": 5.612052917480469,
      "epoch": 1.7427550615323542,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047008914259345986,
      "loss": 5.0963,
      "mean_token_accuracy": 0.17070481330156326,
      "num_tokens": 38481608.0,
      "step": 17560
    },
    {
      "entropy": 5.61853985786438,
      "epoch": 1.7432512901945216,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047007148824724295,
      "loss": 5.1537,
      "mean_token_accuracy": 0.17066892385482788,
      "num_tokens": 38492534.0,
      "step": 17565
    },
    {
      "entropy": 5.615537118911743,
      "epoch": 1.743747518856689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004700538290636734,
      "loss": 5.1523,
      "mean_token_accuracy": 0.17136462032794952,
      "num_tokens": 38503629.0,
      "step": 17570
    },
    {
      "entropy": 5.557282495498657,
      "epoch": 1.7442437475188566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047003616504318925,
      "loss": 5.0269,
      "mean_token_accuracy": 0.17629433274269105,
      "num_tokens": 38514428.0,
      "step": 17575
    },
    {
      "entropy": 5.643941211700439,
      "epoch": 1.744739976181024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047001849618622844,
      "loss": 5.1243,
      "mean_token_accuracy": 0.1742339938879013,
      "num_tokens": 38525468.0,
      "step": 17580
    },
    {
      "entropy": 5.592883968353272,
      "epoch": 1.7452362048431918,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047000082249322916,
      "loss": 5.0857,
      "mean_token_accuracy": 0.1742246776819229,
      "num_tokens": 38535095.0,
      "step": 17585
    },
    {
      "entropy": 5.607070446014404,
      "epoch": 1.7457324335053592,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004699831439646297,
      "loss": 5.0754,
      "mean_token_accuracy": 0.1691793441772461,
      "num_tokens": 38546867.0,
      "step": 17590
    },
    {
      "entropy": 5.538869762420655,
      "epoch": 1.746228662167527,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004699654606008684,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18382130265235902,
      "num_tokens": 38557231.0,
      "step": 17595
    },
    {
      "entropy": 5.616737651824951,
      "epoch": 1.7467248908296944,
      "grad_norm": 1.0,
      "learning_rate": 0.0004699477724023838,
      "loss": 5.1335,
      "mean_token_accuracy": 0.17268901467323303,
      "num_tokens": 38568857.0,
      "step": 17600
    },
    {
      "entropy": 5.632205772399902,
      "epoch": 1.7472211194918619,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004699300793696145,
      "loss": 5.0952,
      "mean_token_accuracy": 0.17258983701467515,
      "num_tokens": 38579177.0,
      "step": 17605
    },
    {
      "entropy": 5.608993339538574,
      "epoch": 1.7477173481540293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004699123815029993,
      "loss": 5.0005,
      "mean_token_accuracy": 0.1823594570159912,
      "num_tokens": 38588683.0,
      "step": 17610
    },
    {
      "entropy": 5.554879140853882,
      "epoch": 1.7482135768161968,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046989467880297704,
      "loss": 5.0819,
      "mean_token_accuracy": 0.175696524977684,
      "num_tokens": 38599636.0,
      "step": 17615
    },
    {
      "entropy": 5.668238973617553,
      "epoch": 1.7487098054783643,
      "grad_norm": 1.125,
      "learning_rate": 0.00046987697126998673,
      "loss": 5.1882,
      "mean_token_accuracy": 0.1745127931237221,
      "num_tokens": 38609360.0,
      "step": 17620
    },
    {
      "entropy": 5.6352959156036375,
      "epoch": 1.749206034140532,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004698592589044674,
      "loss": 5.1029,
      "mean_token_accuracy": 0.17246202230453492,
      "num_tokens": 38620996.0,
      "step": 17625
    },
    {
      "entropy": 5.625862455368042,
      "epoch": 1.7497022628026995,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046984154170685833,
      "loss": 5.2295,
      "mean_token_accuracy": 0.17558062076568604,
      "num_tokens": 38632935.0,
      "step": 17630
    },
    {
      "entropy": 5.667958116531372,
      "epoch": 1.7501984914648672,
      "grad_norm": 1.0,
      "learning_rate": 0.0004698238196775989,
      "loss": 5.1891,
      "mean_token_accuracy": 0.17415012717247008,
      "num_tokens": 38644299.0,
      "step": 17635
    },
    {
      "entropy": 5.628525638580323,
      "epoch": 1.7506947201270346,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046980609281712857,
      "loss": 5.139,
      "mean_token_accuracy": 0.1750885874032974,
      "num_tokens": 38654970.0,
      "step": 17640
    },
    {
      "entropy": 5.642880868911743,
      "epoch": 1.7511909487892021,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004697883611258868,
      "loss": 5.1884,
      "mean_token_accuracy": 0.17299480885267257,
      "num_tokens": 38666302.0,
      "step": 17645
    },
    {
      "entropy": 5.628435325622559,
      "epoch": 1.7516871774513696,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004697706246043136,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17775723338127136,
      "num_tokens": 38678346.0,
      "step": 17650
    },
    {
      "entropy": 5.596378898620605,
      "epoch": 1.752183406113537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046975288325284845,
      "loss": 5.0497,
      "mean_token_accuracy": 0.174201799929142,
      "num_tokens": 38688242.0,
      "step": 17655
    },
    {
      "entropy": 5.628761339187622,
      "epoch": 1.7526796347757045,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697351370719315,
      "loss": 5.2183,
      "mean_token_accuracy": 0.16656507849693297,
      "num_tokens": 38698756.0,
      "step": 17660
    },
    {
      "entropy": 5.6043463230133055,
      "epoch": 1.753175863437872,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004697173860620027,
      "loss": 5.182,
      "mean_token_accuracy": 0.16674468517303467,
      "num_tokens": 38710410.0,
      "step": 17665
    },
    {
      "entropy": 5.684505605697632,
      "epoch": 1.7536720921000397,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004696996302235024,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17018482238054275,
      "num_tokens": 38721917.0,
      "step": 17670
    },
    {
      "entropy": 5.645163583755493,
      "epoch": 1.7541683207622072,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046968186955687074,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1768716886639595,
      "num_tokens": 38733085.0,
      "step": 17675
    },
    {
      "entropy": 5.5920390605926515,
      "epoch": 1.7546645494243749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004696641040625483,
      "loss": 5.1776,
      "mean_token_accuracy": 0.17255741059780122,
      "num_tokens": 38744785.0,
      "step": 17680
    },
    {
      "entropy": 5.613038158416748,
      "epoch": 1.7551607780865424,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00046964633374097544,
      "loss": 5.1414,
      "mean_token_accuracy": 0.17650899887084961,
      "num_tokens": 38755626.0,
      "step": 17685
    },
    {
      "entropy": 5.631186151504517,
      "epoch": 1.7556570067487098,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046962855859259306,
      "loss": 5.0333,
      "mean_token_accuracy": 0.17535644322633742,
      "num_tokens": 38766360.0,
      "step": 17690
    },
    {
      "entropy": 5.603131103515625,
      "epoch": 1.7561532354108773,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046961077861784167,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17513916045427322,
      "num_tokens": 38776763.0,
      "step": 17695
    },
    {
      "entropy": 5.56859188079834,
      "epoch": 1.7566494640730448,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004695929938171624,
      "loss": 5.1612,
      "mean_token_accuracy": 0.17393120676279067,
      "num_tokens": 38787664.0,
      "step": 17700
    },
    {
      "entropy": 5.620319128036499,
      "epoch": 1.7571456927352123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004695752041909962,
      "loss": 5.0913,
      "mean_token_accuracy": 0.17367804050445557,
      "num_tokens": 38798018.0,
      "step": 17705
    },
    {
      "entropy": 5.601537036895752,
      "epoch": 1.75764192139738,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004695574097397842,
      "loss": 5.0533,
      "mean_token_accuracy": 0.1780431389808655,
      "num_tokens": 38810381.0,
      "step": 17710
    },
    {
      "entropy": 5.52885479927063,
      "epoch": 1.7581381500595474,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046953961046396767,
      "loss": 5.0679,
      "mean_token_accuracy": 0.18244463205337524,
      "num_tokens": 38821042.0,
      "step": 17715
    },
    {
      "entropy": 5.542054224014282,
      "epoch": 1.7586343787217151,
      "grad_norm": 1.125,
      "learning_rate": 0.00046952180636398793,
      "loss": 5.023,
      "mean_token_accuracy": 0.17878574579954148,
      "num_tokens": 38830890.0,
      "step": 17720
    },
    {
      "entropy": 5.632383918762207,
      "epoch": 1.7591306073838826,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004695039974402866,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18040262460708617,
      "num_tokens": 38842648.0,
      "step": 17725
    },
    {
      "entropy": 5.615215539932251,
      "epoch": 1.75962683604605,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046948618369330533,
      "loss": 5.0095,
      "mean_token_accuracy": 0.1844614788889885,
      "num_tokens": 38852740.0,
      "step": 17730
    },
    {
      "entropy": 5.612536191940308,
      "epoch": 1.7601230647082176,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046946836512348566,
      "loss": 5.1581,
      "mean_token_accuracy": 0.16656874716281891,
      "num_tokens": 38863371.0,
      "step": 17735
    },
    {
      "entropy": 5.651531934738159,
      "epoch": 1.760619293370385,
      "grad_norm": 1.0,
      "learning_rate": 0.00046945054173126965,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17092522084712983,
      "num_tokens": 38873893.0,
      "step": 17740
    },
    {
      "entropy": 5.637461519241333,
      "epoch": 1.7611155220325525,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046943271351709923,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1721355840563774,
      "num_tokens": 38884778.0,
      "step": 17745
    },
    {
      "entropy": 5.5364155769348145,
      "epoch": 1.76161175069472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004694148804814164,
      "loss": 5.1134,
      "mean_token_accuracy": 0.16912716031074523,
      "num_tokens": 38896011.0,
      "step": 17750
    },
    {
      "entropy": 5.617780017852783,
      "epoch": 1.7621079793568877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004693970426246635,
      "loss": 5.1618,
      "mean_token_accuracy": 0.1699749082326889,
      "num_tokens": 38907686.0,
      "step": 17755
    },
    {
      "entropy": 5.57259087562561,
      "epoch": 1.7626042080190552,
      "grad_norm": 0.90625,
      "learning_rate": 0.00046937919994728287,
      "loss": 5.1089,
      "mean_token_accuracy": 0.1711419105529785,
      "num_tokens": 38919180.0,
      "step": 17760
    },
    {
      "entropy": 5.600240421295166,
      "epoch": 1.7631004366812228,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004693613524497168,
      "loss": 5.0862,
      "mean_token_accuracy": 0.1778033643960953,
      "num_tokens": 38929600.0,
      "step": 17765
    },
    {
      "entropy": 5.593649053573609,
      "epoch": 1.7635966653433903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046934350013240814,
      "loss": 5.0335,
      "mean_token_accuracy": 0.18278988748788833,
      "num_tokens": 38940404.0,
      "step": 17770
    },
    {
      "entropy": 5.608020830154419,
      "epoch": 1.7640928940055578,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004693256429957994,
      "loss": 5.097,
      "mean_token_accuracy": 0.17356131225824356,
      "num_tokens": 38951929.0,
      "step": 17775
    },
    {
      "entropy": 5.546957969665527,
      "epoch": 1.7645891226677253,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004693077810403335,
      "loss": 5.0842,
      "mean_token_accuracy": 0.1803742080926895,
      "num_tokens": 38961954.0,
      "step": 17780
    },
    {
      "entropy": 5.53098611831665,
      "epoch": 1.7650853513298927,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046928991426645325,
      "loss": 5.0551,
      "mean_token_accuracy": 0.18028428852558137,
      "num_tokens": 38970974.0,
      "step": 17785
    },
    {
      "entropy": 5.613315677642822,
      "epoch": 1.7655815799920602,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046927204267460177,
      "loss": 5.1437,
      "mean_token_accuracy": 0.16925946921110152,
      "num_tokens": 38982481.0,
      "step": 17790
    },
    {
      "entropy": 5.645141696929931,
      "epoch": 1.766077808654228,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046925416626522235,
      "loss": 5.0982,
      "mean_token_accuracy": 0.176967054605484,
      "num_tokens": 38993279.0,
      "step": 17795
    },
    {
      "entropy": 5.592154264450073,
      "epoch": 1.7665740373163954,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004692362850387581,
      "loss": 5.0831,
      "mean_token_accuracy": 0.1789329394698143,
      "num_tokens": 39003748.0,
      "step": 17800
    },
    {
      "entropy": 5.552034711837768,
      "epoch": 1.7670702659785629,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046921839899565263,
      "loss": 5.1132,
      "mean_token_accuracy": 0.1666934981942177,
      "num_tokens": 39014902.0,
      "step": 17805
    },
    {
      "entropy": 5.624300813674926,
      "epoch": 1.7675664946407306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004692005081363493,
      "loss": 5.1709,
      "mean_token_accuracy": 0.16976969689130783,
      "num_tokens": 39026484.0,
      "step": 17810
    },
    {
      "entropy": 5.586310243606567,
      "epoch": 1.768062723302898,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046918261246129176,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17041055411100386,
      "num_tokens": 39036268.0,
      "step": 17815
    },
    {
      "entropy": 5.633852624893189,
      "epoch": 1.7685589519650655,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000469164711970924,
      "loss": 5.1979,
      "mean_token_accuracy": 0.17508595436811447,
      "num_tokens": 39049092.0,
      "step": 17820
    },
    {
      "entropy": 5.6601309299469,
      "epoch": 1.769055180627233,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004691468066656896,
      "loss": 5.1315,
      "mean_token_accuracy": 0.1738499656319618,
      "num_tokens": 39059255.0,
      "step": 17825
    },
    {
      "entropy": 5.5493621826171875,
      "epoch": 1.7695514092894005,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046912889654603286,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17702848464250565,
      "num_tokens": 39069913.0,
      "step": 17830
    },
    {
      "entropy": 5.613896703720092,
      "epoch": 1.770047637951568,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004691109816123977,
      "loss": 5.0788,
      "mean_token_accuracy": 0.17963456660509108,
      "num_tokens": 39080666.0,
      "step": 17835
    },
    {
      "entropy": 5.673900318145752,
      "epoch": 1.7705438666137356,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004690930618652286,
      "loss": 5.146,
      "mean_token_accuracy": 0.1744784712791443,
      "num_tokens": 39091562.0,
      "step": 17840
    },
    {
      "entropy": 5.6436944007873535,
      "epoch": 1.7710400952759031,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046907513730496975,
      "loss": 5.219,
      "mean_token_accuracy": 0.16445765197277068,
      "num_tokens": 39102567.0,
      "step": 17845
    },
    {
      "entropy": 5.658611917495728,
      "epoch": 1.7715363239380708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004690572079320657,
      "loss": 5.1116,
      "mean_token_accuracy": 0.17458836138248443,
      "num_tokens": 39113260.0,
      "step": 17850
    },
    {
      "entropy": 5.604324674606323,
      "epoch": 1.7720325526002383,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046903927374696106,
      "loss": 5.1361,
      "mean_token_accuracy": 0.17303506880998612,
      "num_tokens": 39125790.0,
      "step": 17855
    },
    {
      "entropy": 5.599290227890014,
      "epoch": 1.7725287812624058,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004690213347501005,
      "loss": 5.1953,
      "mean_token_accuracy": 0.16766995936632156,
      "num_tokens": 39137220.0,
      "step": 17860
    },
    {
      "entropy": 5.620479917526245,
      "epoch": 1.7730250099245732,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004690033909419289,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1814360111951828,
      "num_tokens": 39146362.0,
      "step": 17865
    },
    {
      "entropy": 5.622931623458863,
      "epoch": 1.7735212385867407,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046898544232289126,
      "loss": 5.1436,
      "mean_token_accuracy": 0.1732465922832489,
      "num_tokens": 39156822.0,
      "step": 17870
    },
    {
      "entropy": 5.635919857025146,
      "epoch": 1.7740174672489082,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046896748889343267,
      "loss": 5.1218,
      "mean_token_accuracy": 0.17051680088043214,
      "num_tokens": 39166693.0,
      "step": 17875
    },
    {
      "entropy": 5.552589321136475,
      "epoch": 1.7745136959110757,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046894953065399836,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17539652138948442,
      "num_tokens": 39176889.0,
      "step": 17880
    },
    {
      "entropy": 5.624076509475708,
      "epoch": 1.7750099245732434,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046893156760503357,
      "loss": 5.1008,
      "mean_token_accuracy": 0.1821555808186531,
      "num_tokens": 39188077.0,
      "step": 17885
    },
    {
      "entropy": 5.556470251083374,
      "epoch": 1.7755061532354108,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004689135997469838,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17249607294797897,
      "num_tokens": 39198182.0,
      "step": 17890
    },
    {
      "entropy": 5.620037603378296,
      "epoch": 1.7760023818975785,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004688956270802945,
      "loss": 5.0638,
      "mean_token_accuracy": 0.17689334005117416,
      "num_tokens": 39208801.0,
      "step": 17895
    },
    {
      "entropy": 5.481461858749389,
      "epoch": 1.776498610559746,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004688776496054117,
      "loss": 4.9398,
      "mean_token_accuracy": 0.18704006522893907,
      "num_tokens": 39220658.0,
      "step": 17900
    },
    {
      "entropy": 5.581477212905884,
      "epoch": 1.7769948392219135,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046885966732278077,
      "loss": 5.1842,
      "mean_token_accuracy": 0.17020562738180162,
      "num_tokens": 39231987.0,
      "step": 17905
    },
    {
      "entropy": 5.630270957946777,
      "epoch": 1.777491067884081,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046884168023284787,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1729697436094284,
      "num_tokens": 39242797.0,
      "step": 17910
    },
    {
      "entropy": 5.637810468673706,
      "epoch": 1.7779872965462484,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046882368833605894,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17558135688304902,
      "num_tokens": 39253621.0,
      "step": 17915
    },
    {
      "entropy": 5.550224018096924,
      "epoch": 1.778483525208416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046880569163286023,
      "loss": 5.1495,
      "mean_token_accuracy": 0.16964975893497466,
      "num_tokens": 39264126.0,
      "step": 17920
    },
    {
      "entropy": 5.592004108428955,
      "epoch": 1.7789797538705836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046878769012369807,
      "loss": 5.0735,
      "mean_token_accuracy": 0.18172547817230225,
      "num_tokens": 39275118.0,
      "step": 17925
    },
    {
      "entropy": 5.628892469406128,
      "epoch": 1.779475982532751,
      "grad_norm": 1.0,
      "learning_rate": 0.00046876968380901864,
      "loss": 5.1001,
      "mean_token_accuracy": 0.17315933108329773,
      "num_tokens": 39287237.0,
      "step": 17930
    },
    {
      "entropy": 5.662260103225708,
      "epoch": 1.7799722111949188,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046875167268926865,
      "loss": 5.1507,
      "mean_token_accuracy": 0.17398077845573426,
      "num_tokens": 39296995.0,
      "step": 17935
    },
    {
      "entropy": 5.579225730895996,
      "epoch": 1.7804684398570862,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004687336567648947,
      "loss": 5.0378,
      "mean_token_accuracy": 0.17514205127954482,
      "num_tokens": 39305980.0,
      "step": 17940
    },
    {
      "entropy": 5.639255094528198,
      "epoch": 1.7809646685192537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046871563603634347,
      "loss": 5.173,
      "mean_token_accuracy": 0.1633278414607048,
      "num_tokens": 39316494.0,
      "step": 17945
    },
    {
      "entropy": 5.625512647628784,
      "epoch": 1.7814608971814212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004686976105040619,
      "loss": 5.1431,
      "mean_token_accuracy": 0.1756330609321594,
      "num_tokens": 39327699.0,
      "step": 17950
    },
    {
      "entropy": 5.638405466079712,
      "epoch": 1.7819571258435887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004686795801684969,
      "loss": 5.1532,
      "mean_token_accuracy": 0.17359215021133423,
      "num_tokens": 39338749.0,
      "step": 17955
    },
    {
      "entropy": 5.580149698257446,
      "epoch": 1.7824533545057561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004686615450300958,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17506850510835648,
      "num_tokens": 39349747.0,
      "step": 17960
    },
    {
      "entropy": 5.529717826843262,
      "epoch": 1.7829495831679236,
      "grad_norm": 1.125,
      "learning_rate": 0.0004686435050893056,
      "loss": 4.9951,
      "mean_token_accuracy": 0.1821524128317833,
      "num_tokens": 39360068.0,
      "step": 17965
    },
    {
      "entropy": 5.545273733139038,
      "epoch": 1.7834458118300913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046862546034657376,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18017940521240233,
      "num_tokens": 39369651.0,
      "step": 17970
    },
    {
      "entropy": 5.622182416915893,
      "epoch": 1.7839420404922588,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004686074108023477,
      "loss": 5.2279,
      "mean_token_accuracy": 0.17189501821994782,
      "num_tokens": 39381354.0,
      "step": 17975
    },
    {
      "entropy": 5.61877760887146,
      "epoch": 1.7844382691544265,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046858935645707506,
      "loss": 5.0845,
      "mean_token_accuracy": 0.17493598610162736,
      "num_tokens": 39391793.0,
      "step": 17980
    },
    {
      "entropy": 5.587228870391845,
      "epoch": 1.784934497816594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004685712973112035,
      "loss": 5.0872,
      "mean_token_accuracy": 0.1767203375697136,
      "num_tokens": 39402194.0,
      "step": 17985
    },
    {
      "entropy": 5.578391790390015,
      "epoch": 1.7854307264787614,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004685532333651809,
      "loss": 5.1063,
      "mean_token_accuracy": 0.17596443146467208,
      "num_tokens": 39413193.0,
      "step": 17990
    },
    {
      "entropy": 5.558865022659302,
      "epoch": 1.785926955140929,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046853516461945523,
      "loss": 5.0818,
      "mean_token_accuracy": 0.17787336558103561,
      "num_tokens": 39423072.0,
      "step": 17995
    },
    {
      "entropy": 5.607248592376709,
      "epoch": 1.7864231838030964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046851709107447446,
      "loss": 5.1866,
      "mean_token_accuracy": 0.16660024374723434,
      "num_tokens": 39435088.0,
      "step": 18000
    },
    {
      "epoch": 1.7864231838030964,
      "eval_entropy": 5.412690199024747,
      "eval_loss": 5.212555885314941,
      "eval_mean_token_accuracy": 0.17520355190802964,
      "eval_num_tokens": 39435088.0,
      "eval_runtime": 20.6145,
      "eval_samples_per_second": 1633.27,
      "eval_steps_per_second": 204.177,
      "step": 18000
    },
    {
      "entropy": 5.574995136260986,
      "epoch": 1.7869194124652639,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004684990127306868,
      "loss": 5.0561,
      "mean_token_accuracy": 0.177810437977314,
      "num_tokens": 39446129.0,
      "step": 18005
    },
    {
      "entropy": 5.563825798034668,
      "epoch": 1.7874156411274316,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046848092958854065,
      "loss": 5.0294,
      "mean_token_accuracy": 0.1909279152750969,
      "num_tokens": 39458127.0,
      "step": 18010
    },
    {
      "entropy": 5.591496849060059,
      "epoch": 1.787911869789599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684628416484843,
      "loss": 5.0717,
      "mean_token_accuracy": 0.18210589289665222,
      "num_tokens": 39468204.0,
      "step": 18015
    },
    {
      "entropy": 5.591556692123413,
      "epoch": 1.7884080984517667,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004684447489109663,
      "loss": 5.1282,
      "mean_token_accuracy": 0.17167278677225112,
      "num_tokens": 39479114.0,
      "step": 18020
    },
    {
      "entropy": 5.62839789390564,
      "epoch": 1.7889043271139342,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046842665137643546,
      "loss": 5.1084,
      "mean_token_accuracy": 0.1768855556845665,
      "num_tokens": 39489691.0,
      "step": 18025
    },
    {
      "entropy": 5.6490621089935305,
      "epoch": 1.7894005557761017,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046840854904534046,
      "loss": 5.2214,
      "mean_token_accuracy": 0.16355034559965134,
      "num_tokens": 39501879.0,
      "step": 18030
    },
    {
      "entropy": 5.639403581619263,
      "epoch": 1.7898967844382692,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004683904419181302,
      "loss": 5.1426,
      "mean_token_accuracy": 0.17735382914543152,
      "num_tokens": 39513413.0,
      "step": 18035
    },
    {
      "entropy": 5.626006269454956,
      "epoch": 1.7903930131004366,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046837232999525366,
      "loss": 5.0722,
      "mean_token_accuracy": 0.17909554839134217,
      "num_tokens": 39522719.0,
      "step": 18040
    },
    {
      "entropy": 5.618683481216431,
      "epoch": 1.7908892417626041,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046835421327716,
      "loss": 5.1425,
      "mean_token_accuracy": 0.17290786653757095,
      "num_tokens": 39532463.0,
      "step": 18045
    },
    {
      "entropy": 5.603424692153931,
      "epoch": 1.7913854704247716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004683360917642986,
      "loss": 5.1065,
      "mean_token_accuracy": 0.16600829511880874,
      "num_tokens": 39543005.0,
      "step": 18050
    },
    {
      "entropy": 5.666201448440551,
      "epoch": 1.7918816990869393,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046831796545711865,
      "loss": 5.1651,
      "mean_token_accuracy": 0.1687757596373558,
      "num_tokens": 39553468.0,
      "step": 18055
    },
    {
      "entropy": 5.679385566711426,
      "epoch": 1.7923779277491068,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046829983435606973,
      "loss": 5.2169,
      "mean_token_accuracy": 0.17306623011827468,
      "num_tokens": 39563539.0,
      "step": 18060
    },
    {
      "entropy": 5.623307991027832,
      "epoch": 1.7928741564112745,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004682816984616014,
      "loss": 5.1524,
      "mean_token_accuracy": 0.17145076543092727,
      "num_tokens": 39574835.0,
      "step": 18065
    },
    {
      "entropy": 5.645779418945312,
      "epoch": 1.793370385073442,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046826355777416346,
      "loss": 5.1143,
      "mean_token_accuracy": 0.17912749350070953,
      "num_tokens": 39586021.0,
      "step": 18070
    },
    {
      "entropy": 5.636172723770142,
      "epoch": 1.7938666137356094,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046824541229420583,
      "loss": 5.0464,
      "mean_token_accuracy": 0.17641086131334305,
      "num_tokens": 39596333.0,
      "step": 18075
    },
    {
      "entropy": 5.57766318321228,
      "epoch": 1.7943628423977769,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046822726202217825,
      "loss": 5.0807,
      "mean_token_accuracy": 0.1752803847193718,
      "num_tokens": 39607553.0,
      "step": 18080
    },
    {
      "entropy": 5.589487552642822,
      "epoch": 1.7948590710599444,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000468209106958531,
      "loss": 5.1636,
      "mean_token_accuracy": 0.1717144653201103,
      "num_tokens": 39619700.0,
      "step": 18085
    },
    {
      "entropy": 5.603281307220459,
      "epoch": 1.7953552997221118,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004681909471037142,
      "loss": 5.0603,
      "mean_token_accuracy": 0.17870819121599196,
      "num_tokens": 39630623.0,
      "step": 18090
    },
    {
      "entropy": 5.620237970352173,
      "epoch": 1.7958515283842795,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046817278245817815,
      "loss": 5.0694,
      "mean_token_accuracy": 0.1808307349681854,
      "num_tokens": 39641089.0,
      "step": 18095
    },
    {
      "entropy": 5.588995552062988,
      "epoch": 1.796347757046447,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681546130223735,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17471811324357986,
      "num_tokens": 39652381.0,
      "step": 18100
    },
    {
      "entropy": 5.576749658584594,
      "epoch": 1.7968439857086145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004681364387967505,
      "loss": 5.0527,
      "mean_token_accuracy": 0.1774971142411232,
      "num_tokens": 39662536.0,
      "step": 18105
    },
    {
      "entropy": 5.609192895889282,
      "epoch": 1.7973402143707822,
      "grad_norm": 1.0,
      "learning_rate": 0.0004681182597817601,
      "loss": 5.0975,
      "mean_token_accuracy": 0.1807541534304619,
      "num_tokens": 39673140.0,
      "step": 18110
    },
    {
      "entropy": 5.584336137771606,
      "epoch": 1.7978364430329496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004681000759778529,
      "loss": 5.106,
      "mean_token_accuracy": 0.16941312551498414,
      "num_tokens": 39683843.0,
      "step": 18115
    },
    {
      "entropy": 5.603990745544434,
      "epoch": 1.7983326716951171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004680818873854799,
      "loss": 5.0538,
      "mean_token_accuracy": 0.17753756940364837,
      "num_tokens": 39694178.0,
      "step": 18120
    },
    {
      "entropy": 5.519855642318726,
      "epoch": 1.7988289003572846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004680636940050922,
      "loss": 5.0557,
      "mean_token_accuracy": 0.18005267679691314,
      "num_tokens": 39704818.0,
      "step": 18125
    },
    {
      "entropy": 5.585251379013061,
      "epoch": 1.799325129019452,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000468045495837141,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17171570807695388,
      "num_tokens": 39715599.0,
      "step": 18130
    },
    {
      "entropy": 5.634206628799438,
      "epoch": 1.7998213576816195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046802729288207735,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17403510808944703,
      "num_tokens": 39725736.0,
      "step": 18135
    },
    {
      "entropy": 5.5933106422424315,
      "epoch": 1.8003175863437872,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046800908514035285,
      "loss": 5.13,
      "mean_token_accuracy": 0.17251380681991577,
      "num_tokens": 39735736.0,
      "step": 18140
    },
    {
      "entropy": 5.615758895874023,
      "epoch": 1.8008138150059547,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046799087261241884,
      "loss": 5.1593,
      "mean_token_accuracy": 0.17298026084899903,
      "num_tokens": 39747443.0,
      "step": 18145
    },
    {
      "entropy": 5.625498676300049,
      "epoch": 1.8013100436681224,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004679726552987271,
      "loss": 5.0991,
      "mean_token_accuracy": 0.1725580006837845,
      "num_tokens": 39758128.0,
      "step": 18150
    },
    {
      "entropy": 5.602609014511108,
      "epoch": 1.80180627233029,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004679544331997294,
      "loss": 5.0228,
      "mean_token_accuracy": 0.18337812423706054,
      "num_tokens": 39769153.0,
      "step": 18155
    },
    {
      "entropy": 5.603472471237183,
      "epoch": 1.8023025009924574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046793620631587743,
      "loss": 5.1224,
      "mean_token_accuracy": 0.16816356480121614,
      "num_tokens": 39779666.0,
      "step": 18160
    },
    {
      "entropy": 5.602590036392212,
      "epoch": 1.8027987296546248,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046791797464762334,
      "loss": 5.141,
      "mean_token_accuracy": 0.17023623287677764,
      "num_tokens": 39790329.0,
      "step": 18165
    },
    {
      "entropy": 5.663318204879761,
      "epoch": 1.8032949583167923,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046789973819541915,
      "loss": 5.1329,
      "mean_token_accuracy": 0.1701362207531929,
      "num_tokens": 39801996.0,
      "step": 18170
    },
    {
      "entropy": 5.6653941631317135,
      "epoch": 1.8037911869789598,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004678814969597171,
      "loss": 5.1055,
      "mean_token_accuracy": 0.17387129962444306,
      "num_tokens": 39814064.0,
      "step": 18175
    },
    {
      "entropy": 5.601706218719483,
      "epoch": 1.8042874156411273,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004678632509409696,
      "loss": 5.0673,
      "mean_token_accuracy": 0.17565013468265533,
      "num_tokens": 39825369.0,
      "step": 18180
    },
    {
      "entropy": 5.5327857494354244,
      "epoch": 1.804783644303295,
      "grad_norm": 1.046875,
      "learning_rate": 0.000467845000139629,
      "loss": 5.1565,
      "mean_token_accuracy": 0.17838125675916672,
      "num_tokens": 39835218.0,
      "step": 18185
    },
    {
      "entropy": 5.619577312469483,
      "epoch": 1.8052798729654624,
      "grad_norm": 1.09375,
      "learning_rate": 0.000467826744556148,
      "loss": 5.173,
      "mean_token_accuracy": 0.17046052664518357,
      "num_tokens": 39844706.0,
      "step": 18190
    },
    {
      "entropy": 5.54996771812439,
      "epoch": 1.8057761016276301,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004678084841909792,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17758762389421462,
      "num_tokens": 39856348.0,
      "step": 18195
    },
    {
      "entropy": 5.578570175170898,
      "epoch": 1.8062723302897976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046779021904457555,
      "loss": 5.1389,
      "mean_token_accuracy": 0.173824942111969,
      "num_tokens": 39866850.0,
      "step": 18200
    },
    {
      "entropy": 5.658380794525146,
      "epoch": 1.806768558951965,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004677719491173898,
      "loss": 5.1616,
      "mean_token_accuracy": 0.1736438527703285,
      "num_tokens": 39877647.0,
      "step": 18205
    },
    {
      "entropy": 5.631809520721435,
      "epoch": 1.8072647876141326,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046775367440987506,
      "loss": 5.2004,
      "mean_token_accuracy": 0.1726662203669548,
      "num_tokens": 39889172.0,
      "step": 18210
    },
    {
      "entropy": 5.620269155502319,
      "epoch": 1.8077610162763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004677353949224846,
      "loss": 5.1017,
      "mean_token_accuracy": 0.17799613177776336,
      "num_tokens": 39900445.0,
      "step": 18215
    },
    {
      "entropy": 5.623287343978882,
      "epoch": 1.8082572449384675,
      "grad_norm": 1.0,
      "learning_rate": 0.00046771711065567167,
      "loss": 5.1258,
      "mean_token_accuracy": 0.1748728036880493,
      "num_tokens": 39912825.0,
      "step": 18220
    },
    {
      "entropy": 5.5520875453948975,
      "epoch": 1.8087534736006352,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004676988216098896,
      "loss": 5.0724,
      "mean_token_accuracy": 0.18135943561792373,
      "num_tokens": 39923370.0,
      "step": 18225
    },
    {
      "entropy": 5.6612917423248295,
      "epoch": 1.8092497022628027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004676805277855921,
      "loss": 5.1447,
      "mean_token_accuracy": 0.17431560456752776,
      "num_tokens": 39934724.0,
      "step": 18230
    },
    {
      "entropy": 5.582058715820312,
      "epoch": 1.8097459309249704,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004676622291832326,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17324392050504683,
      "num_tokens": 39947064.0,
      "step": 18235
    },
    {
      "entropy": 5.5807195663452145,
      "epoch": 1.8102421595871379,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046764392580326506,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17764058113098144,
      "num_tokens": 39957649.0,
      "step": 18240
    },
    {
      "entropy": 5.578929901123047,
      "epoch": 1.8107383882493053,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046762561764614316,
      "loss": 5.1187,
      "mean_token_accuracy": 0.17154781371355057,
      "num_tokens": 39968086.0,
      "step": 18245
    },
    {
      "entropy": 5.602425670623779,
      "epoch": 1.8112346169114728,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000467607304712321,
      "loss": 5.0615,
      "mean_token_accuracy": 0.18002657741308212,
      "num_tokens": 39978794.0,
      "step": 18250
    },
    {
      "entropy": 5.62895884513855,
      "epoch": 1.8117308455736403,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046758898700225283,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1778009995818138,
      "num_tokens": 39990383.0,
      "step": 18255
    },
    {
      "entropy": 5.622438955307007,
      "epoch": 1.8122270742358078,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004675706645163927,
      "loss": 5.1624,
      "mean_token_accuracy": 0.16997318118810653,
      "num_tokens": 40001900.0,
      "step": 18260
    },
    {
      "entropy": 5.6350281715393065,
      "epoch": 1.8127233028979752,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467552337255195,
      "loss": 5.0772,
      "mean_token_accuracy": 0.1731880262494087,
      "num_tokens": 40012482.0,
      "step": 18265
    },
    {
      "entropy": 5.550037479400634,
      "epoch": 1.813219531560143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004675340052191144,
      "loss": 5.11,
      "mean_token_accuracy": 0.17436347156763077,
      "num_tokens": 40023264.0,
      "step": 18270
    },
    {
      "entropy": 5.634382438659668,
      "epoch": 1.8137157602223104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004675156684086052,
      "loss": 5.1584,
      "mean_token_accuracy": 0.17435586899518968,
      "num_tokens": 40034754.0,
      "step": 18275
    },
    {
      "entropy": 5.591562700271607,
      "epoch": 1.814211988884478,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004674973268241224,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18005286604166032,
      "num_tokens": 40045781.0,
      "step": 18280
    },
    {
      "entropy": 5.636264657974243,
      "epoch": 1.8147082175466456,
      "grad_norm": 1.125,
      "learning_rate": 0.0004674789804661205,
      "loss": 5.1579,
      "mean_token_accuracy": 0.1704658344388008,
      "num_tokens": 40055700.0,
      "step": 18285
    },
    {
      "entropy": 5.644700908660889,
      "epoch": 1.815204446208813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046746062933505477,
      "loss": 5.1854,
      "mean_token_accuracy": 0.16776805818080903,
      "num_tokens": 40066702.0,
      "step": 18290
    },
    {
      "entropy": 5.637765979766845,
      "epoch": 1.8157006748709805,
      "grad_norm": 1.0,
      "learning_rate": 0.0004674422734313801,
      "loss": 5.175,
      "mean_token_accuracy": 0.1765969455242157,
      "num_tokens": 40077856.0,
      "step": 18295
    },
    {
      "entropy": 5.542260360717774,
      "epoch": 1.816196903533148,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046742391275555176,
      "loss": 5.0651,
      "mean_token_accuracy": 0.18019032925367356,
      "num_tokens": 40089205.0,
      "step": 18300
    },
    {
      "entropy": 5.638839387893677,
      "epoch": 1.8166931321953155,
      "grad_norm": 0.90625,
      "learning_rate": 0.000467405547308025,
      "loss": 5.1125,
      "mean_token_accuracy": 0.1710200011730194,
      "num_tokens": 40100028.0,
      "step": 18305
    },
    {
      "entropy": 5.695910215377808,
      "epoch": 1.8171893608574832,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004673871770892553,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17260389477014543,
      "num_tokens": 40109635.0,
      "step": 18310
    },
    {
      "entropy": 5.657359981536866,
      "epoch": 1.8176855895196506,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046736880209969814,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17337968349456787,
      "num_tokens": 40120539.0,
      "step": 18315
    },
    {
      "entropy": 5.6284486770629885,
      "epoch": 1.8181818181818183,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046735042233980923,
      "loss": 5.1621,
      "mean_token_accuracy": 0.16550198793411255,
      "num_tokens": 40131738.0,
      "step": 18320
    },
    {
      "entropy": 5.704628038406372,
      "epoch": 1.8186780468439858,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046733203781004426,
      "loss": 5.2194,
      "mean_token_accuracy": 0.1761952355504036,
      "num_tokens": 40142800.0,
      "step": 18325
    },
    {
      "entropy": 5.701707267761231,
      "epoch": 1.8191742755061533,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046731364851085923,
      "loss": 5.1762,
      "mean_token_accuracy": 0.16931460797786713,
      "num_tokens": 40153992.0,
      "step": 18330
    },
    {
      "entropy": 5.609345865249634,
      "epoch": 1.8196705041683208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004672952544427101,
      "loss": 5.1044,
      "mean_token_accuracy": 0.17643154710531234,
      "num_tokens": 40165213.0,
      "step": 18335
    },
    {
      "entropy": 5.600104951858521,
      "epoch": 1.8201667328304882,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046727685560605305,
      "loss": 5.1616,
      "mean_token_accuracy": 0.17244796454906464,
      "num_tokens": 40176763.0,
      "step": 18340
    },
    {
      "entropy": 5.614071941375732,
      "epoch": 1.8206629614926557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004672584520013443,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17457992434501649,
      "num_tokens": 40187958.0,
      "step": 18345
    },
    {
      "entropy": 5.574062395095825,
      "epoch": 1.8211591901548232,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004672400436290402,
      "loss": 5.079,
      "mean_token_accuracy": 0.177436363697052,
      "num_tokens": 40197881.0,
      "step": 18350
    },
    {
      "entropy": 5.640538120269776,
      "epoch": 1.821655418816991,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046722163048959725,
      "loss": 5.0548,
      "mean_token_accuracy": 0.18434662669897078,
      "num_tokens": 40209322.0,
      "step": 18355
    },
    {
      "entropy": 5.626106834411621,
      "epoch": 1.8221516474791584,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004672032125834721,
      "loss": 5.1127,
      "mean_token_accuracy": 0.17721065431833266,
      "num_tokens": 40222186.0,
      "step": 18360
    },
    {
      "entropy": 5.649416255950928,
      "epoch": 1.822647876141326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004671847899111214,
      "loss": 5.1415,
      "mean_token_accuracy": 0.1714126467704773,
      "num_tokens": 40232136.0,
      "step": 18365
    },
    {
      "entropy": 5.706641817092896,
      "epoch": 1.8231441048034935,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046716636247300206,
      "loss": 5.1721,
      "mean_token_accuracy": 0.1712469920516014,
      "num_tokens": 40244468.0,
      "step": 18370
    },
    {
      "entropy": 5.649786520004272,
      "epoch": 1.823640333465661,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046714793026957096,
      "loss": 5.1387,
      "mean_token_accuracy": 0.17171223163604737,
      "num_tokens": 40255441.0,
      "step": 18375
    },
    {
      "entropy": 5.564567947387696,
      "epoch": 1.8241365621278285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004671294933012853,
      "loss": 5.052,
      "mean_token_accuracy": 0.17345405071973802,
      "num_tokens": 40266109.0,
      "step": 18380
    },
    {
      "entropy": 5.5746422290802,
      "epoch": 1.824632790789996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046711105156860223,
      "loss": 5.0657,
      "mean_token_accuracy": 0.17334969937801362,
      "num_tokens": 40277635.0,
      "step": 18385
    },
    {
      "entropy": 5.6212685108184814,
      "epoch": 1.8251290194521634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046709260507197896,
      "loss": 5.2092,
      "mean_token_accuracy": 0.17209556698799133,
      "num_tokens": 40288668.0,
      "step": 18390
    },
    {
      "entropy": 5.605411148071289,
      "epoch": 1.8256252481143311,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046707415381187306,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17697148621082306,
      "num_tokens": 40300262.0,
      "step": 18395
    },
    {
      "entropy": 5.627186489105225,
      "epoch": 1.8261214767764986,
      "grad_norm": 0.90625,
      "learning_rate": 0.000467055697788742,
      "loss": 5.1486,
      "mean_token_accuracy": 0.16308642476797103,
      "num_tokens": 40310931.0,
      "step": 18400
    },
    {
      "entropy": 5.591327238082886,
      "epoch": 1.826617705438666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046703723700304345,
      "loss": 5.0793,
      "mean_token_accuracy": 0.17756234854459763,
      "num_tokens": 40321623.0,
      "step": 18405
    },
    {
      "entropy": 5.568545866012573,
      "epoch": 1.8271139341008338,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004670187714552353,
      "loss": 5.0506,
      "mean_token_accuracy": 0.17478836327791214,
      "num_tokens": 40330951.0,
      "step": 18410
    },
    {
      "entropy": 5.606765604019165,
      "epoch": 1.8276101627630013,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046700030114577533,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1745282530784607,
      "num_tokens": 40342971.0,
      "step": 18415
    },
    {
      "entropy": 5.625487613677978,
      "epoch": 1.8281063914251687,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004669818260751217,
      "loss": 5.0994,
      "mean_token_accuracy": 0.1784167006611824,
      "num_tokens": 40353837.0,
      "step": 18420
    },
    {
      "entropy": 5.630926990509034,
      "epoch": 1.8286026200873362,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004669633462437324,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17491415590047837,
      "num_tokens": 40367389.0,
      "step": 18425
    },
    {
      "entropy": 5.597847366333008,
      "epoch": 1.8290988487495037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046694486165206573,
      "loss": 4.9953,
      "mean_token_accuracy": 0.1829256922006607,
      "num_tokens": 40376793.0,
      "step": 18430
    },
    {
      "entropy": 5.586914730072022,
      "epoch": 1.8295950774116712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004669263723005801,
      "loss": 5.0974,
      "mean_token_accuracy": 0.17654511481523513,
      "num_tokens": 40388025.0,
      "step": 18435
    },
    {
      "entropy": 5.603774166107177,
      "epoch": 1.8300913060738389,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046690787818973406,
      "loss": 5.1813,
      "mean_token_accuracy": 0.17449597120285035,
      "num_tokens": 40399731.0,
      "step": 18440
    },
    {
      "entropy": 5.654325532913208,
      "epoch": 1.8305875347360063,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046688937931998613,
      "loss": 5.2168,
      "mean_token_accuracy": 0.17375346273183823,
      "num_tokens": 40411078.0,
      "step": 18445
    },
    {
      "entropy": 5.7117475986480715,
      "epoch": 1.831083763398174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004668708756917951,
      "loss": 5.1817,
      "mean_token_accuracy": 0.1738217994570732,
      "num_tokens": 40422806.0,
      "step": 18450
    },
    {
      "entropy": 5.731017684936523,
      "epoch": 1.8315799920603415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046685236730561975,
      "loss": 5.2368,
      "mean_token_accuracy": 0.17135872095823287,
      "num_tokens": 40434041.0,
      "step": 18455
    },
    {
      "entropy": 5.607840633392334,
      "epoch": 1.832076220722509,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00046683385416191914,
      "loss": 5.0704,
      "mean_token_accuracy": 0.18276649266481398,
      "num_tokens": 40444658.0,
      "step": 18460
    },
    {
      "entropy": 5.588128519058228,
      "epoch": 1.8325724493846764,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046681533626115226,
      "loss": 5.0867,
      "mean_token_accuracy": 0.1775441661477089,
      "num_tokens": 40455745.0,
      "step": 18465
    },
    {
      "entropy": 5.575847148895264,
      "epoch": 1.833068678046844,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004667968136037784,
      "loss": 5.1835,
      "mean_token_accuracy": 0.1745190292596817,
      "num_tokens": 40468829.0,
      "step": 18470
    },
    {
      "entropy": 5.584440517425537,
      "epoch": 1.8335649067090114,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046677828619025687,
      "loss": 5.0882,
      "mean_token_accuracy": 0.17652197629213334,
      "num_tokens": 40479817.0,
      "step": 18475
    },
    {
      "entropy": 5.72104287147522,
      "epoch": 1.8340611353711789,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046675975402104706,
      "loss": 5.1967,
      "mean_token_accuracy": 0.1701135069131851,
      "num_tokens": 40490398.0,
      "step": 18480
    },
    {
      "entropy": 5.6074341297149655,
      "epoch": 1.8345573640333466,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046674121709660855,
      "loss": 5.0335,
      "mean_token_accuracy": 0.17980420142412185,
      "num_tokens": 40500483.0,
      "step": 18485
    },
    {
      "entropy": 5.653661727905273,
      "epoch": 1.835053592695514,
      "grad_norm": 0.890625,
      "learning_rate": 0.000466722675417401,
      "loss": 5.1428,
      "mean_token_accuracy": 0.16847982555627822,
      "num_tokens": 40511494.0,
      "step": 18490
    },
    {
      "entropy": 5.609594011306763,
      "epoch": 1.8355498213576817,
      "grad_norm": 1.0,
      "learning_rate": 0.00046670412898388425,
      "loss": 5.1023,
      "mean_token_accuracy": 0.18326909244060516,
      "num_tokens": 40523573.0,
      "step": 18495
    },
    {
      "entropy": 5.595342063903809,
      "epoch": 1.8360460500198492,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046668557779651825,
      "loss": 5.1443,
      "mean_token_accuracy": 0.17039330303668976,
      "num_tokens": 40534904.0,
      "step": 18500
    },
    {
      "entropy": 5.61021237373352,
      "epoch": 1.8365422786820167,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046666702185576295,
      "loss": 5.1088,
      "mean_token_accuracy": 0.17461351305246353,
      "num_tokens": 40545053.0,
      "step": 18505
    },
    {
      "entropy": 5.63336706161499,
      "epoch": 1.8370385073441842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004666484611620785,
      "loss": 5.0449,
      "mean_token_accuracy": 0.17969146519899368,
      "num_tokens": 40555305.0,
      "step": 18510
    },
    {
      "entropy": 5.595337152481079,
      "epoch": 1.8375347360063516,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004666298957159251,
      "loss": 5.0794,
      "mean_token_accuracy": 0.17218848019838334,
      "num_tokens": 40565611.0,
      "step": 18515
    },
    {
      "entropy": 5.615338897705078,
      "epoch": 1.8380309646685191,
      "grad_norm": 1.0,
      "learning_rate": 0.00046661132551776327,
      "loss": 5.1596,
      "mean_token_accuracy": 0.1786956310272217,
      "num_tokens": 40578321.0,
      "step": 18520
    },
    {
      "entropy": 5.6702704429626465,
      "epoch": 1.8385271933306868,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004665927505680535,
      "loss": 5.1586,
      "mean_token_accuracy": 0.18008727729320526,
      "num_tokens": 40588696.0,
      "step": 18525
    },
    {
      "entropy": 5.612645435333252,
      "epoch": 1.8390234219928543,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046657417086725625,
      "loss": 5.1128,
      "mean_token_accuracy": 0.17147452235221863,
      "num_tokens": 40599699.0,
      "step": 18530
    },
    {
      "entropy": 5.6028753280639645,
      "epoch": 1.839519650655022,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004665555864158324,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17680850476026536,
      "num_tokens": 40609706.0,
      "step": 18535
    },
    {
      "entropy": 5.604129266738892,
      "epoch": 1.8400158793171895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004665369972142428,
      "loss": 5.1567,
      "mean_token_accuracy": 0.1758778437972069,
      "num_tokens": 40619431.0,
      "step": 18540
    },
    {
      "entropy": 5.653966474533081,
      "epoch": 1.840512107979357,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046651840326294845,
      "loss": 5.1169,
      "mean_token_accuracy": 0.17865515649318695,
      "num_tokens": 40629904.0,
      "step": 18545
    },
    {
      "entropy": 5.601891374588012,
      "epoch": 1.8410083366415244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004664998045624104,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1701461046934128,
      "num_tokens": 40641266.0,
      "step": 18550
    },
    {
      "entropy": 5.572062301635742,
      "epoch": 1.8415045653036919,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004664812011130897,
      "loss": 5.1132,
      "mean_token_accuracy": 0.17343977391719817,
      "num_tokens": 40652454.0,
      "step": 18555
    },
    {
      "entropy": 5.621101379394531,
      "epoch": 1.8420007939658594,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046646259291544787,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1753406971693039,
      "num_tokens": 40664311.0,
      "step": 18560
    },
    {
      "entropy": 5.550297260284424,
      "epoch": 1.8424970226280268,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004664439799699463,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18417392075061798,
      "num_tokens": 40675241.0,
      "step": 18565
    },
    {
      "entropy": 5.585807275772095,
      "epoch": 1.8429932512901945,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046642536227704665,
      "loss": 5.0929,
      "mean_token_accuracy": 0.17489570677280425,
      "num_tokens": 40685972.0,
      "step": 18570
    },
    {
      "entropy": 5.607104253768921,
      "epoch": 1.843489479952362,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046640673983721037,
      "loss": 5.0954,
      "mean_token_accuracy": 0.17366386204957962,
      "num_tokens": 40696967.0,
      "step": 18575
    },
    {
      "entropy": 5.618705463409424,
      "epoch": 1.8439857086145297,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046638811265089944,
      "loss": 5.1294,
      "mean_token_accuracy": 0.1682773694396019,
      "num_tokens": 40707710.0,
      "step": 18580
    },
    {
      "entropy": 5.637105274200439,
      "epoch": 1.8444819372766972,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004663694807185758,
      "loss": 5.2353,
      "mean_token_accuracy": 0.17362237423658372,
      "num_tokens": 40718586.0,
      "step": 18585
    },
    {
      "entropy": 5.651723384857178,
      "epoch": 1.8449781659388647,
      "grad_norm": 1.0,
      "learning_rate": 0.00046635084404070135,
      "loss": 5.12,
      "mean_token_accuracy": 0.1772138759493828,
      "num_tokens": 40729792.0,
      "step": 18590
    },
    {
      "entropy": 5.7891364097595215,
      "epoch": 1.8454743946010321,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004663322026177382,
      "loss": 5.1724,
      "mean_token_accuracy": 0.16804285869002342,
      "num_tokens": 40740974.0,
      "step": 18595
    },
    {
      "entropy": 5.582426404953003,
      "epoch": 1.8459706232631996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046631355645014875,
      "loss": 4.9847,
      "mean_token_accuracy": 0.18482498973608016,
      "num_tokens": 40750208.0,
      "step": 18600
    },
    {
      "entropy": 5.603971385955811,
      "epoch": 1.846466851925367,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046629490553839537,
      "loss": 5.0764,
      "mean_token_accuracy": 0.17751141190528869,
      "num_tokens": 40762163.0,
      "step": 18605
    },
    {
      "entropy": 5.6431221008300785,
      "epoch": 1.8469630805875348,
      "grad_norm": 1.0,
      "learning_rate": 0.0004662762498829405,
      "loss": 5.1916,
      "mean_token_accuracy": 0.17604711651802063,
      "num_tokens": 40773028.0,
      "step": 18610
    },
    {
      "entropy": 5.675988578796387,
      "epoch": 1.8474593092497023,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004662575894842469,
      "loss": 5.1466,
      "mean_token_accuracy": 0.16866226494312286,
      "num_tokens": 40784939.0,
      "step": 18615
    },
    {
      "entropy": 5.660443830490112,
      "epoch": 1.84795553791187,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00046623892434277713,
      "loss": 5.1427,
      "mean_token_accuracy": 0.17590830475091934,
      "num_tokens": 40796170.0,
      "step": 18620
    },
    {
      "entropy": 5.606220293045044,
      "epoch": 1.8484517665740374,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004662202544589941,
      "loss": 5.0798,
      "mean_token_accuracy": 0.18046202957630159,
      "num_tokens": 40807184.0,
      "step": 18625
    },
    {
      "entropy": 5.5677446842193605,
      "epoch": 1.848947995236205,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046620157983336087,
      "loss": 5.0616,
      "mean_token_accuracy": 0.17265552580356597,
      "num_tokens": 40818474.0,
      "step": 18630
    },
    {
      "entropy": 5.642298746109009,
      "epoch": 1.8494442238983724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004661829004663404,
      "loss": 5.1151,
      "mean_token_accuracy": 0.17661844342947006,
      "num_tokens": 40828789.0,
      "step": 18635
    },
    {
      "entropy": 5.621394777297974,
      "epoch": 1.8499404525605398,
      "grad_norm": 1.0,
      "learning_rate": 0.0004661642163583961,
      "loss": 5.0766,
      "mean_token_accuracy": 0.1825094312429428,
      "num_tokens": 40840564.0,
      "step": 18640
    },
    {
      "entropy": 5.575894451141357,
      "epoch": 1.8504366812227073,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00046614552750999093,
      "loss": 5.0579,
      "mean_token_accuracy": 0.17347056269645691,
      "num_tokens": 40852528.0,
      "step": 18645
    },
    {
      "entropy": 5.602608156204224,
      "epoch": 1.8509329098848748,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004661268339215887,
      "loss": 5.0948,
      "mean_token_accuracy": 0.17629306465387345,
      "num_tokens": 40863242.0,
      "step": 18650
    },
    {
      "entropy": 5.587656164169312,
      "epoch": 1.8514291385470425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661081355936528,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18271770179271699,
      "num_tokens": 40873721.0,
      "step": 18655
    },
    {
      "entropy": 5.644473934173584,
      "epoch": 1.85192536720921,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000466089432526647,
      "loss": 5.1457,
      "mean_token_accuracy": 0.1720174252986908,
      "num_tokens": 40884921.0,
      "step": 18660
    },
    {
      "entropy": 5.637965297698974,
      "epoch": 1.8524215958713777,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046607072472103493,
      "loss": 5.0437,
      "mean_token_accuracy": 0.1737710252404213,
      "num_tokens": 40896243.0,
      "step": 18665
    },
    {
      "entropy": 5.6063313484191895,
      "epoch": 1.8529178245335451,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046605201217728067,
      "loss": 5.1693,
      "mean_token_accuracy": 0.1687125399708748,
      "num_tokens": 40907191.0,
      "step": 18670
    },
    {
      "entropy": 5.611230754852295,
      "epoch": 1.8534140531957126,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046603329489584817,
      "loss": 5.1206,
      "mean_token_accuracy": 0.17921803742647172,
      "num_tokens": 40918737.0,
      "step": 18675
    },
    {
      "entropy": 5.6648050308227536,
      "epoch": 1.85391028185788,
      "grad_norm": 1.0,
      "learning_rate": 0.0004660145728772016,
      "loss": 5.1454,
      "mean_token_accuracy": 0.17257591038942338,
      "num_tokens": 40930184.0,
      "step": 18680
    },
    {
      "entropy": 5.6116374969482425,
      "epoch": 1.8544065105200476,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004659958461218052,
      "loss": 5.0503,
      "mean_token_accuracy": 0.1748535841703415,
      "num_tokens": 40941483.0,
      "step": 18685
    },
    {
      "entropy": 5.583458757400512,
      "epoch": 1.854902739182215,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046597711463012344,
      "loss": 5.1329,
      "mean_token_accuracy": 0.17797277122735977,
      "num_tokens": 40952392.0,
      "step": 18690
    },
    {
      "entropy": 5.549638366699218,
      "epoch": 1.8553989678443827,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004659583784026207,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17399363964796066,
      "num_tokens": 40962539.0,
      "step": 18695
    },
    {
      "entropy": 5.592260646820068,
      "epoch": 1.8558951965065502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046593963743976163,
      "loss": 4.9918,
      "mean_token_accuracy": 0.1856878161430359,
      "num_tokens": 40972594.0,
      "step": 18700
    },
    {
      "entropy": 5.625839424133301,
      "epoch": 1.8563914251687177,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000465920891742011,
      "loss": 5.115,
      "mean_token_accuracy": 0.17649046331644058,
      "num_tokens": 40983460.0,
      "step": 18705
    },
    {
      "entropy": 5.58230767250061,
      "epoch": 1.8568876538308854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004659021413098337,
      "loss": 5.1081,
      "mean_token_accuracy": 0.17425971925258638,
      "num_tokens": 40993945.0,
      "step": 18710
    },
    {
      "entropy": 5.5208062648773195,
      "epoch": 1.8573838824930529,
      "grad_norm": 1.0,
      "learning_rate": 0.00046588338614369465,
      "loss": 5.1045,
      "mean_token_accuracy": 0.17678961902856827,
      "num_tokens": 41005781.0,
      "step": 18715
    },
    {
      "entropy": 5.664039373397827,
      "epoch": 1.8578801111552203,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046586462624405897,
      "loss": 5.2074,
      "mean_token_accuracy": 0.17238276451826096,
      "num_tokens": 41017650.0,
      "step": 18720
    },
    {
      "entropy": 5.661844301223755,
      "epoch": 1.8583763398173878,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004658458616113918,
      "loss": 5.0608,
      "mean_token_accuracy": 0.17738498449325563,
      "num_tokens": 41029196.0,
      "step": 18725
    },
    {
      "entropy": 5.637041902542114,
      "epoch": 1.8588725684795553,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004658270922461585,
      "loss": 5.1365,
      "mean_token_accuracy": 0.17985228300094605,
      "num_tokens": 41040204.0,
      "step": 18730
    },
    {
      "entropy": 5.638016986846924,
      "epoch": 1.8593687971417228,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046580831814882446,
      "loss": 5.1236,
      "mean_token_accuracy": 0.17152525186538697,
      "num_tokens": 41050507.0,
      "step": 18735
    },
    {
      "entropy": 5.575051021575928,
      "epoch": 1.8598650258038905,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046578953931985536,
      "loss": 5.0909,
      "mean_token_accuracy": 0.17251402288675308,
      "num_tokens": 41062001.0,
      "step": 18740
    },
    {
      "entropy": 5.629895639419556,
      "epoch": 1.860361254466058,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004657707557597167,
      "loss": 5.1098,
      "mean_token_accuracy": 0.16854377835988998,
      "num_tokens": 41073097.0,
      "step": 18745
    },
    {
      "entropy": 5.622098922729492,
      "epoch": 1.8608574831282256,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046575196746887447,
      "loss": 5.1368,
      "mean_token_accuracy": 0.1705164521932602,
      "num_tokens": 41083883.0,
      "step": 18750
    },
    {
      "entropy": 5.560026979446411,
      "epoch": 1.861353711790393,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046573317444779445,
      "loss": 5.102,
      "mean_token_accuracy": 0.1834790900349617,
      "num_tokens": 41095506.0,
      "step": 18755
    },
    {
      "entropy": 5.634246444702148,
      "epoch": 1.8618499404525606,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004657143766969427,
      "loss": 5.1253,
      "mean_token_accuracy": 0.177387598156929,
      "num_tokens": 41106568.0,
      "step": 18760
    },
    {
      "entropy": 5.602603387832642,
      "epoch": 1.862346169114728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004656955742167853,
      "loss": 5.0224,
      "mean_token_accuracy": 0.17828670144081116,
      "num_tokens": 41117661.0,
      "step": 18765
    },
    {
      "entropy": 5.5952380180358885,
      "epoch": 1.8628423977768955,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046567676700778865,
      "loss": 5.1,
      "mean_token_accuracy": 0.17979365736246108,
      "num_tokens": 41129209.0,
      "step": 18770
    },
    {
      "entropy": 5.61033182144165,
      "epoch": 1.863338626439063,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00046565795507041904,
      "loss": 5.1669,
      "mean_token_accuracy": 0.17390589714050292,
      "num_tokens": 41140715.0,
      "step": 18775
    },
    {
      "entropy": 5.632399702072144,
      "epoch": 1.8638348551012305,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000465639138405143,
      "loss": 5.1407,
      "mean_token_accuracy": 0.17511017769575118,
      "num_tokens": 41151080.0,
      "step": 18780
    },
    {
      "entropy": 5.656877040863037,
      "epoch": 1.8643310837633982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046562031701242703,
      "loss": 5.1523,
      "mean_token_accuracy": 0.1735799565911293,
      "num_tokens": 41162816.0,
      "step": 18785
    },
    {
      "entropy": 5.6147199153900145,
      "epoch": 1.8648273124255657,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046560149089273803,
      "loss": 5.0538,
      "mean_token_accuracy": 0.18187918663024902,
      "num_tokens": 41173456.0,
      "step": 18790
    },
    {
      "entropy": 5.574633646011352,
      "epoch": 1.8653235410877333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004655826600465427,
      "loss": 5.0084,
      "mean_token_accuracy": 0.1793573394417763,
      "num_tokens": 41184436.0,
      "step": 18795
    },
    {
      "entropy": 5.571736907958984,
      "epoch": 1.8658197697499008,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046556382447430816,
      "loss": 5.0797,
      "mean_token_accuracy": 0.1819749280810356,
      "num_tokens": 41196417.0,
      "step": 18800
    },
    {
      "entropy": 5.573965167999267,
      "epoch": 1.8663159984120683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046554498417650134,
      "loss": 5.0977,
      "mean_token_accuracy": 0.17395012378692626,
      "num_tokens": 41206054.0,
      "step": 18805
    },
    {
      "entropy": 5.5783899307250975,
      "epoch": 1.8668122270742358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004655261391535895,
      "loss": 5.0969,
      "mean_token_accuracy": 0.17692628800868987,
      "num_tokens": 41217002.0,
      "step": 18810
    },
    {
      "entropy": 5.724311637878418,
      "epoch": 1.8673084557364032,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004655072894060399,
      "loss": 5.2464,
      "mean_token_accuracy": 0.16830726265907286,
      "num_tokens": 41228485.0,
      "step": 18815
    },
    {
      "entropy": 5.640799951553345,
      "epoch": 1.8678046843985707,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004654884349343201,
      "loss": 5.1294,
      "mean_token_accuracy": 0.17651508003473282,
      "num_tokens": 41240600.0,
      "step": 18820
    },
    {
      "entropy": 5.701280307769776,
      "epoch": 1.8683009130607384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046546957573889757,
      "loss": 5.2087,
      "mean_token_accuracy": 0.16779563575983047,
      "num_tokens": 41251870.0,
      "step": 18825
    },
    {
      "entropy": 5.620864582061768,
      "epoch": 1.868797141722906,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046545071182024,
      "loss": 5.1197,
      "mean_token_accuracy": 0.17024701535701753,
      "num_tokens": 41264679.0,
      "step": 18830
    },
    {
      "entropy": 5.714161443710327,
      "epoch": 1.8692933703850736,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046543184317881516,
      "loss": 5.2145,
      "mean_token_accuracy": 0.1703433007001877,
      "num_tokens": 41275690.0,
      "step": 18835
    },
    {
      "entropy": 5.615005016326904,
      "epoch": 1.869789599047241,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000465412969815091,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17772131115198136,
      "num_tokens": 41285118.0,
      "step": 18840
    },
    {
      "entropy": 5.576655673980713,
      "epoch": 1.8702858277094085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653940917295354,
      "loss": 5.0807,
      "mean_token_accuracy": 0.17933025509119033,
      "num_tokens": 41296330.0,
      "step": 18845
    },
    {
      "entropy": 5.552207136154175,
      "epoch": 1.870782056371576,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004653752089226166,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17402651607990266,
      "num_tokens": 41307051.0,
      "step": 18850
    },
    {
      "entropy": 5.613187789916992,
      "epoch": 1.8712782850337435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004653563213948029,
      "loss": 5.0553,
      "mean_token_accuracy": 0.1830991268157959,
      "num_tokens": 41317111.0,
      "step": 18855
    },
    {
      "entropy": 5.544817209243774,
      "epoch": 1.871774513695911,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046533742914656254,
      "loss": 5.0359,
      "mean_token_accuracy": 0.1795032486319542,
      "num_tokens": 41327509.0,
      "step": 18860
    },
    {
      "entropy": 5.595528030395508,
      "epoch": 1.8722707423580784,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046531853217836414,
      "loss": 5.1308,
      "mean_token_accuracy": 0.17172084450721742,
      "num_tokens": 41338721.0,
      "step": 18865
    },
    {
      "entropy": 5.658637046813965,
      "epoch": 1.8727669710202461,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046529963049067623,
      "loss": 5.156,
      "mean_token_accuracy": 0.17172664552927017,
      "num_tokens": 41349767.0,
      "step": 18870
    },
    {
      "entropy": 5.687707567214966,
      "epoch": 1.8732631996824136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046528072408396747,
      "loss": 5.2845,
      "mean_token_accuracy": 0.16579007506370544,
      "num_tokens": 41361151.0,
      "step": 18875
    },
    {
      "entropy": 5.621701097488403,
      "epoch": 1.8737594283445813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004652618129587069,
      "loss": 5.0697,
      "mean_token_accuracy": 0.1739332154393196,
      "num_tokens": 41372875.0,
      "step": 18880
    },
    {
      "entropy": 5.687017917633057,
      "epoch": 1.8742556570067488,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046524289711536327,
      "loss": 5.1313,
      "mean_token_accuracy": 0.1742318242788315,
      "num_tokens": 41383284.0,
      "step": 18885
    },
    {
      "entropy": 5.604385042190552,
      "epoch": 1.8747518856689163,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004652239765544057,
      "loss": 5.098,
      "mean_token_accuracy": 0.18272093534469605,
      "num_tokens": 41395389.0,
      "step": 18890
    },
    {
      "entropy": 5.5525137424469,
      "epoch": 1.8752481143310837,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046520505127630346,
      "loss": 5.0512,
      "mean_token_accuracy": 0.17520064860582352,
      "num_tokens": 41406740.0,
      "step": 18895
    },
    {
      "entropy": 5.584644079208374,
      "epoch": 1.8757443429932512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046518612128152576,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18051227033138276,
      "num_tokens": 41417279.0,
      "step": 18900
    },
    {
      "entropy": 5.576508617401123,
      "epoch": 1.8762405716554187,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004651671865705421,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1796530768275261,
      "num_tokens": 41428574.0,
      "step": 18905
    },
    {
      "entropy": 5.598288249969483,
      "epoch": 1.8767368003175864,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046514824714382204,
      "loss": 5.0278,
      "mean_token_accuracy": 0.17417267560958863,
      "num_tokens": 41438784.0,
      "step": 18910
    },
    {
      "entropy": 5.605532693862915,
      "epoch": 1.8772330289797539,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004651293030018351,
      "loss": 5.0627,
      "mean_token_accuracy": 0.18129724264144897,
      "num_tokens": 41448481.0,
      "step": 18915
    },
    {
      "entropy": 5.643266201019287,
      "epoch": 1.8777292576419216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046511035414505115,
      "loss": 5.1297,
      "mean_token_accuracy": 0.18007885664701462,
      "num_tokens": 41459097.0,
      "step": 18920
    },
    {
      "entropy": 5.654841089248658,
      "epoch": 1.878225486304089,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004650914005739401,
      "loss": 5.1657,
      "mean_token_accuracy": 0.17090674191713334,
      "num_tokens": 41469647.0,
      "step": 18925
    },
    {
      "entropy": 5.655966758728027,
      "epoch": 1.8787217149662565,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004650724422889719,
      "loss": 5.1012,
      "mean_token_accuracy": 0.1706242248415947,
      "num_tokens": 41479826.0,
      "step": 18930
    },
    {
      "entropy": 5.683741807937622,
      "epoch": 1.879217943628424,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004650534792906167,
      "loss": 5.0728,
      "mean_token_accuracy": 0.1754855751991272,
      "num_tokens": 41490551.0,
      "step": 18935
    },
    {
      "entropy": 5.6175963401794435,
      "epoch": 1.8797141722905915,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004650345115793447,
      "loss": 5.0956,
      "mean_token_accuracy": 0.18067144006490707,
      "num_tokens": 41500908.0,
      "step": 18940
    },
    {
      "entropy": 5.590824270248413,
      "epoch": 1.880210400952759,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046501553915562644,
      "loss": 5.0645,
      "mean_token_accuracy": 0.17658614814281465,
      "num_tokens": 41511311.0,
      "step": 18945
    },
    {
      "entropy": 5.597450923919678,
      "epoch": 1.8807066296149264,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046499656201993213,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17914437204599382,
      "num_tokens": 41522068.0,
      "step": 18950
    },
    {
      "entropy": 5.5858899593353275,
      "epoch": 1.881202858277094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046497758017273255,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17009185403585433,
      "num_tokens": 41534099.0,
      "step": 18955
    },
    {
      "entropy": 5.635938405990601,
      "epoch": 1.8816990869392616,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046495859361449834,
      "loss": 5.1548,
      "mean_token_accuracy": 0.1703226462006569,
      "num_tokens": 41545870.0,
      "step": 18960
    },
    {
      "entropy": 5.6607424259185795,
      "epoch": 1.8821953156014293,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004649396023457003,
      "loss": 5.1526,
      "mean_token_accuracy": 0.17404603064060212,
      "num_tokens": 41558175.0,
      "step": 18965
    },
    {
      "entropy": 5.588940763473511,
      "epoch": 1.8826915442635968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004649206063668094,
      "loss": 5.1142,
      "mean_token_accuracy": 0.18140386790037155,
      "num_tokens": 41569326.0,
      "step": 18970
    },
    {
      "entropy": 5.582805728912353,
      "epoch": 1.8831877729257642,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004649016056782968,
      "loss": 5.0731,
      "mean_token_accuracy": 0.1822219356894493,
      "num_tokens": 41580065.0,
      "step": 18975
    },
    {
      "entropy": 5.692486429214478,
      "epoch": 1.8836840015879317,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004648826002806334,
      "loss": 5.2047,
      "mean_token_accuracy": 0.17627725452184678,
      "num_tokens": 41590915.0,
      "step": 18980
    },
    {
      "entropy": 5.58346848487854,
      "epoch": 1.8841802302500992,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004648635901742909,
      "loss": 5.0439,
      "mean_token_accuracy": 0.17933499217033386,
      "num_tokens": 41602096.0,
      "step": 18985
    },
    {
      "entropy": 5.585684585571289,
      "epoch": 1.8846764589122667,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004648445753597403,
      "loss": 5.1658,
      "mean_token_accuracy": 0.17130571454763413,
      "num_tokens": 41612589.0,
      "step": 18990
    },
    {
      "entropy": 5.585687446594238,
      "epoch": 1.8851726875744343,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00046482555583745336,
      "loss": 5.0425,
      "mean_token_accuracy": 0.176401586830616,
      "num_tokens": 41624163.0,
      "step": 18995
    },
    {
      "entropy": 5.6287901401519775,
      "epoch": 1.8856689162366018,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004648065316079017,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17919156700372696,
      "num_tokens": 41635919.0,
      "step": 19000
    },
    {
      "entropy": 5.616925525665283,
      "epoch": 1.8861651448987695,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046478750267155703,
      "loss": 5.077,
      "mean_token_accuracy": 0.17742199152708055,
      "num_tokens": 41644723.0,
      "step": 19005
    },
    {
      "entropy": 5.596064901351928,
      "epoch": 1.886661373560937,
      "grad_norm": 0.859375,
      "learning_rate": 0.00046476846902889125,
      "loss": 5.1441,
      "mean_token_accuracy": 0.17497727572917937,
      "num_tokens": 41657137.0,
      "step": 19010
    },
    {
      "entropy": 5.649802827835083,
      "epoch": 1.8871576022231045,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004647494306803764,
      "loss": 5.1142,
      "mean_token_accuracy": 0.1749027654528618,
      "num_tokens": 41667773.0,
      "step": 19015
    },
    {
      "entropy": 5.678574323654175,
      "epoch": 1.887653830885272,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046473038762648446,
      "loss": 5.1978,
      "mean_token_accuracy": 0.17025222927331923,
      "num_tokens": 41680549.0,
      "step": 19020
    },
    {
      "entropy": 5.635614204406738,
      "epoch": 1.8881500595474394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046471133986768784,
      "loss": 5.0725,
      "mean_token_accuracy": 0.1806030884385109,
      "num_tokens": 41689700.0,
      "step": 19025
    },
    {
      "entropy": 5.640012454986572,
      "epoch": 1.888646288209607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046469228740445874,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17467286735773085,
      "num_tokens": 41699734.0,
      "step": 19030
    },
    {
      "entropy": 5.59315071105957,
      "epoch": 1.8891425168717744,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046467323023726965,
      "loss": 5.132,
      "mean_token_accuracy": 0.1760464444756508,
      "num_tokens": 41710644.0,
      "step": 19035
    },
    {
      "entropy": 5.5518089771270756,
      "epoch": 1.889638745533942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046465416836659313,
      "loss": 5.0308,
      "mean_token_accuracy": 0.184174282848835,
      "num_tokens": 41721416.0,
      "step": 19040
    },
    {
      "entropy": 5.6175695896148685,
      "epoch": 1.8901349741961095,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000464635101792902,
      "loss": 5.0447,
      "mean_token_accuracy": 0.17594312727451325,
      "num_tokens": 41732139.0,
      "step": 19045
    },
    {
      "entropy": 5.627979373931884,
      "epoch": 1.8906312028582772,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046461603051666883,
      "loss": 5.1442,
      "mean_token_accuracy": 0.1729724168777466,
      "num_tokens": 41744233.0,
      "step": 19050
    },
    {
      "entropy": 5.602883195877075,
      "epoch": 1.8911274315204447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046459695453836676,
      "loss": 5.0763,
      "mean_token_accuracy": 0.18213993906974793,
      "num_tokens": 41755514.0,
      "step": 19055
    },
    {
      "entropy": 5.635495615005493,
      "epoch": 1.8916236601826122,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004645778738584688,
      "loss": 5.0976,
      "mean_token_accuracy": 0.17559309899806977,
      "num_tokens": 41765613.0,
      "step": 19060
    },
    {
      "entropy": 5.65129771232605,
      "epoch": 1.8921198888447797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004645587884774481,
      "loss": 5.1392,
      "mean_token_accuracy": 0.17199637442827226,
      "num_tokens": 41777294.0,
      "step": 19065
    },
    {
      "entropy": 5.5666594982147215,
      "epoch": 1.8926161175069471,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004645396983957778,
      "loss": 5.0393,
      "mean_token_accuracy": 0.17936763763427735,
      "num_tokens": 41788278.0,
      "step": 19070
    },
    {
      "entropy": 5.618350696563721,
      "epoch": 1.8931123461691146,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046452060361393147,
      "loss": 5.1055,
      "mean_token_accuracy": 0.1707746133208275,
      "num_tokens": 41798663.0,
      "step": 19075
    },
    {
      "entropy": 5.627126216888428,
      "epoch": 1.8936085748312823,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004645015041323825,
      "loss": 5.0632,
      "mean_token_accuracy": 0.18226485550403596,
      "num_tokens": 41808711.0,
      "step": 19080
    },
    {
      "entropy": 5.614083766937256,
      "epoch": 1.8941048034934498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046448239995160455,
      "loss": 5.1097,
      "mean_token_accuracy": 0.17078835666179656,
      "num_tokens": 41819679.0,
      "step": 19085
    },
    {
      "entropy": 5.651131725311279,
      "epoch": 1.8946010321556173,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004644632910720715,
      "loss": 5.1012,
      "mean_token_accuracy": 0.17519588470458985,
      "num_tokens": 41829649.0,
      "step": 19090
    },
    {
      "entropy": 5.63698935508728,
      "epoch": 1.895097260817785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000464444177494257,
      "loss": 5.142,
      "mean_token_accuracy": 0.16884741187095642,
      "num_tokens": 41842523.0,
      "step": 19095
    },
    {
      "entropy": 5.615895986557007,
      "epoch": 1.8955934894799524,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004644250592186351,
      "loss": 5.1134,
      "mean_token_accuracy": 0.17199523746967316,
      "num_tokens": 41854514.0,
      "step": 19100
    },
    {
      "entropy": 5.6436210632324215,
      "epoch": 1.89608971814212,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046440593624568,
      "loss": 5.0851,
      "mean_token_accuracy": 0.17718427032232284,
      "num_tokens": 41865166.0,
      "step": 19105
    },
    {
      "entropy": 5.629592037200927,
      "epoch": 1.8965859468042874,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004643868085758657,
      "loss": 5.0501,
      "mean_token_accuracy": 0.1794016405940056,
      "num_tokens": 41877264.0,
      "step": 19110
    },
    {
      "entropy": 5.59939432144165,
      "epoch": 1.8970821754664549,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046436767620966675,
      "loss": 5.0639,
      "mean_token_accuracy": 0.17188819497823715,
      "num_tokens": 41887345.0,
      "step": 19115
    },
    {
      "entropy": 5.629842042922974,
      "epoch": 1.8975784041286223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046434853914755744,
      "loss": 5.1136,
      "mean_token_accuracy": 0.17434400618076323,
      "num_tokens": 41898877.0,
      "step": 19120
    },
    {
      "entropy": 5.584552192687989,
      "epoch": 1.89807463279079,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046432939739001237,
      "loss": 5.0506,
      "mean_token_accuracy": 0.17773896753787993,
      "num_tokens": 41910593.0,
      "step": 19125
    },
    {
      "entropy": 5.609117698669434,
      "epoch": 1.8985708614529575,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046431025093750613,
      "loss": 5.0776,
      "mean_token_accuracy": 0.18243939578533172,
      "num_tokens": 41920719.0,
      "step": 19130
    },
    {
      "entropy": 5.65151629447937,
      "epoch": 1.8990670901151252,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046429109979051367,
      "loss": 5.1205,
      "mean_token_accuracy": 0.1649581789970398,
      "num_tokens": 41931630.0,
      "step": 19135
    },
    {
      "entropy": 5.624905157089233,
      "epoch": 1.8995633187772927,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046427194394950985,
      "loss": 5.1538,
      "mean_token_accuracy": 0.1774039685726166,
      "num_tokens": 41943076.0,
      "step": 19140
    },
    {
      "entropy": 5.649903345108032,
      "epoch": 1.9000595474394602,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004642527834149696,
      "loss": 5.0459,
      "mean_token_accuracy": 0.17752454727888106,
      "num_tokens": 41954294.0,
      "step": 19145
    },
    {
      "entropy": 5.588830804824829,
      "epoch": 1.9005557761016276,
      "grad_norm": 1.125,
      "learning_rate": 0.00046423361818736814,
      "loss": 5.0355,
      "mean_token_accuracy": 0.1725974202156067,
      "num_tokens": 41964943.0,
      "step": 19150
    },
    {
      "entropy": 5.5443497657775875,
      "epoch": 1.901052004763795,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004642144482671807,
      "loss": 5.0881,
      "mean_token_accuracy": 0.1781328558921814,
      "num_tokens": 41975211.0,
      "step": 19155
    },
    {
      "entropy": 5.6175612449646,
      "epoch": 1.9015482334259626,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046419527365488275,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17325039356946945,
      "num_tokens": 41986492.0,
      "step": 19160
    },
    {
      "entropy": 5.596925067901611,
      "epoch": 1.90204446208813,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004641760943509496,
      "loss": 5.0732,
      "mean_token_accuracy": 0.17799487560987473,
      "num_tokens": 41997613.0,
      "step": 19165
    },
    {
      "entropy": 5.590560054779052,
      "epoch": 1.9025406907502977,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000464156910355857,
      "loss": 5.1307,
      "mean_token_accuracy": 0.16934071630239486,
      "num_tokens": 42008597.0,
      "step": 19170
    },
    {
      "entropy": 5.584636449813843,
      "epoch": 1.9030369194124652,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046413772167008057,
      "loss": 5.1289,
      "mean_token_accuracy": 0.17767439782619476,
      "num_tokens": 42019698.0,
      "step": 19175
    },
    {
      "entropy": 5.621352672576904,
      "epoch": 1.903533148074633,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046411852829409626,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1791344776749611,
      "num_tokens": 42029473.0,
      "step": 19180
    },
    {
      "entropy": 5.593325519561768,
      "epoch": 1.9040293767368004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046409933022837984,
      "loss": 5.1622,
      "mean_token_accuracy": 0.16624840199947358,
      "num_tokens": 42041494.0,
      "step": 19185
    },
    {
      "entropy": 5.616284990310669,
      "epoch": 1.9045256053989679,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640801274734076,
      "loss": 5.1262,
      "mean_token_accuracy": 0.17743525505065919,
      "num_tokens": 42051798.0,
      "step": 19190
    },
    {
      "entropy": 5.616162538528442,
      "epoch": 1.9050218340611353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004640609200296557,
      "loss": 5.0443,
      "mean_token_accuracy": 0.17735201567411424,
      "num_tokens": 42061802.0,
      "step": 19195
    },
    {
      "entropy": 5.660167169570923,
      "epoch": 1.9055180627233028,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004640417078976003,
      "loss": 5.1704,
      "mean_token_accuracy": 0.16831487268209458,
      "num_tokens": 42071503.0,
      "step": 19200
    },
    {
      "entropy": 5.5126549243927006,
      "epoch": 1.9060142913854703,
      "grad_norm": 1.125,
      "learning_rate": 0.0004640224910777179,
      "loss": 5.0149,
      "mean_token_accuracy": 0.17719407081604005,
      "num_tokens": 42082104.0,
      "step": 19205
    },
    {
      "entropy": 5.6353997707366945,
      "epoch": 1.906510520047638,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000464003269570485,
      "loss": 5.0828,
      "mean_token_accuracy": 0.17648355066776275,
      "num_tokens": 42092798.0,
      "step": 19210
    },
    {
      "entropy": 5.634766530990601,
      "epoch": 1.9070067487098055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004639840433763783,
      "loss": 5.1291,
      "mean_token_accuracy": 0.17383530139923095,
      "num_tokens": 42103062.0,
      "step": 19215
    },
    {
      "entropy": 5.624455785751342,
      "epoch": 1.9075029773719732,
      "grad_norm": 1.0,
      "learning_rate": 0.00046396481249587454,
      "loss": 5.1547,
      "mean_token_accuracy": 0.1786423832178116,
      "num_tokens": 42115062.0,
      "step": 19220
    },
    {
      "entropy": 5.679159355163574,
      "epoch": 1.9079992060341406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004639455769294506,
      "loss": 5.1501,
      "mean_token_accuracy": 0.17160236835479736,
      "num_tokens": 42125937.0,
      "step": 19225
    },
    {
      "entropy": 5.655734395980835,
      "epoch": 1.9084954346963081,
      "grad_norm": 1.0,
      "learning_rate": 0.0004639263366775835,
      "loss": 5.0839,
      "mean_token_accuracy": 0.17577982991933822,
      "num_tokens": 42136916.0,
      "step": 19230
    },
    {
      "entropy": 5.66775918006897,
      "epoch": 1.9089916633584756,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046390709174075036,
      "loss": 5.156,
      "mean_token_accuracy": 0.17332761138677596,
      "num_tokens": 42148348.0,
      "step": 19235
    },
    {
      "entropy": 5.65292706489563,
      "epoch": 1.909487892020643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004638878421194284,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18566261231899261,
      "num_tokens": 42157586.0,
      "step": 19240
    },
    {
      "entropy": 5.613950729370117,
      "epoch": 1.9099841206828105,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000463868587814095,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17589639872312546,
      "num_tokens": 42169536.0,
      "step": 19245
    },
    {
      "entropy": 5.557609987258911,
      "epoch": 1.910480349344978,
      "grad_norm": 1.0,
      "learning_rate": 0.0004638493288252276,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1803385615348816,
      "num_tokens": 42180024.0,
      "step": 19250
    },
    {
      "entropy": 5.6079905986785885,
      "epoch": 1.9109765780071457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046383006515330374,
      "loss": 5.0661,
      "mean_token_accuracy": 0.17720706462860109,
      "num_tokens": 42191070.0,
      "step": 19255
    },
    {
      "entropy": 5.648953294754028,
      "epoch": 1.9114728066693132,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004638107967988012,
      "loss": 5.1194,
      "mean_token_accuracy": 0.1783156216144562,
      "num_tokens": 42203643.0,
      "step": 19260
    },
    {
      "entropy": 5.665133237838745,
      "epoch": 1.9119690353314809,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046379152376219775,
      "loss": 5.1553,
      "mean_token_accuracy": 0.16805948615074157,
      "num_tokens": 42214979.0,
      "step": 19265
    },
    {
      "entropy": 5.72856822013855,
      "epoch": 1.9124652639936484,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046377224604397127,
      "loss": 5.1381,
      "mean_token_accuracy": 0.17146701961755753,
      "num_tokens": 42225152.0,
      "step": 19270
    },
    {
      "entropy": 5.669720315933228,
      "epoch": 1.9129614926558158,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004637529636445998,
      "loss": 5.0724,
      "mean_token_accuracy": 0.1794816240668297,
      "num_tokens": 42235151.0,
      "step": 19275
    },
    {
      "entropy": 5.614039278030395,
      "epoch": 1.9134577213179833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004637336765645617,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18438845723867417,
      "num_tokens": 42245419.0,
      "step": 19280
    },
    {
      "entropy": 5.538188838958741,
      "epoch": 1.9139539499801508,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046371438480433504,
      "loss": 5.0489,
      "mean_token_accuracy": 0.1774456173181534,
      "num_tokens": 42255836.0,
      "step": 19285
    },
    {
      "entropy": 5.576700258255005,
      "epoch": 1.9144501786423183,
      "grad_norm": 0.875,
      "learning_rate": 0.0004636950883643983,
      "loss": 5.1447,
      "mean_token_accuracy": 0.17315268963575364,
      "num_tokens": 42267490.0,
      "step": 19290
    },
    {
      "entropy": 5.6920592308044435,
      "epoch": 1.914946407304486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046367578724523,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1757742553949356,
      "num_tokens": 42276957.0,
      "step": 19295
    },
    {
      "entropy": 5.651052618026734,
      "epoch": 1.9154426359666534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004636564814473087,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18405548334121705,
      "num_tokens": 42288469.0,
      "step": 19300
    },
    {
      "entropy": 5.599853229522705,
      "epoch": 1.9159388646288211,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004636371709711131,
      "loss": 5.0358,
      "mean_token_accuracy": 0.1882678210735321,
      "num_tokens": 42299141.0,
      "step": 19305
    },
    {
      "entropy": 5.5870616912841795,
      "epoch": 1.9164350932909886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046361785581712224,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17595580965280533,
      "num_tokens": 42310651.0,
      "step": 19310
    },
    {
      "entropy": 5.569118785858154,
      "epoch": 1.916931321953156,
      "grad_norm": 1.046875,
      "learning_rate": 0.000463598535985815,
      "loss": 4.9523,
      "mean_token_accuracy": 0.1900081902742386,
      "num_tokens": 42320529.0,
      "step": 19315
    },
    {
      "entropy": 5.598519277572632,
      "epoch": 1.9174275506153236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004635792114776704,
      "loss": 5.0042,
      "mean_token_accuracy": 0.18219348043203354,
      "num_tokens": 42331216.0,
      "step": 19320
    },
    {
      "entropy": 5.529211902618409,
      "epoch": 1.917923779277491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004635598822931677,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18210722059011458,
      "num_tokens": 42341730.0,
      "step": 19325
    },
    {
      "entropy": 5.649684095382691,
      "epoch": 1.9184200079396585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046354054843278624,
      "loss": 5.1178,
      "mean_token_accuracy": 0.1746904581785202,
      "num_tokens": 42352643.0,
      "step": 19330
    },
    {
      "entropy": 5.675132894515992,
      "epoch": 1.918916236601826,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046352120989700553,
      "loss": 5.1665,
      "mean_token_accuracy": 0.1707271859049797,
      "num_tokens": 42363979.0,
      "step": 19335
    },
    {
      "entropy": 5.650002956390381,
      "epoch": 1.9194124652639937,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000463501866686305,
      "loss": 5.1005,
      "mean_token_accuracy": 0.1789351299405098,
      "num_tokens": 42374104.0,
      "step": 19340
    },
    {
      "entropy": 5.6553057670593265,
      "epoch": 1.9199086939261611,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046348251880116433,
      "loss": 5.0855,
      "mean_token_accuracy": 0.17710110396146775,
      "num_tokens": 42383576.0,
      "step": 19345
    },
    {
      "entropy": 5.634666395187378,
      "epoch": 1.9204049225883288,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046346316624206325,
      "loss": 5.1019,
      "mean_token_accuracy": 0.17811881005764008,
      "num_tokens": 42394625.0,
      "step": 19350
    },
    {
      "entropy": 5.6422919750213625,
      "epoch": 1.9209011512504963,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004634438090094819,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18091819286346436,
      "num_tokens": 42404764.0,
      "step": 19355
    },
    {
      "entropy": 5.6109778881073,
      "epoch": 1.9213973799126638,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046342444710390006,
      "loss": 5.039,
      "mean_token_accuracy": 0.1787576675415039,
      "num_tokens": 42414386.0,
      "step": 19360
    },
    {
      "entropy": 5.602859354019165,
      "epoch": 1.9218936085748313,
      "grad_norm": 1.03125,
      "learning_rate": 0.000463405080525798,
      "loss": 5.1179,
      "mean_token_accuracy": 0.17147554606199264,
      "num_tokens": 42425671.0,
      "step": 19365
    },
    {
      "entropy": 5.619019222259522,
      "epoch": 1.9223898372369987,
      "grad_norm": 1.0,
      "learning_rate": 0.0004633857092756559,
      "loss": 5.1462,
      "mean_token_accuracy": 0.17126853466033937,
      "num_tokens": 42436365.0,
      "step": 19370
    },
    {
      "entropy": 5.5986024856567385,
      "epoch": 1.9228860658991662,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004633663333539542,
      "loss": 5.0122,
      "mean_token_accuracy": 0.17735473066568375,
      "num_tokens": 42447624.0,
      "step": 19375
    },
    {
      "entropy": 5.679117107391358,
      "epoch": 1.923382294561334,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046334695276117324,
      "loss": 5.1103,
      "mean_token_accuracy": 0.17637975960969926,
      "num_tokens": 42458013.0,
      "step": 19380
    },
    {
      "entropy": 5.6473499774932865,
      "epoch": 1.9238785232235014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004633275674977937,
      "loss": 5.1188,
      "mean_token_accuracy": 0.1752373218536377,
      "num_tokens": 42469807.0,
      "step": 19385
    },
    {
      "entropy": 5.587552309036255,
      "epoch": 1.9243747518856689,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004633081775642964,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17930302023887634,
      "num_tokens": 42479240.0,
      "step": 19390
    },
    {
      "entropy": 5.6145471096038815,
      "epoch": 1.9248709805478366,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046328878296116196,
      "loss": 5.1203,
      "mean_token_accuracy": 0.1781229168176651,
      "num_tokens": 42489558.0,
      "step": 19395
    },
    {
      "entropy": 5.635738468170166,
      "epoch": 1.925367209210004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046326938368887146,
      "loss": 5.1255,
      "mean_token_accuracy": 0.17548428624868392,
      "num_tokens": 42501252.0,
      "step": 19400
    },
    {
      "entropy": 5.657933950424194,
      "epoch": 1.9258634378721715,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046324997974790596,
      "loss": 5.1612,
      "mean_token_accuracy": 0.17761810421943663,
      "num_tokens": 42512330.0,
      "step": 19405
    },
    {
      "entropy": 5.652893114089966,
      "epoch": 1.926359666534339,
      "grad_norm": 1.0,
      "learning_rate": 0.0004632305711387466,
      "loss": 5.1479,
      "mean_token_accuracy": 0.17240212857723236,
      "num_tokens": 42525216.0,
      "step": 19410
    },
    {
      "entropy": 5.649395942687988,
      "epoch": 1.9268558951965065,
      "grad_norm": 1.0,
      "learning_rate": 0.00046321115786187466,
      "loss": 5.0786,
      "mean_token_accuracy": 0.17519006431102752,
      "num_tokens": 42537047.0,
      "step": 19415
    },
    {
      "entropy": 5.624427223205567,
      "epoch": 1.927352123858674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046319173991777153,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17797528058290482,
      "num_tokens": 42549784.0,
      "step": 19420
    },
    {
      "entropy": 5.616265058517456,
      "epoch": 1.9278483525208416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004631723173069188,
      "loss": 5.0687,
      "mean_token_accuracy": 0.17784585952758789,
      "num_tokens": 42561079.0,
      "step": 19425
    },
    {
      "entropy": 5.636670112609863,
      "epoch": 1.928344581183009,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004631528900297981,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17064312398433684,
      "num_tokens": 42572531.0,
      "step": 19430
    },
    {
      "entropy": 5.618291759490967,
      "epoch": 1.9288408098451768,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046313345808689113,
      "loss": 5.0553,
      "mean_token_accuracy": 0.17775977104902269,
      "num_tokens": 42583836.0,
      "step": 19435
    },
    {
      "entropy": 5.641771984100342,
      "epoch": 1.9293370385073443,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004631140214786798,
      "loss": 5.1708,
      "mean_token_accuracy": 0.16881755590438843,
      "num_tokens": 42595338.0,
      "step": 19440
    },
    {
      "entropy": 5.619860887527466,
      "epoch": 1.9298332671695118,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046309458020564606,
      "loss": 5.1091,
      "mean_token_accuracy": 0.17218412458896637,
      "num_tokens": 42606804.0,
      "step": 19445
    },
    {
      "entropy": 5.604937791824341,
      "epoch": 1.9303294958316792,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046307513426827214,
      "loss": 5.0186,
      "mean_token_accuracy": 0.1821754202246666,
      "num_tokens": 42616987.0,
      "step": 19450
    },
    {
      "entropy": 5.637788486480713,
      "epoch": 1.9308257244938467,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046305568366704006,
      "loss": 5.1035,
      "mean_token_accuracy": 0.17422610521316528,
      "num_tokens": 42627633.0,
      "step": 19455
    },
    {
      "entropy": 5.654410362243652,
      "epoch": 1.9313219531560142,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004630362284024322,
      "loss": 5.083,
      "mean_token_accuracy": 0.17787501513957976,
      "num_tokens": 42638802.0,
      "step": 19460
    },
    {
      "entropy": 5.614974403381348,
      "epoch": 1.9318181818181817,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046301676847493116,
      "loss": 5.0688,
      "mean_token_accuracy": 0.18088817298412324,
      "num_tokens": 42650802.0,
      "step": 19465
    },
    {
      "entropy": 5.604468536376953,
      "epoch": 1.9323144104803494,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046299730388501935,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18095044940710067,
      "num_tokens": 42660290.0,
      "step": 19470
    },
    {
      "entropy": 5.634112024307251,
      "epoch": 1.9328106391425168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046297783463317957,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17484190613031386,
      "num_tokens": 42671182.0,
      "step": 19475
    },
    {
      "entropy": 5.658721208572388,
      "epoch": 1.9333068678046845,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004629583607198945,
      "loss": 5.1222,
      "mean_token_accuracy": 0.1768755942583084,
      "num_tokens": 42682499.0,
      "step": 19480
    },
    {
      "entropy": 5.5974561214447025,
      "epoch": 1.933803096466852,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046293888214564704,
      "loss": 5.1249,
      "mean_token_accuracy": 0.1744990974664688,
      "num_tokens": 42693728.0,
      "step": 19485
    },
    {
      "entropy": 5.632141733169556,
      "epoch": 1.9342993251290195,
      "grad_norm": 1.0,
      "learning_rate": 0.00046291939891092036,
      "loss": 5.1278,
      "mean_token_accuracy": 0.17454599887132644,
      "num_tokens": 42704877.0,
      "step": 19490
    },
    {
      "entropy": 5.689607954025268,
      "epoch": 1.934795553791187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004628999110161975,
      "loss": 5.1536,
      "mean_token_accuracy": 0.16930246651172637,
      "num_tokens": 42715787.0,
      "step": 19495
    },
    {
      "entropy": 5.665239715576172,
      "epoch": 1.9352917824533544,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046288041846196174,
      "loss": 5.1043,
      "mean_token_accuracy": 0.1729670986533165,
      "num_tokens": 42727557.0,
      "step": 19500
    },
    {
      "entropy": 5.638056707382202,
      "epoch": 1.935788011115522,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046286092124869647,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17202329635620117,
      "num_tokens": 42737540.0,
      "step": 19505
    },
    {
      "entropy": 5.68894476890564,
      "epoch": 1.9362842397776896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046284141937688513,
      "loss": 5.1314,
      "mean_token_accuracy": 0.18351756483316423,
      "num_tokens": 42748716.0,
      "step": 19510
    },
    {
      "entropy": 5.748136281967163,
      "epoch": 1.936780468439857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046282191284701123,
      "loss": 5.1742,
      "mean_token_accuracy": 0.17116613686084747,
      "num_tokens": 42758991.0,
      "step": 19515
    },
    {
      "entropy": 5.699264812469482,
      "epoch": 1.9372766971020248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004628024016595587,
      "loss": 5.1394,
      "mean_token_accuracy": 0.1711787387728691,
      "num_tokens": 42771105.0,
      "step": 19520
    },
    {
      "entropy": 5.59972562789917,
      "epoch": 1.9377729257641922,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00046278288581501136,
      "loss": 5.1098,
      "mean_token_accuracy": 0.17940554171800613,
      "num_tokens": 42783317.0,
      "step": 19525
    },
    {
      "entropy": 5.669765996932983,
      "epoch": 1.9382691544263597,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046276336531385305,
      "loss": 5.2167,
      "mean_token_accuracy": 0.16810741871595383,
      "num_tokens": 42794998.0,
      "step": 19530
    },
    {
      "entropy": 5.654805660247803,
      "epoch": 1.9387653830885272,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046274384015656785,
      "loss": 5.0907,
      "mean_token_accuracy": 0.18001317977905273,
      "num_tokens": 42806197.0,
      "step": 19535
    },
    {
      "entropy": 5.602545547485351,
      "epoch": 1.9392616117506947,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046272431034363997,
      "loss": 5.0111,
      "mean_token_accuracy": 0.173931984603405,
      "num_tokens": 42816522.0,
      "step": 19540
    },
    {
      "entropy": 5.589268732070923,
      "epoch": 1.9397578404128621,
      "grad_norm": 1.125,
      "learning_rate": 0.0004627047758755538,
      "loss": 5.0203,
      "mean_token_accuracy": 0.18109623789787294,
      "num_tokens": 42826003.0,
      "step": 19545
    },
    {
      "entropy": 5.622390985488892,
      "epoch": 1.9402540690750296,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004626852367527935,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1697564885020256,
      "num_tokens": 42836346.0,
      "step": 19550
    },
    {
      "entropy": 5.6265623569488525,
      "epoch": 1.9407502977371973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004626656929758438,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17363854199647905,
      "num_tokens": 42846296.0,
      "step": 19555
    },
    {
      "entropy": 5.693745136260986,
      "epoch": 1.9412465263993648,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004626461445451894,
      "loss": 5.1586,
      "mean_token_accuracy": 0.17259424924850464,
      "num_tokens": 42856323.0,
      "step": 19560
    },
    {
      "entropy": 5.709331274032593,
      "epoch": 1.9417427550615325,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004626265914613149,
      "loss": 5.1658,
      "mean_token_accuracy": 0.17209894359111785,
      "num_tokens": 42866937.0,
      "step": 19565
    },
    {
      "entropy": 5.653783321380615,
      "epoch": 1.9422389837237,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004626070337247052,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17046387642621993,
      "num_tokens": 42877194.0,
      "step": 19570
    },
    {
      "entropy": 5.641523647308349,
      "epoch": 1.9427352123858674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004625874713358453,
      "loss": 5.1317,
      "mean_token_accuracy": 0.1721916139125824,
      "num_tokens": 42887741.0,
      "step": 19575
    },
    {
      "entropy": 5.6333293437957765,
      "epoch": 1.943231441048035,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046256790429522035,
      "loss": 5.1033,
      "mean_token_accuracy": 0.18363362550735474,
      "num_tokens": 42898195.0,
      "step": 19580
    },
    {
      "entropy": 5.728744554519653,
      "epoch": 1.9437276697102024,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004625483326033155,
      "loss": 5.1318,
      "mean_token_accuracy": 0.17626961171627045,
      "num_tokens": 42908852.0,
      "step": 19585
    },
    {
      "entropy": 5.598348140716553,
      "epoch": 1.9442238983723699,
      "grad_norm": 1.0,
      "learning_rate": 0.0004625287562606162,
      "loss": 5.0548,
      "mean_token_accuracy": 0.17729806751012803,
      "num_tokens": 42920166.0,
      "step": 19590
    },
    {
      "entropy": 5.651196718215942,
      "epoch": 1.9447201270345376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004625091752676078,
      "loss": 5.1116,
      "mean_token_accuracy": 0.17633812725543976,
      "num_tokens": 42931677.0,
      "step": 19595
    },
    {
      "entropy": 5.645641613006592,
      "epoch": 1.945216355696705,
      "grad_norm": 1.0625,
      "learning_rate": 0.000462489589624776,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17761460542678834,
      "num_tokens": 42942346.0,
      "step": 19600
    },
    {
      "entropy": 5.539219760894776,
      "epoch": 1.9457125843588727,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004624699993326063,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1781325712800026,
      "num_tokens": 42953821.0,
      "step": 19605
    },
    {
      "entropy": 5.606766223907471,
      "epoch": 1.9462088130210402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046245040439158456,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1787712201476097,
      "num_tokens": 42964358.0,
      "step": 19610
    },
    {
      "entropy": 5.610420751571655,
      "epoch": 1.9467050416832077,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046243080480219673,
      "loss": 5.0708,
      "mean_token_accuracy": 0.17430377751588821,
      "num_tokens": 42976459.0,
      "step": 19615
    },
    {
      "entropy": 5.715780639648438,
      "epoch": 1.9472012703453752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004624112005649288,
      "loss": 5.2219,
      "mean_token_accuracy": 0.16757089793682098,
      "num_tokens": 42988603.0,
      "step": 19620
    },
    {
      "entropy": 5.643850898742675,
      "epoch": 1.9476974990075426,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046239159168026694,
      "loss": 5.1249,
      "mean_token_accuracy": 0.17509018927812575,
      "num_tokens": 43000262.0,
      "step": 19625
    },
    {
      "entropy": 5.597548770904541,
      "epoch": 1.94819372766971,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004623719781486974,
      "loss": 5.0624,
      "mean_token_accuracy": 0.18060435205698014,
      "num_tokens": 43009363.0,
      "step": 19630
    },
    {
      "entropy": 5.663141822814941,
      "epoch": 1.9486899563318776,
      "grad_norm": 1.0,
      "learning_rate": 0.00046235235997070664,
      "loss": 5.1225,
      "mean_token_accuracy": 0.17553853541612624,
      "num_tokens": 43019893.0,
      "step": 19635
    },
    {
      "entropy": 5.581222105026245,
      "epoch": 1.9491861849940453,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000462332737146781,
      "loss": 5.0347,
      "mean_token_accuracy": 0.18125436305999756,
      "num_tokens": 43031152.0,
      "step": 19640
    },
    {
      "entropy": 5.617210388183594,
      "epoch": 1.9496824136562128,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004623131096774071,
      "loss": 5.174,
      "mean_token_accuracy": 0.16962123811244964,
      "num_tokens": 43041247.0,
      "step": 19645
    },
    {
      "entropy": 5.665547895431518,
      "epoch": 1.9501786423183805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004622934775630717,
      "loss": 5.0913,
      "mean_token_accuracy": 0.16754546761512756,
      "num_tokens": 43052065.0,
      "step": 19650
    },
    {
      "entropy": 5.633870553970337,
      "epoch": 1.950674870980548,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00046227384080426163,
      "loss": 5.1232,
      "mean_token_accuracy": 0.1700010597705841,
      "num_tokens": 43063217.0,
      "step": 19655
    },
    {
      "entropy": 5.6282402038574215,
      "epoch": 1.9511710996427154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004622541994014639,
      "loss": 5.1231,
      "mean_token_accuracy": 0.1810326933860779,
      "num_tokens": 43074463.0,
      "step": 19660
    },
    {
      "entropy": 5.743797445297242,
      "epoch": 1.9516673283048829,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004622345533551655,
      "loss": 5.173,
      "mean_token_accuracy": 0.17116323560476304,
      "num_tokens": 43086004.0,
      "step": 19665
    },
    {
      "entropy": 5.671821975708008,
      "epoch": 1.9521635569670504,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004622149026658537,
      "loss": 5.0692,
      "mean_token_accuracy": 0.18262009024620057,
      "num_tokens": 43096973.0,
      "step": 19670
    },
    {
      "entropy": 5.67931809425354,
      "epoch": 1.9526597856292178,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004621952473340156,
      "loss": 5.1634,
      "mean_token_accuracy": 0.17587673664093018,
      "num_tokens": 43107337.0,
      "step": 19675
    },
    {
      "entropy": 5.655240535736084,
      "epoch": 1.9531560142913855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004621755873601388,
      "loss": 5.0853,
      "mean_token_accuracy": 0.18433144092559814,
      "num_tokens": 43118668.0,
      "step": 19680
    },
    {
      "entropy": 5.707370901107788,
      "epoch": 1.953652242953553,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004621559227447108,
      "loss": 5.1894,
      "mean_token_accuracy": 0.17193802148103715,
      "num_tokens": 43129030.0,
      "step": 19685
    },
    {
      "entropy": 5.56262526512146,
      "epoch": 1.9541484716157205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004621362534882192,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1818743735551834,
      "num_tokens": 43139120.0,
      "step": 19690
    },
    {
      "entropy": 5.639174938201904,
      "epoch": 1.9546447002778882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004621165795911517,
      "loss": 5.1339,
      "mean_token_accuracy": 0.17271141856908798,
      "num_tokens": 43152390.0,
      "step": 19695
    },
    {
      "entropy": 5.598840713500977,
      "epoch": 1.9551409289400556,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046209690105399623,
      "loss": 4.9352,
      "mean_token_accuracy": 0.1861257955431938,
      "num_tokens": 43162230.0,
      "step": 19700
    },
    {
      "entropy": 5.574537181854248,
      "epoch": 1.9556371576022231,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046207721787724085,
      "loss": 5.0715,
      "mean_token_accuracy": 0.17704373300075532,
      "num_tokens": 43172751.0,
      "step": 19705
    },
    {
      "entropy": 5.737267875671387,
      "epoch": 1.9561333862643906,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046205753006137347,
      "loss": 5.1726,
      "mean_token_accuracy": 0.17390822023153304,
      "num_tokens": 43183835.0,
      "step": 19710
    },
    {
      "entropy": 5.680675935745239,
      "epoch": 1.956629614926558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046203783760688245,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17555489540100097,
      "num_tokens": 43193178.0,
      "step": 19715
    },
    {
      "entropy": 5.656751918792724,
      "epoch": 1.9571258435887255,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004620181405142561,
      "loss": 5.203,
      "mean_token_accuracy": 0.16552540212869643,
      "num_tokens": 43204968.0,
      "step": 19720
    },
    {
      "entropy": 5.642568969726563,
      "epoch": 1.9576220722508932,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004619984387839829,
      "loss": 5.2182,
      "mean_token_accuracy": 0.17174778133630753,
      "num_tokens": 43214916.0,
      "step": 19725
    },
    {
      "entropy": 5.764087867736817,
      "epoch": 1.9581183009130607,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046197873241655133,
      "loss": 5.2055,
      "mean_token_accuracy": 0.17178250402212142,
      "num_tokens": 43226374.0,
      "step": 19730
    },
    {
      "entropy": 5.6472601890563965,
      "epoch": 1.9586145295752284,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004619590214124501,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1723868265748024,
      "num_tokens": 43237296.0,
      "step": 19735
    },
    {
      "entropy": 5.741288185119629,
      "epoch": 1.9591107582373959,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046193930577216797,
      "loss": 5.2428,
      "mean_token_accuracy": 0.16276839822530748,
      "num_tokens": 43248220.0,
      "step": 19740
    },
    {
      "entropy": 5.594349765777588,
      "epoch": 1.9596069868995634,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004619195854961939,
      "loss": 5.0168,
      "mean_token_accuracy": 0.18728587329387664,
      "num_tokens": 43259913.0,
      "step": 19745
    },
    {
      "entropy": 5.656796455383301,
      "epoch": 1.9601032155617308,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046189986058501687,
      "loss": 5.1389,
      "mean_token_accuracy": 0.16964245736598968,
      "num_tokens": 43272317.0,
      "step": 19750
    },
    {
      "entropy": 5.569743394851685,
      "epoch": 1.9605994442238983,
      "grad_norm": 1.078125,
      "learning_rate": 0.000461880131039126,
      "loss": 5.0317,
      "mean_token_accuracy": 0.17913611680269242,
      "num_tokens": 43283170.0,
      "step": 19755
    },
    {
      "entropy": 5.600822305679321,
      "epoch": 1.9610956728860658,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004618603968590107,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17718788236379623,
      "num_tokens": 43293899.0,
      "step": 19760
    },
    {
      "entropy": 5.6731456279754635,
      "epoch": 1.9615919015482333,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046184065804516007,
      "loss": 5.1394,
      "mean_token_accuracy": 0.17816756516695023,
      "num_tokens": 43305638.0,
      "step": 19765
    },
    {
      "entropy": 5.686150026321411,
      "epoch": 1.962088130210401,
      "grad_norm": 1.0,
      "learning_rate": 0.0004618209145980638,
      "loss": 5.2139,
      "mean_token_accuracy": 0.16742821782827377,
      "num_tokens": 43315900.0,
      "step": 19770
    },
    {
      "entropy": 5.707617282867432,
      "epoch": 1.9625843588725684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004618011665182114,
      "loss": 5.2619,
      "mean_token_accuracy": 0.16489529013633727,
      "num_tokens": 43328464.0,
      "step": 19775
    },
    {
      "entropy": 5.587302494049072,
      "epoch": 1.9630805875347361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004617814138060926,
      "loss": 4.9651,
      "mean_token_accuracy": 0.1862844333052635,
      "num_tokens": 43339303.0,
      "step": 19780
    },
    {
      "entropy": 5.705175971984863,
      "epoch": 1.9635768161969036,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004617616564621972,
      "loss": 5.1917,
      "mean_token_accuracy": 0.17002160102128983,
      "num_tokens": 43349269.0,
      "step": 19785
    },
    {
      "entropy": 5.588354253768921,
      "epoch": 1.964073044859071,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046174189448701516,
      "loss": 5.047,
      "mean_token_accuracy": 0.17804120630025863,
      "num_tokens": 43361090.0,
      "step": 19790
    },
    {
      "entropy": 5.591315698623657,
      "epoch": 1.9645692735212386,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004617221278810365,
      "loss": 5.0432,
      "mean_token_accuracy": 0.17720565646886827,
      "num_tokens": 43373448.0,
      "step": 19795
    },
    {
      "entropy": 5.7258875370025635,
      "epoch": 1.965065502183406,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046170235664475147,
      "loss": 5.1861,
      "mean_token_accuracy": 0.17273018062114714,
      "num_tokens": 43386202.0,
      "step": 19800
    },
    {
      "entropy": 5.6432960510253904,
      "epoch": 1.9655617308455735,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046168258077865026,
      "loss": 5.0867,
      "mean_token_accuracy": 0.17994345873594284,
      "num_tokens": 43396363.0,
      "step": 19805
    },
    {
      "entropy": 5.627947807312012,
      "epoch": 1.9660579595077412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004616628002832233,
      "loss": 5.005,
      "mean_token_accuracy": 0.18254876434803008,
      "num_tokens": 43406076.0,
      "step": 19810
    },
    {
      "entropy": 5.663095617294312,
      "epoch": 1.9665541881699087,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046164301515896124,
      "loss": 5.1649,
      "mean_token_accuracy": 0.17426572293043135,
      "num_tokens": 43416226.0,
      "step": 19815
    },
    {
      "entropy": 5.600831699371338,
      "epoch": 1.9670504168320764,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004616232254063544,
      "loss": 5.0006,
      "mean_token_accuracy": 0.18140294402837753,
      "num_tokens": 43427364.0,
      "step": 19820
    },
    {
      "entropy": 5.624504041671753,
      "epoch": 1.9675466454942439,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046160343102589384,
      "loss": 5.1179,
      "mean_token_accuracy": 0.18107366263866426,
      "num_tokens": 43437854.0,
      "step": 19825
    },
    {
      "entropy": 5.638732004165649,
      "epoch": 1.9680428741564113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004615836320180702,
      "loss": 5.1528,
      "mean_token_accuracy": 0.17727501541376114,
      "num_tokens": 43449370.0,
      "step": 19830
    },
    {
      "entropy": 5.613202905654907,
      "epoch": 1.9685391028185788,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046156382838337455,
      "loss": 5.0926,
      "mean_token_accuracy": 0.1839180201292038,
      "num_tokens": 43461631.0,
      "step": 19835
    },
    {
      "entropy": 5.6553307056427,
      "epoch": 1.9690353314807463,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046154402012229807,
      "loss": 5.072,
      "mean_token_accuracy": 0.17497395426034928,
      "num_tokens": 43472357.0,
      "step": 19840
    },
    {
      "entropy": 5.597118330001831,
      "epoch": 1.9695315601429138,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046152420723533175,
      "loss": 5.056,
      "mean_token_accuracy": 0.178684264421463,
      "num_tokens": 43483642.0,
      "step": 19845
    },
    {
      "entropy": 5.67268557548523,
      "epoch": 1.9700277888050812,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046150438972296706,
      "loss": 5.1309,
      "mean_token_accuracy": 0.1751613900065422,
      "num_tokens": 43495136.0,
      "step": 19850
    },
    {
      "entropy": 5.587759304046631,
      "epoch": 1.970524017467249,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046148456758569536,
      "loss": 5.1145,
      "mean_token_accuracy": 0.17365544587373732,
      "num_tokens": 43506376.0,
      "step": 19855
    },
    {
      "entropy": 5.634241724014283,
      "epoch": 1.9710202461294164,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046146474082400823,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1753728359937668,
      "num_tokens": 43516658.0,
      "step": 19860
    },
    {
      "entropy": 5.6450581550598145,
      "epoch": 1.971516474791584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046144490943839735,
      "loss": 5.0435,
      "mean_token_accuracy": 0.18048814833164215,
      "num_tokens": 43526907.0,
      "step": 19865
    },
    {
      "entropy": 5.7079297542572025,
      "epoch": 1.9720127034537516,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004614250734293544,
      "loss": 5.134,
      "mean_token_accuracy": 0.16916865408420562,
      "num_tokens": 43537739.0,
      "step": 19870
    },
    {
      "entropy": 5.586780452728272,
      "epoch": 1.972508932115919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046140523279737146,
      "loss": 5.1047,
      "mean_token_accuracy": 0.17753399759531022,
      "num_tokens": 43549240.0,
      "step": 19875
    },
    {
      "entropy": 5.634145784378052,
      "epoch": 1.9730051607780865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004613853875429403,
      "loss": 5.0611,
      "mean_token_accuracy": 0.17431079000234603,
      "num_tokens": 43559857.0,
      "step": 19880
    },
    {
      "entropy": 5.568658304214478,
      "epoch": 1.973501389440254,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004613655376665532,
      "loss": 5.0367,
      "mean_token_accuracy": 0.1755430966615677,
      "num_tokens": 43570262.0,
      "step": 19885
    },
    {
      "entropy": 5.6900359153747555,
      "epoch": 1.9739976181024215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004613456831687024,
      "loss": 5.1125,
      "mean_token_accuracy": 0.17293996512889862,
      "num_tokens": 43581965.0,
      "step": 19890
    },
    {
      "entropy": 5.637897729873657,
      "epoch": 1.9744938467645892,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046132582404988004,
      "loss": 5.0699,
      "mean_token_accuracy": 0.17860767394304275,
      "num_tokens": 43592728.0,
      "step": 19895
    },
    {
      "entropy": 5.647013711929321,
      "epoch": 1.9749900754267566,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046130596031057885,
      "loss": 5.0483,
      "mean_token_accuracy": 0.17283768057823182,
      "num_tokens": 43602437.0,
      "step": 19900
    },
    {
      "entropy": 5.62953896522522,
      "epoch": 1.9754863040889243,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046128609195129135,
      "loss": 5.1637,
      "mean_token_accuracy": 0.171910460293293,
      "num_tokens": 43613115.0,
      "step": 19905
    },
    {
      "entropy": 5.6511565208435055,
      "epoch": 1.9759825327510918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004612662189725101,
      "loss": 5.1419,
      "mean_token_accuracy": 0.1704648479819298,
      "num_tokens": 43623700.0,
      "step": 19910
    },
    {
      "entropy": 5.702182579040527,
      "epoch": 1.9764787614132593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000461246341374728,
      "loss": 5.1505,
      "mean_token_accuracy": 0.17151180729269982,
      "num_tokens": 43633712.0,
      "step": 19915
    },
    {
      "entropy": 5.676784420013428,
      "epoch": 1.9769749900754268,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046122645915843796,
      "loss": 5.1859,
      "mean_token_accuracy": 0.1747810184955597,
      "num_tokens": 43643705.0,
      "step": 19920
    },
    {
      "entropy": 5.649128770828247,
      "epoch": 1.9774712187375942,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046120657232413305,
      "loss": 5.1462,
      "mean_token_accuracy": 0.17623161822557448,
      "num_tokens": 43655376.0,
      "step": 19925
    },
    {
      "entropy": 5.6188414096832275,
      "epoch": 1.9779674473997617,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004611866808723064,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18351953625679016,
      "num_tokens": 43666329.0,
      "step": 19930
    },
    {
      "entropy": 5.570856523513794,
      "epoch": 1.9784636760619292,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046116678480345117,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17931071519851685,
      "num_tokens": 43676415.0,
      "step": 19935
    },
    {
      "entropy": 5.564628219604492,
      "epoch": 1.9789599047240969,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00046114688411806083,
      "loss": 5.0563,
      "mean_token_accuracy": 0.17990588992834092,
      "num_tokens": 43686997.0,
      "step": 19940
    },
    {
      "entropy": 5.602685594558716,
      "epoch": 1.9794561333862644,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00046112697881662895,
      "loss": 5.037,
      "mean_token_accuracy": 0.1800253540277481,
      "num_tokens": 43697342.0,
      "step": 19945
    },
    {
      "entropy": 5.6927553653717045,
      "epoch": 1.979952362048432,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00046110706889964905,
      "loss": 5.1538,
      "mean_token_accuracy": 0.17560568004846572,
      "num_tokens": 43707876.0,
      "step": 19950
    },
    {
      "entropy": 5.628222560882568,
      "epoch": 1.9804485907105995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004610871543676149,
      "loss": 5.009,
      "mean_token_accuracy": 0.1802981674671173,
      "num_tokens": 43718268.0,
      "step": 19955
    },
    {
      "entropy": 5.629111909866333,
      "epoch": 1.980944819372767,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046106723522102024,
      "loss": 5.1453,
      "mean_token_accuracy": 0.1690903589129448,
      "num_tokens": 43729853.0,
      "step": 19960
    },
    {
      "entropy": 5.638933086395264,
      "epoch": 1.9814410480349345,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046104731146035904,
      "loss": 5.1396,
      "mean_token_accuracy": 0.17534925192594528,
      "num_tokens": 43739886.0,
      "step": 19965
    },
    {
      "entropy": 5.669044017791748,
      "epoch": 1.981937276697102,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004610273830861255,
      "loss": 5.1262,
      "mean_token_accuracy": 0.17256870716810227,
      "num_tokens": 43750536.0,
      "step": 19970
    },
    {
      "entropy": 5.664005947113037,
      "epoch": 1.9824335053592694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004610074500988137,
      "loss": 5.0971,
      "mean_token_accuracy": 0.178105628490448,
      "num_tokens": 43763192.0,
      "step": 19975
    },
    {
      "entropy": 5.6602846622467045,
      "epoch": 1.9829297340214371,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004609875124989179,
      "loss": 5.1356,
      "mean_token_accuracy": 0.1757287785410881,
      "num_tokens": 43774117.0,
      "step": 19980
    },
    {
      "entropy": 5.637751340866089,
      "epoch": 1.9834259626836046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004609675702869326,
      "loss": 5.0795,
      "mean_token_accuracy": 0.17594913244247437,
      "num_tokens": 43784708.0,
      "step": 19985
    },
    {
      "entropy": 5.64250078201294,
      "epoch": 1.983922191345772,
      "grad_norm": 1.0,
      "learning_rate": 0.0004609476234633523,
      "loss": 5.1015,
      "mean_token_accuracy": 0.18091051727533342,
      "num_tokens": 43796767.0,
      "step": 19990
    },
    {
      "entropy": 5.6567394733428955,
      "epoch": 1.9844184200079398,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046092767202867156,
      "loss": 5.1959,
      "mean_token_accuracy": 0.1653544157743454,
      "num_tokens": 43807203.0,
      "step": 19995
    },
    {
      "entropy": 5.602111053466797,
      "epoch": 1.9849146486701073,
      "grad_norm": 2.234375,
      "learning_rate": 0.00046090771598338525,
      "loss": 5.1275,
      "mean_token_accuracy": 0.18043386489152907,
      "num_tokens": 43817123.0,
      "step": 20000
    },
    {
      "entropy": 5.653313207626343,
      "epoch": 1.9854108773322747,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004608877553279881,
      "loss": 5.0867,
      "mean_token_accuracy": 0.182343290746212,
      "num_tokens": 43828625.0,
      "step": 20005
    },
    {
      "entropy": 5.615716981887817,
      "epoch": 1.9859071059944422,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004608677900629753,
      "loss": 5.0611,
      "mean_token_accuracy": 0.18033138215541838,
      "num_tokens": 43840012.0,
      "step": 20010
    },
    {
      "entropy": 5.543627452850342,
      "epoch": 1.9864033346566097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046084782018884177,
      "loss": 5.0577,
      "mean_token_accuracy": 0.17489999383687974,
      "num_tokens": 43851253.0,
      "step": 20015
    },
    {
      "entropy": 5.673974466323853,
      "epoch": 1.9868995633187772,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046082784570608266,
      "loss": 5.1723,
      "mean_token_accuracy": 0.1755812495946884,
      "num_tokens": 43864389.0,
      "step": 20020
    },
    {
      "entropy": 5.660207462310791,
      "epoch": 1.9873957919809448,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046080786661519354,
      "loss": 5.162,
      "mean_token_accuracy": 0.1707623317837715,
      "num_tokens": 43876446.0,
      "step": 20025
    },
    {
      "entropy": 5.667764806747437,
      "epoch": 1.9878920206431123,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004607878829166696,
      "loss": 5.1155,
      "mean_token_accuracy": 0.16955841928720475,
      "num_tokens": 43888318.0,
      "step": 20030
    },
    {
      "entropy": 5.608180522918701,
      "epoch": 1.98838824930528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046076789461100653,
      "loss": 5.0524,
      "mean_token_accuracy": 0.18234322369098663,
      "num_tokens": 43898358.0,
      "step": 20035
    },
    {
      "entropy": 5.660120058059692,
      "epoch": 1.9888844779674475,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607479016987001,
      "loss": 5.1005,
      "mean_token_accuracy": 0.17964099496603012,
      "num_tokens": 43909831.0,
      "step": 20040
    },
    {
      "entropy": 5.664791297912598,
      "epoch": 1.989380706629615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004607279041802458,
      "loss": 5.0703,
      "mean_token_accuracy": 0.1711942195892334,
      "num_tokens": 43919648.0,
      "step": 20045
    },
    {
      "entropy": 5.635083770751953,
      "epoch": 1.9898769352917824,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00046070790205613967,
      "loss": 5.1284,
      "mean_token_accuracy": 0.17310467809438707,
      "num_tokens": 43930563.0,
      "step": 20050
    },
    {
      "entropy": 5.688690710067749,
      "epoch": 1.99037316395395,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046068789532687784,
      "loss": 5.191,
      "mean_token_accuracy": 0.16881656497716904,
      "num_tokens": 43942490.0,
      "step": 20055
    },
    {
      "entropy": 5.5842584609985355,
      "epoch": 1.9908693926161174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004606678839929563,
      "loss": 4.9716,
      "mean_token_accuracy": 0.18736052066087722,
      "num_tokens": 43952187.0,
      "step": 20060
    },
    {
      "entropy": 5.575282049179077,
      "epoch": 1.9913656212782849,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004606478680548713,
      "loss": 5.0401,
      "mean_token_accuracy": 0.18031959235668182,
      "num_tokens": 43964203.0,
      "step": 20065
    },
    {
      "entropy": 5.617230415344238,
      "epoch": 1.9918618499404526,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046062784751311917,
      "loss": 5.1194,
      "mean_token_accuracy": 0.17689718902111054,
      "num_tokens": 43974048.0,
      "step": 20070
    },
    {
      "entropy": 5.664203643798828,
      "epoch": 1.99235807860262,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046060782236819647,
      "loss": 5.1034,
      "mean_token_accuracy": 0.1775547206401825,
      "num_tokens": 43984469.0,
      "step": 20075
    },
    {
      "entropy": 5.617772150039673,
      "epoch": 1.9928543072647877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004605877926205996,
      "loss": 5.1218,
      "mean_token_accuracy": 0.1753300815820694,
      "num_tokens": 43995494.0,
      "step": 20080
    },
    {
      "entropy": 5.654769706726074,
      "epoch": 1.9933505359269552,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004605677582708256,
      "loss": 5.0541,
      "mean_token_accuracy": 0.17663255631923674,
      "num_tokens": 44005869.0,
      "step": 20085
    },
    {
      "entropy": 5.716827344894409,
      "epoch": 1.9938467645891227,
      "grad_norm": 1.0,
      "learning_rate": 0.0004605477193193709,
      "loss": 5.121,
      "mean_token_accuracy": 0.1704731047153473,
      "num_tokens": 44016956.0,
      "step": 20090
    },
    {
      "entropy": 5.672459363937378,
      "epoch": 1.9943429932512902,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004605276757667326,
      "loss": 5.1711,
      "mean_token_accuracy": 0.16537946462631226,
      "num_tokens": 44029524.0,
      "step": 20095
    },
    {
      "entropy": 5.648651456832885,
      "epoch": 1.9948392219134576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004605076276134078,
      "loss": 5.0663,
      "mean_token_accuracy": 0.17916521281003953,
      "num_tokens": 44039523.0,
      "step": 20100
    },
    {
      "entropy": 5.69873275756836,
      "epoch": 1.9953354505756251,
      "grad_norm": 1.125,
      "learning_rate": 0.00046048757485989345,
      "loss": 5.1667,
      "mean_token_accuracy": 0.17014862447977067,
      "num_tokens": 44050538.0,
      "step": 20105
    },
    {
      "entropy": 5.696758270263672,
      "epoch": 1.9958316792377928,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000460467517506687,
      "loss": 5.1587,
      "mean_token_accuracy": 0.16863725632429122,
      "num_tokens": 44061108.0,
      "step": 20110
    },
    {
      "entropy": 5.62265043258667,
      "epoch": 1.9963279078999603,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004604474555542858,
      "loss": 5.087,
      "mean_token_accuracy": 0.1802946224808693,
      "num_tokens": 44072706.0,
      "step": 20115
    },
    {
      "entropy": 5.654217910766602,
      "epoch": 1.996824136562128,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004604273890031872,
      "loss": 5.0942,
      "mean_token_accuracy": 0.1763904109597206,
      "num_tokens": 44082697.0,
      "step": 20120
    },
    {
      "entropy": 5.5824418544769285,
      "epoch": 1.9973203652242955,
      "grad_norm": 1.0,
      "learning_rate": 0.000460407317853889,
      "loss": 5.0708,
      "mean_token_accuracy": 0.1815418392419815,
      "num_tokens": 44093265.0,
      "step": 20125
    },
    {
      "entropy": 5.609546899795532,
      "epoch": 1.997816593886463,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004603872421068888,
      "loss": 5.0613,
      "mean_token_accuracy": 0.17682579755783082,
      "num_tokens": 44103535.0,
      "step": 20130
    },
    {
      "entropy": 5.632092094421386,
      "epoch": 1.9983128225486304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004603671617626845,
      "loss": 5.0464,
      "mean_token_accuracy": 0.18016596734523774,
      "num_tokens": 44113903.0,
      "step": 20135
    },
    {
      "entropy": 5.6257734298706055,
      "epoch": 1.9988090512107979,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046034707682177393,
      "loss": 5.0969,
      "mean_token_accuracy": 0.18011657446622847,
      "num_tokens": 44125148.0,
      "step": 20140
    },
    {
      "entropy": 5.711775112152099,
      "epoch": 1.9993052798729654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004603269872846554,
      "loss": 5.187,
      "mean_token_accuracy": 0.16472611874341964,
      "num_tokens": 44136563.0,
      "step": 20145
    },
    {
      "entropy": 5.74063286781311,
      "epoch": 1.9998015085351328,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046030689315182676,
      "loss": 5.1481,
      "mean_token_accuracy": 0.1765286684036255,
      "num_tokens": 44147521.0,
      "step": 20150
    },
    {
      "entropy": 5.70450177192688,
      "epoch": 2.0002977371973003,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004602867944237866,
      "loss": 5.1472,
      "mean_token_accuracy": 0.17878676056861878,
      "num_tokens": 44158977.0,
      "step": 20155
    },
    {
      "entropy": 5.66407585144043,
      "epoch": 2.0007939658594682,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046026669110103315,
      "loss": 5.1061,
      "mean_token_accuracy": 0.17746686786413193,
      "num_tokens": 44170986.0,
      "step": 20160
    },
    {
      "entropy": 5.718288421630859,
      "epoch": 2.0012901945216357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046024658318406493,
      "loss": 5.1566,
      "mean_token_accuracy": 0.17524030357599257,
      "num_tokens": 44180936.0,
      "step": 20165
    },
    {
      "entropy": 5.726045036315918,
      "epoch": 2.001786423183803,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004602264706733807,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17430340051651,
      "num_tokens": 44191741.0,
      "step": 20170
    },
    {
      "entropy": 5.617003488540649,
      "epoch": 2.0022826518459707,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004602063535694791,
      "loss": 5.0274,
      "mean_token_accuracy": 0.176580448448658,
      "num_tokens": 44202310.0,
      "step": 20175
    },
    {
      "entropy": 5.632394027709961,
      "epoch": 2.002778880508138,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000460186231872859,
      "loss": 5.0806,
      "mean_token_accuracy": 0.17930894643068312,
      "num_tokens": 44213514.0,
      "step": 20180
    },
    {
      "entropy": 5.711247777938842,
      "epoch": 2.0032751091703056,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00046016610558401936,
      "loss": 5.1189,
      "mean_token_accuracy": 0.17302849739789963,
      "num_tokens": 44224718.0,
      "step": 20185
    },
    {
      "entropy": 5.67397894859314,
      "epoch": 2.003771337832473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046014597470345935,
      "loss": 5.1267,
      "mean_token_accuracy": 0.1774274080991745,
      "num_tokens": 44235999.0,
      "step": 20190
    },
    {
      "entropy": 5.612166881561279,
      "epoch": 2.0042675664946406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004601258392316781,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1820056527853012,
      "num_tokens": 44246323.0,
      "step": 20195
    },
    {
      "entropy": 5.613177680969239,
      "epoch": 2.0047637951568085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046010569916917496,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1821150526404381,
      "num_tokens": 44256323.0,
      "step": 20200
    },
    {
      "entropy": 5.654539632797241,
      "epoch": 2.005260023818976,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004600855545164493,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18431341648101807,
      "num_tokens": 44266845.0,
      "step": 20205
    },
    {
      "entropy": 5.666074275970459,
      "epoch": 2.0057562524811434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004600654052740008,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17724825739860534,
      "num_tokens": 44277861.0,
      "step": 20210
    },
    {
      "entropy": 5.642697095870972,
      "epoch": 2.006252481143311,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004600452514423289,
      "loss": 5.0547,
      "mean_token_accuracy": 0.18387431204319,
      "num_tokens": 44288511.0,
      "step": 20215
    },
    {
      "entropy": 5.648847961425782,
      "epoch": 2.0067487098054784,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004600250930219336,
      "loss": 5.1479,
      "mean_token_accuracy": 0.17016663402318954,
      "num_tokens": 44299752.0,
      "step": 20220
    },
    {
      "entropy": 5.605523586273193,
      "epoch": 2.007244938467646,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046000493001331465,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18185025304555893,
      "num_tokens": 44309443.0,
      "step": 20225
    },
    {
      "entropy": 5.665759944915772,
      "epoch": 2.0077411671298133,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045998476241697214,
      "loss": 5.0304,
      "mean_token_accuracy": 0.1838818222284317,
      "num_tokens": 44320030.0,
      "step": 20230
    },
    {
      "entropy": 5.579647588729858,
      "epoch": 2.008237395791981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004599645902334061,
      "loss": 4.9465,
      "mean_token_accuracy": 0.1898818016052246,
      "num_tokens": 44330077.0,
      "step": 20235
    },
    {
      "entropy": 5.531774377822876,
      "epoch": 2.0087336244541483,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045994441346311677,
      "loss": 5.0137,
      "mean_token_accuracy": 0.18540925681591033,
      "num_tokens": 44340336.0,
      "step": 20240
    },
    {
      "entropy": 5.65589599609375,
      "epoch": 2.009229853116316,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004599242321066045,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18714803010225295,
      "num_tokens": 44350704.0,
      "step": 20245
    },
    {
      "entropy": 5.6567505359649655,
      "epoch": 2.0097260817784837,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045990404616436976,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18800253868103028,
      "num_tokens": 44360541.0,
      "step": 20250
    },
    {
      "entropy": 5.5464715480804445,
      "epoch": 2.010222310440651,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004598838556369131,
      "loss": 4.9579,
      "mean_token_accuracy": 0.1847516417503357,
      "num_tokens": 44372132.0,
      "step": 20255
    },
    {
      "entropy": 5.540609121322632,
      "epoch": 2.0107185391028186,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045986366052473525,
      "loss": 4.9545,
      "mean_token_accuracy": 0.18732175678014756,
      "num_tokens": 44383594.0,
      "step": 20260
    },
    {
      "entropy": 5.57135910987854,
      "epoch": 2.011214767764986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004598434608283369,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18478865325450897,
      "num_tokens": 44394101.0,
      "step": 20265
    },
    {
      "entropy": 5.651159143447876,
      "epoch": 2.0117109964271536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004598232565482191,
      "loss": 5.0937,
      "mean_token_accuracy": 0.18093925416469575,
      "num_tokens": 44405670.0,
      "step": 20270
    },
    {
      "entropy": 5.664268541336059,
      "epoch": 2.012207225089321,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045980304768488274,
      "loss": 5.064,
      "mean_token_accuracy": 0.18099830895662308,
      "num_tokens": 44417748.0,
      "step": 20275
    },
    {
      "entropy": 5.704509162902832,
      "epoch": 2.0127034537514885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045978283423882906,
      "loss": 5.038,
      "mean_token_accuracy": 0.17757360637187958,
      "num_tokens": 44429648.0,
      "step": 20280
    },
    {
      "entropy": 5.612417125701905,
      "epoch": 2.0131996824136564,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045976261621055915,
      "loss": 5.0371,
      "mean_token_accuracy": 0.17488421499729156,
      "num_tokens": 44440867.0,
      "step": 20285
    },
    {
      "entropy": 5.592725419998169,
      "epoch": 2.013695911075824,
      "grad_norm": 1.0,
      "learning_rate": 0.00045974239360057457,
      "loss": 5.0762,
      "mean_token_accuracy": 0.17659918516874312,
      "num_tokens": 44452981.0,
      "step": 20290
    },
    {
      "entropy": 5.6428056240081785,
      "epoch": 2.0141921397379914,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045972216640937676,
      "loss": 5.0059,
      "mean_token_accuracy": 0.18295723348855972,
      "num_tokens": 44464243.0,
      "step": 20295
    },
    {
      "entropy": 5.663775491714477,
      "epoch": 2.014688368400159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004597019346374672,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17584344893693923,
      "num_tokens": 44474376.0,
      "step": 20300
    },
    {
      "entropy": 5.603692054748535,
      "epoch": 2.0151845970623263,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004596816982853477,
      "loss": 4.9628,
      "mean_token_accuracy": 0.18845965713262558,
      "num_tokens": 44485686.0,
      "step": 20305
    },
    {
      "entropy": 5.519081735610962,
      "epoch": 2.015680825724494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045966145735352,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18540526777505875,
      "num_tokens": 44496096.0,
      "step": 20310
    },
    {
      "entropy": 5.680245208740234,
      "epoch": 2.0161770543866613,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045964121184248604,
      "loss": 5.0898,
      "mean_token_accuracy": 0.17758666425943376,
      "num_tokens": 44507555.0,
      "step": 20315
    },
    {
      "entropy": 5.669816160202027,
      "epoch": 2.0166732830488288,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004596209617527479,
      "loss": 5.0609,
      "mean_token_accuracy": 0.17458617985248565,
      "num_tokens": 44519288.0,
      "step": 20320
    },
    {
      "entropy": 5.696958732604981,
      "epoch": 2.0171695117109962,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004596007070848078,
      "loss": 5.0953,
      "mean_token_accuracy": 0.1691756159067154,
      "num_tokens": 44530146.0,
      "step": 20325
    },
    {
      "entropy": 5.65507664680481,
      "epoch": 2.017665740373164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045958044783916795,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18175897300243377,
      "num_tokens": 44541411.0,
      "step": 20330
    },
    {
      "entropy": 5.5920491218566895,
      "epoch": 2.0181619690353316,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004595601840163306,
      "loss": 5.0406,
      "mean_token_accuracy": 0.17596777975559236,
      "num_tokens": 44552556.0,
      "step": 20335
    },
    {
      "entropy": 5.617976379394531,
      "epoch": 2.018658197697499,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045953991561679855,
      "loss": 5.0586,
      "mean_token_accuracy": 0.18115630745887756,
      "num_tokens": 44563052.0,
      "step": 20340
    },
    {
      "entropy": 5.593982124328614,
      "epoch": 2.0191544263596666,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004595196426410742,
      "loss": 5.0107,
      "mean_token_accuracy": 0.17999665737152098,
      "num_tokens": 44574020.0,
      "step": 20345
    },
    {
      "entropy": 5.552698040008545,
      "epoch": 2.019650655021834,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045949936508966015,
      "loss": 4.9782,
      "mean_token_accuracy": 0.18209113776683808,
      "num_tokens": 44586261.0,
      "step": 20350
    },
    {
      "entropy": 5.638844680786133,
      "epoch": 2.0201468836840015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004594790829630596,
      "loss": 5.0243,
      "mean_token_accuracy": 0.17250214219093324,
      "num_tokens": 44596391.0,
      "step": 20355
    },
    {
      "entropy": 5.590059804916382,
      "epoch": 2.020643112346169,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004594587962617752,
      "loss": 4.985,
      "mean_token_accuracy": 0.18226446211338043,
      "num_tokens": 44606263.0,
      "step": 20360
    },
    {
      "entropy": 5.642559051513672,
      "epoch": 2.0211393410083365,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004594385049863102,
      "loss": 5.0415,
      "mean_token_accuracy": 0.1775316134095192,
      "num_tokens": 44617332.0,
      "step": 20365
    },
    {
      "entropy": 5.628878545761109,
      "epoch": 2.021635569670504,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045941820913716766,
      "loss": 5.0507,
      "mean_token_accuracy": 0.19042633324861527,
      "num_tokens": 44629784.0,
      "step": 20370
    },
    {
      "entropy": 5.705547618865967,
      "epoch": 2.022131798332672,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045939790871485093,
      "loss": 5.1599,
      "mean_token_accuracy": 0.17194160372018813,
      "num_tokens": 44641398.0,
      "step": 20375
    },
    {
      "entropy": 5.546917343139649,
      "epoch": 2.0226280269948393,
      "grad_norm": 1.0,
      "learning_rate": 0.00045937760371986334,
      "loss": 4.9122,
      "mean_token_accuracy": 0.19317369610071183,
      "num_tokens": 44652592.0,
      "step": 20380
    },
    {
      "entropy": 5.605185031890869,
      "epoch": 2.023124255657007,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045935729415270857,
      "loss": 4.9881,
      "mean_token_accuracy": 0.19511576741933823,
      "num_tokens": 44663896.0,
      "step": 20385
    },
    {
      "entropy": 5.719769620895386,
      "epoch": 2.0236204843191743,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004593369800138901,
      "loss": 5.1418,
      "mean_token_accuracy": 0.17238679677248,
      "num_tokens": 44675004.0,
      "step": 20390
    },
    {
      "entropy": 5.659353876113892,
      "epoch": 2.0241167129813418,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004593166613039118,
      "loss": 5.0932,
      "mean_token_accuracy": 0.17840562164783477,
      "num_tokens": 44686778.0,
      "step": 20395
    },
    {
      "entropy": 5.59436264038086,
      "epoch": 2.0246129416435092,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045929633802327743,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18352058976888658,
      "num_tokens": 44697600.0,
      "step": 20400
    },
    {
      "entropy": 5.612512111663818,
      "epoch": 2.0251091703056767,
      "grad_norm": 1.0625,
      "learning_rate": 0.000459276010172491,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17728497236967086,
      "num_tokens": 44707453.0,
      "step": 20405
    },
    {
      "entropy": 5.637515592575073,
      "epoch": 2.025605398967844,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045925567775205656,
      "loss": 5.1307,
      "mean_token_accuracy": 0.17490776032209396,
      "num_tokens": 44718351.0,
      "step": 20410
    },
    {
      "entropy": 5.6487936019897464,
      "epoch": 2.026101627630012,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045923534076247846,
      "loss": 4.9997,
      "mean_token_accuracy": 0.18585601598024368,
      "num_tokens": 44728884.0,
      "step": 20415
    },
    {
      "entropy": 5.6426972389221195,
      "epoch": 2.0265978562921796,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045921499920426083,
      "loss": 5.0388,
      "mean_token_accuracy": 0.17958557605743408,
      "num_tokens": 44740769.0,
      "step": 20420
    },
    {
      "entropy": 5.677074337005616,
      "epoch": 2.027094084954347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045919465307790817,
      "loss": 5.1457,
      "mean_token_accuracy": 0.175258132815361,
      "num_tokens": 44753094.0,
      "step": 20425
    },
    {
      "entropy": 5.532149887084961,
      "epoch": 2.0275903136165145,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045917430238392505,
      "loss": 4.9066,
      "mean_token_accuracy": 0.1933710217475891,
      "num_tokens": 44762996.0,
      "step": 20430
    },
    {
      "entropy": 5.569025993347168,
      "epoch": 2.028086542278682,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000459153947122816,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18218472599983215,
      "num_tokens": 44775343.0,
      "step": 20435
    },
    {
      "entropy": 5.556184530258179,
      "epoch": 2.0285827709408495,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004591335872950859,
      "loss": 4.9457,
      "mean_token_accuracy": 0.18577127009630204,
      "num_tokens": 44785383.0,
      "step": 20440
    },
    {
      "entropy": 5.595469570159912,
      "epoch": 2.029078999603017,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004591132229012397,
      "loss": 5.0041,
      "mean_token_accuracy": 0.18162795007228852,
      "num_tokens": 44796805.0,
      "step": 20445
    },
    {
      "entropy": 5.605247068405151,
      "epoch": 2.0295752282651844,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045909285394178217,
      "loss": 4.9939,
      "mean_token_accuracy": 0.18669559359550475,
      "num_tokens": 44806800.0,
      "step": 20450
    },
    {
      "entropy": 5.62912278175354,
      "epoch": 2.030071456927352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004590724804172187,
      "loss": 5.1346,
      "mean_token_accuracy": 0.16852396577596665,
      "num_tokens": 44818745.0,
      "step": 20455
    },
    {
      "entropy": 5.569318199157715,
      "epoch": 2.03056768558952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004590521023280543,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18120095878839493,
      "num_tokens": 44829825.0,
      "step": 20460
    },
    {
      "entropy": 5.605826902389526,
      "epoch": 2.0310639142516873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045903171967479425,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18504736423492432,
      "num_tokens": 44841742.0,
      "step": 20465
    },
    {
      "entropy": 5.683988380432129,
      "epoch": 2.031560142913855,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045901133245794426,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17978113442659377,
      "num_tokens": 44853019.0,
      "step": 20470
    },
    {
      "entropy": 5.6298322677612305,
      "epoch": 2.0320563715760223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004589909406780096,
      "loss": 4.9743,
      "mean_token_accuracy": 0.18350158333778382,
      "num_tokens": 44863147.0,
      "step": 20475
    },
    {
      "entropy": 5.6404155731201175,
      "epoch": 2.0325526002381897,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045897054433549617,
      "loss": 5.0243,
      "mean_token_accuracy": 0.1792710915207863,
      "num_tokens": 44873638.0,
      "step": 20480
    },
    {
      "entropy": 5.587981510162353,
      "epoch": 2.033048828900357,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045895014343090963,
      "loss": 5.0384,
      "mean_token_accuracy": 0.17848518192768098,
      "num_tokens": 44885752.0,
      "step": 20485
    },
    {
      "entropy": 5.570626735687256,
      "epoch": 2.0335450575625247,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045892973796475594,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18263436406850814,
      "num_tokens": 44897471.0,
      "step": 20490
    },
    {
      "entropy": 5.694743204116821,
      "epoch": 2.034041286224692,
      "grad_norm": 1.125,
      "learning_rate": 0.000458909327937541,
      "loss": 5.0742,
      "mean_token_accuracy": 0.17367905080318452,
      "num_tokens": 44907236.0,
      "step": 20495
    },
    {
      "entropy": 5.61212420463562,
      "epoch": 2.03453751488686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045888891334977105,
      "loss": 5.007,
      "mean_token_accuracy": 0.18512410223484038,
      "num_tokens": 44918103.0,
      "step": 20500
    },
    {
      "entropy": 5.576538562774658,
      "epoch": 2.0350337435490276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004588684942019524,
      "loss": 5.0552,
      "mean_token_accuracy": 0.1798553869128227,
      "num_tokens": 44928263.0,
      "step": 20505
    },
    {
      "entropy": 5.586361598968506,
      "epoch": 2.035529972211195,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045884807049459116,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18072873800992967,
      "num_tokens": 44939487.0,
      "step": 20510
    },
    {
      "entropy": 5.64599380493164,
      "epoch": 2.0360262008733625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000458827642228194,
      "loss": 5.1027,
      "mean_token_accuracy": 0.17479217797517776,
      "num_tokens": 44949749.0,
      "step": 20515
    },
    {
      "entropy": 5.609309101104737,
      "epoch": 2.03652242953553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045880720940326744,
      "loss": 5.0346,
      "mean_token_accuracy": 0.18232240229845048,
      "num_tokens": 44960347.0,
      "step": 20520
    },
    {
      "entropy": 5.661993408203125,
      "epoch": 2.0370186581976975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045878677202031804,
      "loss": 5.107,
      "mean_token_accuracy": 0.1735198274254799,
      "num_tokens": 44970420.0,
      "step": 20525
    },
    {
      "entropy": 5.629017496109009,
      "epoch": 2.037514886859865,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004587663300798529,
      "loss": 5.0335,
      "mean_token_accuracy": 0.17656304836273193,
      "num_tokens": 44980888.0,
      "step": 20530
    },
    {
      "entropy": 5.646453905105591,
      "epoch": 2.0380111155220324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004587458835823786,
      "loss": 5.0426,
      "mean_token_accuracy": 0.1770864799618721,
      "num_tokens": 44992721.0,
      "step": 20535
    },
    {
      "entropy": 5.62697958946228,
      "epoch": 2.0385073441842,
      "grad_norm": 1.125,
      "learning_rate": 0.00045872543252840236,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18534399271011354,
      "num_tokens": 45003603.0,
      "step": 20540
    },
    {
      "entropy": 5.600315999984741,
      "epoch": 2.039003572846368,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045870497691843134,
      "loss": 5.0363,
      "mean_token_accuracy": 0.19039897322654725,
      "num_tokens": 45014864.0,
      "step": 20545
    },
    {
      "entropy": 5.651213455200195,
      "epoch": 2.0394998015085353,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004586845167529727,
      "loss": 5.0653,
      "mean_token_accuracy": 0.17903974652290344,
      "num_tokens": 45027766.0,
      "step": 20550
    },
    {
      "entropy": 5.598121643066406,
      "epoch": 2.0399960301707027,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00045866405203253384,
      "loss": 5.0379,
      "mean_token_accuracy": 0.175766059756279,
      "num_tokens": 45039524.0,
      "step": 20555
    },
    {
      "entropy": 5.620949697494507,
      "epoch": 2.04049225883287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004586435827576223,
      "loss": 5.06,
      "mean_token_accuracy": 0.17629922777414322,
      "num_tokens": 45050848.0,
      "step": 20560
    },
    {
      "entropy": 5.628915071487427,
      "epoch": 2.0409884874950377,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004586231089287456,
      "loss": 4.9937,
      "mean_token_accuracy": 0.17865631431341172,
      "num_tokens": 45061229.0,
      "step": 20565
    },
    {
      "entropy": 5.569423723220825,
      "epoch": 2.041484716157205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004586026305464115,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18567425310611724,
      "num_tokens": 45070466.0,
      "step": 20570
    },
    {
      "entropy": 5.614850425720215,
      "epoch": 2.0419809448193726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045858214761112777,
      "loss": 5.052,
      "mean_token_accuracy": 0.17881349772214888,
      "num_tokens": 45081839.0,
      "step": 20575
    },
    {
      "entropy": 5.673265361785889,
      "epoch": 2.04247717348154,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045856166012340235,
      "loss": 5.1123,
      "mean_token_accuracy": 0.17548524737358093,
      "num_tokens": 45092458.0,
      "step": 20580
    },
    {
      "entropy": 5.678771686553955,
      "epoch": 2.0429734021437076,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004585411680837433,
      "loss": 5.0892,
      "mean_token_accuracy": 0.17559907734394073,
      "num_tokens": 45104629.0,
      "step": 20585
    },
    {
      "entropy": 5.647320318222046,
      "epoch": 2.0434696308058755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045852067149265875,
      "loss": 4.946,
      "mean_token_accuracy": 0.18219396024942397,
      "num_tokens": 45115556.0,
      "step": 20590
    },
    {
      "entropy": 5.593822383880616,
      "epoch": 2.043965859468043,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000458500170350657,
      "loss": 5.026,
      "mean_token_accuracy": 0.17745040655136107,
      "num_tokens": 45126489.0,
      "step": 20595
    },
    {
      "entropy": 5.635648107528686,
      "epoch": 2.0444620881302105,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045847966465824654,
      "loss": 5.1424,
      "mean_token_accuracy": 0.17581753730773925,
      "num_tokens": 45138882.0,
      "step": 20600
    },
    {
      "entropy": 5.6749122619628904,
      "epoch": 2.044958316792378,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004584591544159356,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1826891928911209,
      "num_tokens": 45149941.0,
      "step": 20605
    },
    {
      "entropy": 5.721663904190064,
      "epoch": 2.0454545454545454,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004584386396242331,
      "loss": 5.0942,
      "mean_token_accuracy": 0.18032219707965852,
      "num_tokens": 45162267.0,
      "step": 20610
    },
    {
      "entropy": 5.667798471450806,
      "epoch": 2.045950774116713,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045841812028364756,
      "loss": 5.1805,
      "mean_token_accuracy": 0.17585235685110093,
      "num_tokens": 45174444.0,
      "step": 20615
    },
    {
      "entropy": 5.607372188568116,
      "epoch": 2.0464470027788804,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045839759639468774,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18283371031284332,
      "num_tokens": 45183957.0,
      "step": 20620
    },
    {
      "entropy": 5.662184143066407,
      "epoch": 2.046943231441048,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004583770679578628,
      "loss": 5.1249,
      "mean_token_accuracy": 0.17979470193386077,
      "num_tokens": 45195607.0,
      "step": 20625
    },
    {
      "entropy": 5.599042987823486,
      "epoch": 2.0474394601032158,
      "grad_norm": 1.125,
      "learning_rate": 0.0004583565349736818,
      "loss": 5.0045,
      "mean_token_accuracy": 0.18107030987739564,
      "num_tokens": 45205627.0,
      "step": 20630
    },
    {
      "entropy": 5.560740518569946,
      "epoch": 2.0479356887653832,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004583359974426538,
      "loss": 5.0627,
      "mean_token_accuracy": 0.17532941848039627,
      "num_tokens": 45216331.0,
      "step": 20635
    },
    {
      "entropy": 5.595874786376953,
      "epoch": 2.0484319174275507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045831545536528806,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17037209272384643,
      "num_tokens": 45227826.0,
      "step": 20640
    },
    {
      "entropy": 5.629573726654053,
      "epoch": 2.048928146089718,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045829490874209404,
      "loss": 5.0148,
      "mean_token_accuracy": 0.17951969504356385,
      "num_tokens": 45238093.0,
      "step": 20645
    },
    {
      "entropy": 5.598856401443482,
      "epoch": 2.0494243747518857,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045827435757358136,
      "loss": 5.0406,
      "mean_token_accuracy": 0.1843299984931946,
      "num_tokens": 45248660.0,
      "step": 20650
    },
    {
      "entropy": 5.670406770706177,
      "epoch": 2.049920603414053,
      "grad_norm": 1.0,
      "learning_rate": 0.0004582538018602594,
      "loss": 5.1054,
      "mean_token_accuracy": 0.17371099591255187,
      "num_tokens": 45261059.0,
      "step": 20655
    },
    {
      "entropy": 5.629499197006226,
      "epoch": 2.0504168320762206,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004582332416026381,
      "loss": 5.0247,
      "mean_token_accuracy": 0.17426238656044007,
      "num_tokens": 45272210.0,
      "step": 20660
    },
    {
      "entropy": 5.661193084716797,
      "epoch": 2.050913060738388,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045821267680122725,
      "loss": 5.1013,
      "mean_token_accuracy": 0.18157029300928115,
      "num_tokens": 45283488.0,
      "step": 20665
    },
    {
      "entropy": 5.609155225753784,
      "epoch": 2.0514092894005556,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004581921074565368,
      "loss": 4.9519,
      "mean_token_accuracy": 0.18564700037240983,
      "num_tokens": 45292824.0,
      "step": 20670
    },
    {
      "entropy": 5.580414819717407,
      "epoch": 2.0519055180627235,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004581715335690768,
      "loss": 5.0528,
      "mean_token_accuracy": 0.183115254342556,
      "num_tokens": 45303102.0,
      "step": 20675
    },
    {
      "entropy": 5.598613309860229,
      "epoch": 2.052401746724891,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045815095513935747,
      "loss": 4.9714,
      "mean_token_accuracy": 0.17831853181123733,
      "num_tokens": 45313825.0,
      "step": 20680
    },
    {
      "entropy": 5.606683826446533,
      "epoch": 2.0528979753870584,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004581303721678891,
      "loss": 4.9982,
      "mean_token_accuracy": 0.18850478082895278,
      "num_tokens": 45325789.0,
      "step": 20685
    },
    {
      "entropy": 5.634306287765503,
      "epoch": 2.053394204049226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004581097846551821,
      "loss": 5.0895,
      "mean_token_accuracy": 0.17321229577064515,
      "num_tokens": 45337630.0,
      "step": 20690
    },
    {
      "entropy": 5.67279748916626,
      "epoch": 2.0538904327113934,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000458089192601747,
      "loss": 5.1007,
      "mean_token_accuracy": 0.17317246198654174,
      "num_tokens": 45349815.0,
      "step": 20695
    },
    {
      "entropy": 5.613268184661865,
      "epoch": 2.054386661373561,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045806859600809455,
      "loss": 5.0466,
      "mean_token_accuracy": 0.1784007117152214,
      "num_tokens": 45360741.0,
      "step": 20700
    },
    {
      "entropy": 5.538904142379761,
      "epoch": 2.0548828900357283,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004580479948747353,
      "loss": 4.9173,
      "mean_token_accuracy": 0.19041681736707688,
      "num_tokens": 45371258.0,
      "step": 20705
    },
    {
      "entropy": 5.627281522750854,
      "epoch": 2.055379118697896,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045802738920218015,
      "loss": 5.0416,
      "mean_token_accuracy": 0.17804782539606095,
      "num_tokens": 45381914.0,
      "step": 20710
    },
    {
      "entropy": 5.584945011138916,
      "epoch": 2.0558753473600637,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004580067789909403,
      "loss": 5.0562,
      "mean_token_accuracy": 0.18035735487937926,
      "num_tokens": 45393159.0,
      "step": 20715
    },
    {
      "entropy": 5.588097143173218,
      "epoch": 2.056371576022231,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045798616424152654,
      "loss": 4.942,
      "mean_token_accuracy": 0.18653278350830077,
      "num_tokens": 45403076.0,
      "step": 20720
    },
    {
      "entropy": 5.616199350357055,
      "epoch": 2.0568678046843987,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045796554495445023,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17905967831611633,
      "num_tokens": 45412962.0,
      "step": 20725
    },
    {
      "entropy": 5.623038005828858,
      "epoch": 2.057364033346566,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004579449211302228,
      "loss": 5.0104,
      "mean_token_accuracy": 0.1843923807144165,
      "num_tokens": 45423620.0,
      "step": 20730
    },
    {
      "entropy": 5.622395038604736,
      "epoch": 2.0578602620087336,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004579242927693554,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1718085989356041,
      "num_tokens": 45435901.0,
      "step": 20735
    },
    {
      "entropy": 5.608104276657104,
      "epoch": 2.058356490670901,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004579036598723598,
      "loss": 4.9654,
      "mean_token_accuracy": 0.1904067263007164,
      "num_tokens": 45445409.0,
      "step": 20740
    },
    {
      "entropy": 5.683897256851196,
      "epoch": 2.0588527193330686,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045788302243974745,
      "loss": 5.1154,
      "mean_token_accuracy": 0.18147664666175842,
      "num_tokens": 45455794.0,
      "step": 20745
    },
    {
      "entropy": 5.623568534851074,
      "epoch": 2.059348947995236,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004578623804720302,
      "loss": 5.0091,
      "mean_token_accuracy": 0.18202087730169297,
      "num_tokens": 45465922.0,
      "step": 20750
    },
    {
      "entropy": 5.571804666519165,
      "epoch": 2.0598451766574035,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004578417339697201,
      "loss": 5.111,
      "mean_token_accuracy": 0.17677166163921357,
      "num_tokens": 45476314.0,
      "step": 20755
    },
    {
      "entropy": 5.640298557281494,
      "epoch": 2.0603414053195714,
      "grad_norm": 1.125,
      "learning_rate": 0.00045782108293332884,
      "loss": 5.0358,
      "mean_token_accuracy": 0.17730555385351182,
      "num_tokens": 45487232.0,
      "step": 20760
    },
    {
      "entropy": 5.705715894699097,
      "epoch": 2.060837633981739,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004578004273633687,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1772255077958107,
      "num_tokens": 45498368.0,
      "step": 20765
    },
    {
      "entropy": 5.620805454254151,
      "epoch": 2.0613338626439064,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045777976726035194,
      "loss": 5.0289,
      "mean_token_accuracy": 0.17745719254016876,
      "num_tokens": 45507600.0,
      "step": 20770
    },
    {
      "entropy": 5.643569087982177,
      "epoch": 2.061830091306074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004577591026247907,
      "loss": 5.1214,
      "mean_token_accuracy": 0.17317403107881546,
      "num_tokens": 45518594.0,
      "step": 20775
    },
    {
      "entropy": 5.608948373794556,
      "epoch": 2.0623263199682413,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004577384334571976,
      "loss": 4.9956,
      "mean_token_accuracy": 0.18250137716531753,
      "num_tokens": 45529929.0,
      "step": 20780
    },
    {
      "entropy": 5.609036207199097,
      "epoch": 2.062822548630409,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004577177597580851,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18391760140657426,
      "num_tokens": 45540228.0,
      "step": 20785
    },
    {
      "entropy": 5.581872081756591,
      "epoch": 2.0633187772925763,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045769708152796593,
      "loss": 4.9821,
      "mean_token_accuracy": 0.18293698579072953,
      "num_tokens": 45552026.0,
      "step": 20790
    },
    {
      "entropy": 5.553340005874634,
      "epoch": 2.0638150059547438,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045767639876735276,
      "loss": 5.0536,
      "mean_token_accuracy": 0.1836379960179329,
      "num_tokens": 45563579.0,
      "step": 20795
    },
    {
      "entropy": 5.677938652038574,
      "epoch": 2.0643112346169117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004576557114767585,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1738264963030815,
      "num_tokens": 45573800.0,
      "step": 20800
    },
    {
      "entropy": 5.581170463562012,
      "epoch": 2.064807463279079,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045763501965669627,
      "loss": 5.0012,
      "mean_token_accuracy": 0.18033675253391265,
      "num_tokens": 45583758.0,
      "step": 20805
    },
    {
      "entropy": 5.647362136840821,
      "epoch": 2.0653036919412466,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000457614323307679,
      "loss": 5.0022,
      "mean_token_accuracy": 0.1843077689409256,
      "num_tokens": 45595154.0,
      "step": 20810
    },
    {
      "entropy": 5.706191635131836,
      "epoch": 2.065799920603414,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045759362243022,
      "loss": 5.1119,
      "mean_token_accuracy": 0.1742821604013443,
      "num_tokens": 45606557.0,
      "step": 20815
    },
    {
      "entropy": 5.562309265136719,
      "epoch": 2.0662961492655816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575729170248327,
      "loss": 4.9612,
      "mean_token_accuracy": 0.18902714252471925,
      "num_tokens": 45615531.0,
      "step": 20820
    },
    {
      "entropy": 5.535184240341186,
      "epoch": 2.066792377927749,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575522070920305,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18435661643743514,
      "num_tokens": 45626512.0,
      "step": 20825
    },
    {
      "entropy": 5.6708731174469,
      "epoch": 2.0672886065899165,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045753149263232687,
      "loss": 5.0935,
      "mean_token_accuracy": 0.17396873533725737,
      "num_tokens": 45638005.0,
      "step": 20830
    },
    {
      "entropy": 5.6525230884552,
      "epoch": 2.067784835252084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004575107736462355,
      "loss": 5.0809,
      "mean_token_accuracy": 0.1749485045671463,
      "num_tokens": 45647740.0,
      "step": 20835
    },
    {
      "entropy": 5.570057010650634,
      "epoch": 2.0682810639142515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004574900501342703,
      "loss": 5.0086,
      "mean_token_accuracy": 0.1860436737537384,
      "num_tokens": 45659021.0,
      "step": 20840
    },
    {
      "entropy": 5.61150369644165,
      "epoch": 2.0687772925764194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004574693220969451,
      "loss": 5.0833,
      "mean_token_accuracy": 0.1769574448466301,
      "num_tokens": 45669413.0,
      "step": 20845
    },
    {
      "entropy": 5.704837846755981,
      "epoch": 2.069273521238587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004574485895347738,
      "loss": 5.135,
      "mean_token_accuracy": 0.17580965161323547,
      "num_tokens": 45680719.0,
      "step": 20850
    },
    {
      "entropy": 5.5631993293762205,
      "epoch": 2.0697697499007544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045742785244827066,
      "loss": 4.9704,
      "mean_token_accuracy": 0.1825826033949852,
      "num_tokens": 45690393.0,
      "step": 20855
    },
    {
      "entropy": 5.591906213760376,
      "epoch": 2.070265978562922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004574071108379499,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17424320727586745,
      "num_tokens": 45701304.0,
      "step": 20860
    },
    {
      "entropy": 5.607755088806153,
      "epoch": 2.0707622072250893,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004573863647043259,
      "loss": 4.9452,
      "mean_token_accuracy": 0.1852981686592102,
      "num_tokens": 45711892.0,
      "step": 20865
    },
    {
      "entropy": 5.616005611419678,
      "epoch": 2.0712584358872568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004573656140479129,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18020918369293212,
      "num_tokens": 45723693.0,
      "step": 20870
    },
    {
      "entropy": 5.577564907073975,
      "epoch": 2.0717546645494243,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004573448588692258,
      "loss": 5.0092,
      "mean_token_accuracy": 0.17858051806688308,
      "num_tokens": 45734392.0,
      "step": 20875
    },
    {
      "entropy": 5.513390588760376,
      "epoch": 2.0722508932115917,
      "grad_norm": 1.109375,
      "learning_rate": 0.000457324099168779,
      "loss": 4.9394,
      "mean_token_accuracy": 0.19076876491308212,
      "num_tokens": 45745732.0,
      "step": 20880
    },
    {
      "entropy": 5.575284957885742,
      "epoch": 2.0727471218737596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004573033349470874,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18706925213336945,
      "num_tokens": 45756671.0,
      "step": 20885
    },
    {
      "entropy": 5.677130937576294,
      "epoch": 2.073243350535927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004572825662046661,
      "loss": 5.1095,
      "mean_token_accuracy": 0.17474198639392852,
      "num_tokens": 45767685.0,
      "step": 20890
    },
    {
      "entropy": 5.6809868812561035,
      "epoch": 2.0737395791980946,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004572617929420298,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18168893456459045,
      "num_tokens": 45777251.0,
      "step": 20895
    },
    {
      "entropy": 5.687027931213379,
      "epoch": 2.074235807860262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045724101515969375,
      "loss": 5.0804,
      "mean_token_accuracy": 0.17958610951900483,
      "num_tokens": 45790037.0,
      "step": 20900
    },
    {
      "entropy": 5.635348129272461,
      "epoch": 2.0747320365224295,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045722023285817327,
      "loss": 5.1194,
      "mean_token_accuracy": 0.17006149441003798,
      "num_tokens": 45800803.0,
      "step": 20905
    },
    {
      "entropy": 5.662548303604126,
      "epoch": 2.075228265184597,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004571994460379837,
      "loss": 5.0585,
      "mean_token_accuracy": 0.17162808924913406,
      "num_tokens": 45812447.0,
      "step": 20910
    },
    {
      "entropy": 5.64333667755127,
      "epoch": 2.0757244938467645,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045717865469964045,
      "loss": 5.0175,
      "mean_token_accuracy": 0.178761288523674,
      "num_tokens": 45822172.0,
      "step": 20915
    },
    {
      "entropy": 5.7101054191589355,
      "epoch": 2.076220722508932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045715785884365914,
      "loss": 5.1579,
      "mean_token_accuracy": 0.17317310422658921,
      "num_tokens": 45832583.0,
      "step": 20920
    },
    {
      "entropy": 5.693865728378296,
      "epoch": 2.0767169511710994,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045713705847055537,
      "loss": 5.1598,
      "mean_token_accuracy": 0.17798333913087844,
      "num_tokens": 45842640.0,
      "step": 20925
    },
    {
      "entropy": 5.687538957595825,
      "epoch": 2.0772131798332674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004571162535808451,
      "loss": 5.1495,
      "mean_token_accuracy": 0.1716572657227516,
      "num_tokens": 45854618.0,
      "step": 20930
    },
    {
      "entropy": 5.631864881515503,
      "epoch": 2.077709408495435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004570954441750441,
      "loss": 5.033,
      "mean_token_accuracy": 0.18153142631053926,
      "num_tokens": 45866130.0,
      "step": 20935
    },
    {
      "entropy": 5.667792224884034,
      "epoch": 2.0782056371576023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004570746302536685,
      "loss": 5.066,
      "mean_token_accuracy": 0.17777009755373002,
      "num_tokens": 45877187.0,
      "step": 20940
    },
    {
      "entropy": 5.612111139297485,
      "epoch": 2.07870186581977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045705381181723444,
      "loss": 5.0836,
      "mean_token_accuracy": 0.17681358456611634,
      "num_tokens": 45889557.0,
      "step": 20945
    },
    {
      "entropy": 5.635267114639282,
      "epoch": 2.0791980944819373,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045703298886625804,
      "loss": 5.0827,
      "mean_token_accuracy": 0.1737922817468643,
      "num_tokens": 45900860.0,
      "step": 20950
    },
    {
      "entropy": 5.635767078399658,
      "epoch": 2.0796943231441047,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004570121614012558,
      "loss": 5.0928,
      "mean_token_accuracy": 0.17680176496505737,
      "num_tokens": 45912231.0,
      "step": 20955
    },
    {
      "entropy": 5.690398025512695,
      "epoch": 2.080190551806272,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045699132942274425,
      "loss": 5.1401,
      "mean_token_accuracy": 0.17586998045444488,
      "num_tokens": 45922722.0,
      "step": 20960
    },
    {
      "entropy": 5.612362813949585,
      "epoch": 2.0806867804684397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045697049293123984,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18248387277126313,
      "num_tokens": 45932444.0,
      "step": 20965
    },
    {
      "entropy": 5.666155099868774,
      "epoch": 2.0811830091306076,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004569496519272593,
      "loss": 5.0563,
      "mean_token_accuracy": 0.1782078415155411,
      "num_tokens": 45942826.0,
      "step": 20970
    },
    {
      "entropy": 5.651955795288086,
      "epoch": 2.081679237792775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045692880641131945,
      "loss": 5.0532,
      "mean_token_accuracy": 0.18117749989032744,
      "num_tokens": 45953935.0,
      "step": 20975
    },
    {
      "entropy": 5.696436405181885,
      "epoch": 2.0821754664549426,
      "grad_norm": 1.0,
      "learning_rate": 0.0004569079563839372,
      "loss": 5.1474,
      "mean_token_accuracy": 0.16810794025659562,
      "num_tokens": 45964693.0,
      "step": 20980
    },
    {
      "entropy": 5.700645923614502,
      "epoch": 2.08267169511711,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045688710184562965,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17641431242227554,
      "num_tokens": 45976481.0,
      "step": 20985
    },
    {
      "entropy": 5.671458387374878,
      "epoch": 2.0831679237792775,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045686624279691387,
      "loss": 5.1304,
      "mean_token_accuracy": 0.18055204749107362,
      "num_tokens": 45988879.0,
      "step": 20990
    },
    {
      "entropy": 5.586560916900635,
      "epoch": 2.083664152441445,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004568453792383072,
      "loss": 5.016,
      "mean_token_accuracy": 0.18408265411853791,
      "num_tokens": 45999332.0,
      "step": 20995
    },
    {
      "entropy": 5.621128034591675,
      "epoch": 2.0841603811036125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000456824511170327,
      "loss": 5.0412,
      "mean_token_accuracy": 0.1780989333987236,
      "num_tokens": 46009841.0,
      "step": 21000
    },
    {
      "epoch": 2.0841603811036125,
      "eval_entropy": 5.394302048459985,
      "eval_loss": 5.181803226470947,
      "eval_mean_token_accuracy": 0.1786847699377805,
      "eval_num_tokens": 46009841.0,
      "eval_runtime": 20.5583,
      "eval_samples_per_second": 1637.733,
      "eval_steps_per_second": 204.735,
      "step": 21000
    },
    {
      "entropy": 5.583067178726196,
      "epoch": 2.08465660976578,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004568036385934906,
      "loss": 4.9845,
      "mean_token_accuracy": 0.18094138205051422,
      "num_tokens": 46020678.0,
      "step": 21005
    },
    {
      "entropy": 5.6595344066619875,
      "epoch": 2.0851528384279474,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045678276150831586,
      "loss": 5.0628,
      "mean_token_accuracy": 0.17687734663486482,
      "num_tokens": 46032036.0,
      "step": 21010
    },
    {
      "entropy": 5.668038845062256,
      "epoch": 2.0856490670901153,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004567618799153202,
      "loss": 5.0617,
      "mean_token_accuracy": 0.18117256760597228,
      "num_tokens": 46042675.0,
      "step": 21015
    },
    {
      "entropy": 5.559159469604492,
      "epoch": 2.086145295752283,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045674099381502167,
      "loss": 5.0607,
      "mean_token_accuracy": 0.18003827184438706,
      "num_tokens": 46053616.0,
      "step": 21020
    },
    {
      "entropy": 5.645719289779663,
      "epoch": 2.0866415244144503,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004567201032079381,
      "loss": 5.074,
      "mean_token_accuracy": 0.18364328444004058,
      "num_tokens": 46063755.0,
      "step": 21025
    },
    {
      "entropy": 5.673125314712524,
      "epoch": 2.0871377530766178,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045669920809458754,
      "loss": 5.0749,
      "mean_token_accuracy": 0.17694755494594575,
      "num_tokens": 46074459.0,
      "step": 21030
    },
    {
      "entropy": 5.718306159973144,
      "epoch": 2.0876339817387852,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004566783084754881,
      "loss": 5.13,
      "mean_token_accuracy": 0.17814599275588988,
      "num_tokens": 46085053.0,
      "step": 21035
    },
    {
      "entropy": 5.610039854049683,
      "epoch": 2.0881302104009527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045665740435115815,
      "loss": 4.957,
      "mean_token_accuracy": 0.19061294049024582,
      "num_tokens": 46094697.0,
      "step": 21040
    },
    {
      "entropy": 5.616088342666626,
      "epoch": 2.08862643906312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000456636495722116,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1733097866177559,
      "num_tokens": 46105617.0,
      "step": 21045
    },
    {
      "entropy": 5.681585359573364,
      "epoch": 2.0891226677252877,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004566155825888801,
      "loss": 5.0406,
      "mean_token_accuracy": 0.18822656869888305,
      "num_tokens": 46115450.0,
      "step": 21050
    },
    {
      "entropy": 5.637346029281616,
      "epoch": 2.089618896387455,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004565946649519691,
      "loss": 5.1077,
      "mean_token_accuracy": 0.17794696390628814,
      "num_tokens": 46126511.0,
      "step": 21055
    },
    {
      "entropy": 5.625922060012817,
      "epoch": 2.090115125049623,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045657374281190177,
      "loss": 5.008,
      "mean_token_accuracy": 0.18406940549612044,
      "num_tokens": 46136294.0,
      "step": 21060
    },
    {
      "entropy": 5.598113059997559,
      "epoch": 2.0906113537117905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004565528161691969,
      "loss": 5.0805,
      "mean_token_accuracy": 0.17337207198143006,
      "num_tokens": 46148926.0,
      "step": 21065
    },
    {
      "entropy": 5.6595946788787845,
      "epoch": 2.091107582373958,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004565318850243733,
      "loss": 5.0862,
      "mean_token_accuracy": 0.17846753895282746,
      "num_tokens": 46160130.0,
      "step": 21070
    },
    {
      "entropy": 5.624213743209839,
      "epoch": 2.0916038110361255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004565109493779502,
      "loss": 5.0022,
      "mean_token_accuracy": 0.18349802792072295,
      "num_tokens": 46169919.0,
      "step": 21075
    },
    {
      "entropy": 5.65905909538269,
      "epoch": 2.092100039698293,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045649000923044664,
      "loss": 5.0839,
      "mean_token_accuracy": 0.1762904092669487,
      "num_tokens": 46181186.0,
      "step": 21080
    },
    {
      "entropy": 5.609322595596313,
      "epoch": 2.0925962683604604,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045646906458238196,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17478445917367935,
      "num_tokens": 46192172.0,
      "step": 21085
    },
    {
      "entropy": 5.640009784698487,
      "epoch": 2.093092497022628,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004564481154342754,
      "loss": 5.0705,
      "mean_token_accuracy": 0.17124419808387756,
      "num_tokens": 46202678.0,
      "step": 21090
    },
    {
      "entropy": 5.653727245330811,
      "epoch": 2.0935887256847954,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004564271617866467,
      "loss": 5.0611,
      "mean_token_accuracy": 0.17964097559452058,
      "num_tokens": 46212213.0,
      "step": 21095
    },
    {
      "entropy": 5.707536935806274,
      "epoch": 2.0940849543469633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045640620364001524,
      "loss": 5.0909,
      "mean_token_accuracy": 0.17995578199625015,
      "num_tokens": 46223044.0,
      "step": 21100
    },
    {
      "entropy": 5.724713754653931,
      "epoch": 2.0945811830091308,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045638524099490086,
      "loss": 5.1022,
      "mean_token_accuracy": 0.17687397003173827,
      "num_tokens": 46236008.0,
      "step": 21105
    },
    {
      "entropy": 5.5972353458404545,
      "epoch": 2.0950774116712982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045636427385182334,
      "loss": 5.0527,
      "mean_token_accuracy": 0.17328155189752578,
      "num_tokens": 46247631.0,
      "step": 21110
    },
    {
      "entropy": 5.616739273071289,
      "epoch": 2.0955736403334657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045634330221130265,
      "loss": 5.015,
      "mean_token_accuracy": 0.1785146951675415,
      "num_tokens": 46258294.0,
      "step": 21115
    },
    {
      "entropy": 5.648260545730591,
      "epoch": 2.096069868995633,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004563223260738589,
      "loss": 5.023,
      "mean_token_accuracy": 0.1788094460964203,
      "num_tokens": 46269611.0,
      "step": 21120
    },
    {
      "entropy": 5.655237245559692,
      "epoch": 2.0965660976578007,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000456301345440012,
      "loss": 5.0916,
      "mean_token_accuracy": 0.18860187083482743,
      "num_tokens": 46280731.0,
      "step": 21125
    },
    {
      "entropy": 5.6173255443573,
      "epoch": 2.097062326319968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004562803603102825,
      "loss": 5.0402,
      "mean_token_accuracy": 0.175561386346817,
      "num_tokens": 46290226.0,
      "step": 21130
    },
    {
      "entropy": 5.65693416595459,
      "epoch": 2.0975585549821356,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045625937068519075,
      "loss": 5.0415,
      "mean_token_accuracy": 0.1783863425254822,
      "num_tokens": 46300388.0,
      "step": 21135
    },
    {
      "entropy": 5.628778600692749,
      "epoch": 2.098054783644303,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004562383765652571,
      "loss": 5.0759,
      "mean_token_accuracy": 0.1787893682718277,
      "num_tokens": 46312214.0,
      "step": 21140
    },
    {
      "entropy": 5.641170406341553,
      "epoch": 2.098551012306471,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004562173779510023,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1702629655599594,
      "num_tokens": 46322728.0,
      "step": 21145
    },
    {
      "entropy": 5.5976049423217775,
      "epoch": 2.0990472409686385,
      "grad_norm": 1.0,
      "learning_rate": 0.000456196374842947,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18539891690015792,
      "num_tokens": 46334169.0,
      "step": 21150
    },
    {
      "entropy": 5.657478904724121,
      "epoch": 2.099543469630806,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045617536724161203,
      "loss": 5.0014,
      "mean_token_accuracy": 0.17577320635318755,
      "num_tokens": 46344601.0,
      "step": 21155
    },
    {
      "entropy": 5.622035360336303,
      "epoch": 2.1000396982929734,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045615435514751834,
      "loss": 5.0861,
      "mean_token_accuracy": 0.17358327955007552,
      "num_tokens": 46355633.0,
      "step": 21160
    },
    {
      "entropy": 5.644324398040771,
      "epoch": 2.100535926955141,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000456133338561187,
      "loss": 5.1385,
      "mean_token_accuracy": 0.16421952098608017,
      "num_tokens": 46365981.0,
      "step": 21165
    },
    {
      "entropy": 5.642922019958496,
      "epoch": 2.1010321556173084,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004561123174831391,
      "loss": 5.099,
      "mean_token_accuracy": 0.1829488381743431,
      "num_tokens": 46376664.0,
      "step": 21170
    },
    {
      "entropy": 5.591289806365967,
      "epoch": 2.101528384279476,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004560912919138962,
      "loss": 5.0125,
      "mean_token_accuracy": 0.17959095239639283,
      "num_tokens": 46387918.0,
      "step": 21175
    },
    {
      "entropy": 5.6920445442199705,
      "epoch": 2.1020246129416433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045607026185397924,
      "loss": 5.0399,
      "mean_token_accuracy": 0.1810615211725235,
      "num_tokens": 46398660.0,
      "step": 21180
    },
    {
      "entropy": 5.643990850448608,
      "epoch": 2.102520841603811,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004560492273039101,
      "loss": 4.9866,
      "mean_token_accuracy": 0.18697368502616882,
      "num_tokens": 46409365.0,
      "step": 21185
    },
    {
      "entropy": 5.62970724105835,
      "epoch": 2.1030170702659787,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045602818826421024,
      "loss": 5.0065,
      "mean_token_accuracy": 0.1879824295639992,
      "num_tokens": 46419457.0,
      "step": 21190
    },
    {
      "entropy": 5.6158514499664305,
      "epoch": 2.103513298928146,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004560071447354013,
      "loss": 5.0711,
      "mean_token_accuracy": 0.1824338659644127,
      "num_tokens": 46430189.0,
      "step": 21195
    },
    {
      "entropy": 5.553476858139038,
      "epoch": 2.1040095275903137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004559860967180053,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18301851749420167,
      "num_tokens": 46440555.0,
      "step": 21200
    },
    {
      "entropy": 5.701135683059692,
      "epoch": 2.104505756252481,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004559650442125441,
      "loss": 5.0719,
      "mean_token_accuracy": 0.17495754063129426,
      "num_tokens": 46451582.0,
      "step": 21205
    },
    {
      "entropy": 5.73794264793396,
      "epoch": 2.1050019849146486,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004559439872195397,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17719461768865585,
      "num_tokens": 46463614.0,
      "step": 21210
    },
    {
      "entropy": 5.663989734649658,
      "epoch": 2.105498213576816,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004559229257395144,
      "loss": 5.0885,
      "mean_token_accuracy": 0.17566506564617157,
      "num_tokens": 46474279.0,
      "step": 21215
    },
    {
      "entropy": 5.67828221321106,
      "epoch": 2.1059944422389836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045590185977299025,
      "loss": 5.1815,
      "mean_token_accuracy": 0.17510544657707214,
      "num_tokens": 46485531.0,
      "step": 21220
    },
    {
      "entropy": 5.614159202575683,
      "epoch": 2.106490670901151,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004558807893204899,
      "loss": 4.9954,
      "mean_token_accuracy": 0.19085327237844468,
      "num_tokens": 46496197.0,
      "step": 21225
    },
    {
      "entropy": 5.686308670043945,
      "epoch": 2.106986899563319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004558597143825357,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17932274639606477,
      "num_tokens": 46506759.0,
      "step": 21230
    },
    {
      "entropy": 5.6138512134552006,
      "epoch": 2.1074831282254864,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004558386349596503,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17685170024633406,
      "num_tokens": 46518956.0,
      "step": 21235
    },
    {
      "entropy": 5.711998081207275,
      "epoch": 2.107979356887654,
      "grad_norm": 1.125,
      "learning_rate": 0.0004558175510523564,
      "loss": 5.1465,
      "mean_token_accuracy": 0.17624360024929048,
      "num_tokens": 46530109.0,
      "step": 21240
    },
    {
      "entropy": 5.666682434082031,
      "epoch": 2.1084755855498214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045579646266117696,
      "loss": 4.9889,
      "mean_token_accuracy": 0.1801063522696495,
      "num_tokens": 46542206.0,
      "step": 21245
    },
    {
      "entropy": 5.676545143127441,
      "epoch": 2.108971814211989,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045577536978663475,
      "loss": 5.1505,
      "mean_token_accuracy": 0.1794459417462349,
      "num_tokens": 46554143.0,
      "step": 21250
    },
    {
      "entropy": 5.572560596466064,
      "epoch": 2.1094680428741563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004557542724292529,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18196433335542678,
      "num_tokens": 46565697.0,
      "step": 21255
    },
    {
      "entropy": 5.667012453079224,
      "epoch": 2.109964271536324,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004557331705895545,
      "loss": 5.0509,
      "mean_token_accuracy": 0.1807334765791893,
      "num_tokens": 46576358.0,
      "step": 21260
    },
    {
      "entropy": 5.694110965728759,
      "epoch": 2.1104605001984913,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004557120642680631,
      "loss": 5.0859,
      "mean_token_accuracy": 0.17382164895534516,
      "num_tokens": 46589422.0,
      "step": 21265
    },
    {
      "entropy": 5.611817359924316,
      "epoch": 2.1109567288606588,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004556909534653018,
      "loss": 5.0617,
      "mean_token_accuracy": 0.18331800550222396,
      "num_tokens": 46599623.0,
      "step": 21270
    },
    {
      "entropy": 5.624138402938843,
      "epoch": 2.1114529575228267,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004556698381817942,
      "loss": 4.9967,
      "mean_token_accuracy": 0.18838846534490586,
      "num_tokens": 46609321.0,
      "step": 21275
    },
    {
      "entropy": 5.586575889587403,
      "epoch": 2.111949186184994,
      "grad_norm": 1.03125,
      "learning_rate": 0.000455648718418064,
      "loss": 5.0828,
      "mean_token_accuracy": 0.17780685871839524,
      "num_tokens": 46620265.0,
      "step": 21280
    },
    {
      "entropy": 5.663124084472656,
      "epoch": 2.1124454148471616,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004556275941746348,
      "loss": 5.0824,
      "mean_token_accuracy": 0.18283750414848327,
      "num_tokens": 46630776.0,
      "step": 21285
    },
    {
      "entropy": 5.5829521179199215,
      "epoch": 2.112941643509329,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045560646545203044,
      "loss": 4.9702,
      "mean_token_accuracy": 0.1822322353720665,
      "num_tokens": 46641382.0,
      "step": 21290
    },
    {
      "entropy": 5.54496169090271,
      "epoch": 2.1134378721714966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045558533225077496,
      "loss": 4.9816,
      "mean_token_accuracy": 0.18537165671586991,
      "num_tokens": 46653589.0,
      "step": 21295
    },
    {
      "entropy": 5.580789709091187,
      "epoch": 2.113934100833664,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045556419457139235,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18282255828380584,
      "num_tokens": 46664595.0,
      "step": 21300
    },
    {
      "entropy": 5.642827987670898,
      "epoch": 2.1144303294958315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045554305241440684,
      "loss": 5.0451,
      "mean_token_accuracy": 0.1761321559548378,
      "num_tokens": 46675453.0,
      "step": 21305
    },
    {
      "entropy": 5.583301782608032,
      "epoch": 2.114926558157999,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004555219057803426,
      "loss": 4.9587,
      "mean_token_accuracy": 0.18699311316013337,
      "num_tokens": 46685668.0,
      "step": 21310
    },
    {
      "entropy": 5.6245489597320555,
      "epoch": 2.115422786820167,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004555007546697242,
      "loss": 5.079,
      "mean_token_accuracy": 0.17724706679582597,
      "num_tokens": 46696144.0,
      "step": 21315
    },
    {
      "entropy": 5.646666240692139,
      "epoch": 2.1159190154823344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045547959908307585,
      "loss": 5.0498,
      "mean_token_accuracy": 0.17741041034460067,
      "num_tokens": 46707892.0,
      "step": 21320
    },
    {
      "entropy": 5.659394216537476,
      "epoch": 2.116415244144502,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004554584390209225,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17986789643764495,
      "num_tokens": 46721099.0,
      "step": 21325
    },
    {
      "entropy": 5.64666519165039,
      "epoch": 2.1169114728066694,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004554372744837886,
      "loss": 5.0937,
      "mean_token_accuracy": 0.17598705887794494,
      "num_tokens": 46732974.0,
      "step": 21330
    },
    {
      "entropy": 5.6331676006317135,
      "epoch": 2.117407701468837,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004554161054721992,
      "loss": 5.1201,
      "mean_token_accuracy": 0.1773753821849823,
      "num_tokens": 46743148.0,
      "step": 21335
    },
    {
      "entropy": 5.626925468444824,
      "epoch": 2.1179039301310043,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004553949319866791,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18524772375822068,
      "num_tokens": 46754214.0,
      "step": 21340
    },
    {
      "entropy": 5.624210691452026,
      "epoch": 2.118400158793172,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045537375402775334,
      "loss": 5.0452,
      "mean_token_accuracy": 0.179503957927227,
      "num_tokens": 46764112.0,
      "step": 21345
    },
    {
      "entropy": 5.606901836395264,
      "epoch": 2.1188963874553393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045535257159594724,
      "loss": 5.0081,
      "mean_token_accuracy": 0.1817215234041214,
      "num_tokens": 46775202.0,
      "step": 21350
    },
    {
      "entropy": 5.744702482223511,
      "epoch": 2.1193926161175067,
      "grad_norm": 1.125,
      "learning_rate": 0.000455331384691786,
      "loss": 5.0675,
      "mean_token_accuracy": 0.18083506375551223,
      "num_tokens": 46786557.0,
      "step": 21355
    },
    {
      "entropy": 5.603355550765992,
      "epoch": 2.1198888447796747,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004553101933157949,
      "loss": 5.0315,
      "mean_token_accuracy": 0.17318292558193207,
      "num_tokens": 46797082.0,
      "step": 21360
    },
    {
      "entropy": 5.582869958877564,
      "epoch": 2.120385073441842,
      "grad_norm": 1.125,
      "learning_rate": 0.00045528899746849965,
      "loss": 5.0722,
      "mean_token_accuracy": 0.17881860136985778,
      "num_tokens": 46806679.0,
      "step": 21365
    },
    {
      "entropy": 5.7002167224884035,
      "epoch": 2.1208813021040096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004552677971504257,
      "loss": 5.1887,
      "mean_token_accuracy": 0.16391037702560424,
      "num_tokens": 46818169.0,
      "step": 21370
    },
    {
      "entropy": 5.811241579055786,
      "epoch": 2.121377530766177,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004552465923620988,
      "loss": 5.2365,
      "mean_token_accuracy": 0.1704838529229164,
      "num_tokens": 46829552.0,
      "step": 21375
    },
    {
      "entropy": 5.714204502105713,
      "epoch": 2.1218737594283446,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004552253831040448,
      "loss": 5.1451,
      "mean_token_accuracy": 0.17868593633174895,
      "num_tokens": 46840808.0,
      "step": 21380
    },
    {
      "entropy": 5.698163270950317,
      "epoch": 2.122369988090512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045520416937678973,
      "loss": 5.0556,
      "mean_token_accuracy": 0.1732631042599678,
      "num_tokens": 46852066.0,
      "step": 21385
    },
    {
      "entropy": 5.603039360046386,
      "epoch": 2.1228662167526795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045518295118085955,
      "loss": 4.9971,
      "mean_token_accuracy": 0.18134532421827315,
      "num_tokens": 46863163.0,
      "step": 21390
    },
    {
      "entropy": 5.664174938201905,
      "epoch": 2.123362445414847,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045516172851678036,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17403849214315414,
      "num_tokens": 46873246.0,
      "step": 21395
    },
    {
      "entropy": 5.627801513671875,
      "epoch": 2.123858674077015,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045514050138507857,
      "loss": 5.0574,
      "mean_token_accuracy": 0.1781091198325157,
      "num_tokens": 46883881.0,
      "step": 21400
    },
    {
      "entropy": 5.668616676330567,
      "epoch": 2.1243549027391824,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004551192697862805,
      "loss": 5.0968,
      "mean_token_accuracy": 0.1752529188990593,
      "num_tokens": 46894907.0,
      "step": 21405
    },
    {
      "entropy": 5.612592983245849,
      "epoch": 2.12485113140135,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004550980337209127,
      "loss": 5.0151,
      "mean_token_accuracy": 0.17152939587831498,
      "num_tokens": 46905220.0,
      "step": 21410
    },
    {
      "entropy": 5.603359127044678,
      "epoch": 2.1253473600635173,
      "grad_norm": 1.359375,
      "learning_rate": 0.00045507679318950166,
      "loss": 5.0133,
      "mean_token_accuracy": 0.186904938519001,
      "num_tokens": 46916328.0,
      "step": 21415
    },
    {
      "entropy": 5.832666730880737,
      "epoch": 2.125843588725685,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004550555481925743,
      "loss": 5.2201,
      "mean_token_accuracy": 0.16313505321741104,
      "num_tokens": 46927450.0,
      "step": 21420
    },
    {
      "entropy": 5.581094360351562,
      "epoch": 2.1263398173878523,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045503429873065723,
      "loss": 4.977,
      "mean_token_accuracy": 0.18246981501579285,
      "num_tokens": 46936758.0,
      "step": 21425
    },
    {
      "entropy": 5.525733995437622,
      "epoch": 2.1268360460500197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004550130448042775,
      "loss": 4.9296,
      "mean_token_accuracy": 0.1909484848380089,
      "num_tokens": 46945776.0,
      "step": 21430
    },
    {
      "entropy": 5.599197721481323,
      "epoch": 2.127332274712187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549917864139622,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18801047056913375,
      "num_tokens": 46955621.0,
      "step": 21435
    },
    {
      "entropy": 5.7298845767974855,
      "epoch": 2.1278285033743547,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004549705235602384,
      "loss": 5.1499,
      "mean_token_accuracy": 0.1790675237774849,
      "num_tokens": 46966890.0,
      "step": 21440
    },
    {
      "entropy": 5.684625053405762,
      "epoch": 2.1283247320365226,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549492562436334,
      "loss": 5.1498,
      "mean_token_accuracy": 0.1764849007129669,
      "num_tokens": 46978380.0,
      "step": 21445
    },
    {
      "entropy": 5.636482763290405,
      "epoch": 2.12882096069869,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045492798446467465,
      "loss": 4.976,
      "mean_token_accuracy": 0.18499033749103547,
      "num_tokens": 46988286.0,
      "step": 21450
    },
    {
      "entropy": 5.595537185668945,
      "epoch": 2.1293171893608576,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045490670822388957,
      "loss": 4.981,
      "mean_token_accuracy": 0.1825837731361389,
      "num_tokens": 46997795.0,
      "step": 21455
    },
    {
      "entropy": 5.546331548690796,
      "epoch": 2.129813418023025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004548854275218059,
      "loss": 4.8849,
      "mean_token_accuracy": 0.19140267372131348,
      "num_tokens": 47007970.0,
      "step": 21460
    },
    {
      "entropy": 5.594887399673462,
      "epoch": 2.1303096466851925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045486414235895115,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18021750301122666,
      "num_tokens": 47019548.0,
      "step": 21465
    },
    {
      "entropy": 5.628417015075684,
      "epoch": 2.13080587534736,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004548428527358532,
      "loss": 5.1169,
      "mean_token_accuracy": 0.1768563747406006,
      "num_tokens": 47029568.0,
      "step": 21470
    },
    {
      "entropy": 5.68505825996399,
      "epoch": 2.1313021040095275,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004548215586530401,
      "loss": 5.1306,
      "mean_token_accuracy": 0.1716662585735321,
      "num_tokens": 47042048.0,
      "step": 21475
    },
    {
      "entropy": 5.643602085113526,
      "epoch": 2.131798332671695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045480026011103975,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18001498579978942,
      "num_tokens": 47052099.0,
      "step": 21480
    },
    {
      "entropy": 5.652211380004883,
      "epoch": 2.132294561333863,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045477895711038045,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18340561389923096,
      "num_tokens": 47063310.0,
      "step": 21485
    },
    {
      "entropy": 5.624781227111816,
      "epoch": 2.1327907899960303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004547576496515904,
      "loss": 5.038,
      "mean_token_accuracy": 0.1745828777551651,
      "num_tokens": 47073545.0,
      "step": 21490
    },
    {
      "entropy": 5.607168769836425,
      "epoch": 2.133287018658198,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045473633773519795,
      "loss": 5.0848,
      "mean_token_accuracy": 0.18083644956350325,
      "num_tokens": 47085456.0,
      "step": 21495
    },
    {
      "entropy": 5.6698657989501955,
      "epoch": 2.1337832473203653,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045471502136173173,
      "loss": 5.0692,
      "mean_token_accuracy": 0.18310429900884628,
      "num_tokens": 47096590.0,
      "step": 21500
    },
    {
      "entropy": 5.694825935363769,
      "epoch": 2.1342794759825328,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546937005317201,
      "loss": 5.0108,
      "mean_token_accuracy": 0.1897694945335388,
      "num_tokens": 47107876.0,
      "step": 21505
    },
    {
      "entropy": 5.633671474456787,
      "epoch": 2.1347757046447002,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045467237524569193,
      "loss": 5.0449,
      "mean_token_accuracy": 0.1829899713397026,
      "num_tokens": 47117662.0,
      "step": 21510
    },
    {
      "entropy": 5.596507740020752,
      "epoch": 2.1352719333068677,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045465104550417604,
      "loss": 5.0817,
      "mean_token_accuracy": 0.17588527500629425,
      "num_tokens": 47129283.0,
      "step": 21515
    },
    {
      "entropy": 5.701072645187378,
      "epoch": 2.135768161969035,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045462971130770123,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1790285661816597,
      "num_tokens": 47140824.0,
      "step": 21520
    },
    {
      "entropy": 5.625451278686524,
      "epoch": 2.1362643906312027,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546083726567968,
      "loss": 5.0006,
      "mean_token_accuracy": 0.18578442931175232,
      "num_tokens": 47151733.0,
      "step": 21525
    },
    {
      "entropy": 5.642958354949951,
      "epoch": 2.1367606192933706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004545870295519917,
      "loss": 5.1334,
      "mean_token_accuracy": 0.16901808530092238,
      "num_tokens": 47162846.0,
      "step": 21530
    },
    {
      "entropy": 5.685342121124267,
      "epoch": 2.137256847955538,
      "grad_norm": 1.125,
      "learning_rate": 0.00045456568199381524,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18457923084497452,
      "num_tokens": 47174969.0,
      "step": 21535
    },
    {
      "entropy": 5.730123519897461,
      "epoch": 2.1377530766177055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045454432998279676,
      "loss": 5.1108,
      "mean_token_accuracy": 0.17629973441362382,
      "num_tokens": 47186903.0,
      "step": 21540
    },
    {
      "entropy": 5.641247320175171,
      "epoch": 2.138249305279873,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004545229735194658,
      "loss": 5.0719,
      "mean_token_accuracy": 0.18022799342870713,
      "num_tokens": 47197716.0,
      "step": 21545
    },
    {
      "entropy": 5.613566637039185,
      "epoch": 2.1387455339420405,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045450161260435196,
      "loss": 5.0149,
      "mean_token_accuracy": 0.18210412710905075,
      "num_tokens": 47208623.0,
      "step": 21550
    },
    {
      "entropy": 5.588279151916504,
      "epoch": 2.139241762604208,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004544802472379849,
      "loss": 5.0405,
      "mean_token_accuracy": 0.1775360882282257,
      "num_tokens": 47218248.0,
      "step": 21555
    },
    {
      "entropy": 5.643615341186523,
      "epoch": 2.1397379912663754,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004544588774208945,
      "loss": 5.1452,
      "mean_token_accuracy": 0.1756714940071106,
      "num_tokens": 47232028.0,
      "step": 21560
    },
    {
      "entropy": 5.6966187953948975,
      "epoch": 2.140234219928543,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045443750315361055,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18597073704004288,
      "num_tokens": 47241976.0,
      "step": 21565
    },
    {
      "entropy": 5.654049444198608,
      "epoch": 2.140730448590711,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004544161244366632,
      "loss": 5.0243,
      "mean_token_accuracy": 0.1824648693203926,
      "num_tokens": 47253228.0,
      "step": 21570
    },
    {
      "entropy": 5.616700506210327,
      "epoch": 2.1412266772528783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004543947412705826,
      "loss": 5.0392,
      "mean_token_accuracy": 0.17600257098674774,
      "num_tokens": 47265210.0,
      "step": 21575
    },
    {
      "entropy": 5.586776208877564,
      "epoch": 2.1417229059150458,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045437335365589905,
      "loss": 5.0447,
      "mean_token_accuracy": 0.17824703752994536,
      "num_tokens": 47275723.0,
      "step": 21580
    },
    {
      "entropy": 5.669466972351074,
      "epoch": 2.1422191345772132,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045435196159314276,
      "loss": 5.0978,
      "mean_token_accuracy": 0.17419553846120833,
      "num_tokens": 47286469.0,
      "step": 21585
    },
    {
      "entropy": 5.724857759475708,
      "epoch": 2.1427153632393807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045433056508284425,
      "loss": 5.1208,
      "mean_token_accuracy": 0.17654566168785096,
      "num_tokens": 47296461.0,
      "step": 21590
    },
    {
      "entropy": 5.655224418640136,
      "epoch": 2.143211591901548,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004543091641255342,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18420733362436295,
      "num_tokens": 47307217.0,
      "step": 21595
    },
    {
      "entropy": 5.6680100440979,
      "epoch": 2.1437078205637157,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004542877587217432,
      "loss": 5.139,
      "mean_token_accuracy": 0.17019178122282028,
      "num_tokens": 47317041.0,
      "step": 21600
    },
    {
      "entropy": 5.6503331661224365,
      "epoch": 2.144204049225883,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045426634887200224,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1806392565369606,
      "num_tokens": 47328185.0,
      "step": 21605
    },
    {
      "entropy": 5.658969497680664,
      "epoch": 2.1447002778880506,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000454244934576842,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17453552335500716,
      "num_tokens": 47339984.0,
      "step": 21610
    },
    {
      "entropy": 5.660405492782592,
      "epoch": 2.1451965065502185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045422351583679365,
      "loss": 5.0504,
      "mean_token_accuracy": 0.18408181965351106,
      "num_tokens": 47350911.0,
      "step": 21615
    },
    {
      "entropy": 5.683227968215943,
      "epoch": 2.145692735212386,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004542020926523883,
      "loss": 5.0484,
      "mean_token_accuracy": 0.17690181434154512,
      "num_tokens": 47361703.0,
      "step": 21620
    },
    {
      "entropy": 5.6585112571716305,
      "epoch": 2.1461889638745535,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004541806650241572,
      "loss": 5.0198,
      "mean_token_accuracy": 0.1761561378836632,
      "num_tokens": 47372878.0,
      "step": 21625
    },
    {
      "entropy": 5.616320276260376,
      "epoch": 2.146685192536721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004541592329526317,
      "loss": 5.0494,
      "mean_token_accuracy": 0.17447454631328582,
      "num_tokens": 47383056.0,
      "step": 21630
    },
    {
      "entropy": 5.602045774459839,
      "epoch": 2.1471814211988884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004541377964383434,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18270945996046067,
      "num_tokens": 47394160.0,
      "step": 21635
    },
    {
      "entropy": 5.668890810012817,
      "epoch": 2.147677649861056,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004541163554818237,
      "loss": 5.0198,
      "mean_token_accuracy": 0.17925110161304475,
      "num_tokens": 47404861.0,
      "step": 21640
    },
    {
      "entropy": 5.648185110092163,
      "epoch": 2.1481738785232234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045409491008360423,
      "loss": 5.1026,
      "mean_token_accuracy": 0.17934130132198334,
      "num_tokens": 47415236.0,
      "step": 21645
    },
    {
      "entropy": 5.650694084167481,
      "epoch": 2.148670107185391,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004540734602442169,
      "loss": 5.0621,
      "mean_token_accuracy": 0.17942364662885665,
      "num_tokens": 47425789.0,
      "step": 21650
    },
    {
      "entropy": 5.641299819946289,
      "epoch": 2.149166335847559,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004540520059641937,
      "loss": 4.9353,
      "mean_token_accuracy": 0.1860051304101944,
      "num_tokens": 47436236.0,
      "step": 21655
    },
    {
      "entropy": 5.569094514846801,
      "epoch": 2.1496625645097263,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004540305472440666,
      "loss": 4.9537,
      "mean_token_accuracy": 0.1860675558447838,
      "num_tokens": 47447041.0,
      "step": 21660
    },
    {
      "entropy": 5.622613048553466,
      "epoch": 2.1501587931718937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004540090840843677,
      "loss": 5.1029,
      "mean_token_accuracy": 0.1701733112335205,
      "num_tokens": 47458001.0,
      "step": 21665
    },
    {
      "entropy": 5.622630405426025,
      "epoch": 2.150655021834061,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004539876164856292,
      "loss": 4.9721,
      "mean_token_accuracy": 0.18971407413482666,
      "num_tokens": 47468621.0,
      "step": 21670
    },
    {
      "entropy": 5.608687877655029,
      "epoch": 2.1511512504962287,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004539661444483835,
      "loss": 5.085,
      "mean_token_accuracy": 0.18315064311027526,
      "num_tokens": 47478957.0,
      "step": 21675
    },
    {
      "entropy": 5.5822583675384525,
      "epoch": 2.151647479158396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004539446679731631,
      "loss": 5.0297,
      "mean_token_accuracy": 0.17962195128202438,
      "num_tokens": 47489640.0,
      "step": 21680
    },
    {
      "entropy": 5.687397480010986,
      "epoch": 2.1521437078205636,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004539231870605005,
      "loss": 5.1633,
      "mean_token_accuracy": 0.1720585271716118,
      "num_tokens": 47501581.0,
      "step": 21685
    },
    {
      "entropy": 5.667452049255371,
      "epoch": 2.152639936482731,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045390170171092845,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17527162730693818,
      "num_tokens": 47513605.0,
      "step": 21690
    },
    {
      "entropy": 5.699994277954102,
      "epoch": 2.1531361651448986,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045388021192497956,
      "loss": 5.0863,
      "mean_token_accuracy": 0.18231877088546752,
      "num_tokens": 47523784.0,
      "step": 21695
    },
    {
      "entropy": 5.576864242553711,
      "epoch": 2.1536323938070665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000453858717703187,
      "loss": 4.9872,
      "mean_token_accuracy": 0.17686835378408433,
      "num_tokens": 47534643.0,
      "step": 21700
    },
    {
      "entropy": 5.6110903263092045,
      "epoch": 2.154128622469234,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004538372190460836,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17194697260856628,
      "num_tokens": 47545920.0,
      "step": 21705
    },
    {
      "entropy": 5.616489505767822,
      "epoch": 2.1546248511314015,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004538157159542025,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17756803333759308,
      "num_tokens": 47557024.0,
      "step": 21710
    },
    {
      "entropy": 5.646591091156006,
      "epoch": 2.155121079793569,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004537942084280771,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18511921167373657,
      "num_tokens": 47567531.0,
      "step": 21715
    },
    {
      "entropy": 5.644152402877808,
      "epoch": 2.1556173084557364,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045377269646824036,
      "loss": 5.1394,
      "mean_token_accuracy": 0.17711188346147538,
      "num_tokens": 47578642.0,
      "step": 21720
    },
    {
      "entropy": 5.648008728027344,
      "epoch": 2.156113537117904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045375118007522607,
      "loss": 5.1771,
      "mean_token_accuracy": 0.1766583427786827,
      "num_tokens": 47589346.0,
      "step": 21725
    },
    {
      "entropy": 5.584103155136108,
      "epoch": 2.1566097657800714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004537296592495678,
      "loss": 4.9949,
      "mean_token_accuracy": 0.18299344033002854,
      "num_tokens": 47599689.0,
      "step": 21730
    },
    {
      "entropy": 5.7107765674591064,
      "epoch": 2.157105994442239,
      "grad_norm": 1.125,
      "learning_rate": 0.000453708133991799,
      "loss": 5.0801,
      "mean_token_accuracy": 0.1766654744744301,
      "num_tokens": 47609214.0,
      "step": 21735
    },
    {
      "entropy": 5.670279455184937,
      "epoch": 2.1576022231044063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045368660430245364,
      "loss": 5.0314,
      "mean_token_accuracy": 0.1798383742570877,
      "num_tokens": 47620119.0,
      "step": 21740
    },
    {
      "entropy": 5.625681781768799,
      "epoch": 2.158098451766574,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004536650701820654,
      "loss": 5.0063,
      "mean_token_accuracy": 0.18083962500095369,
      "num_tokens": 47630085.0,
      "step": 21745
    },
    {
      "entropy": 5.635739994049072,
      "epoch": 2.1585946804287417,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004536435316311685,
      "loss": 4.9974,
      "mean_token_accuracy": 0.1827731430530548,
      "num_tokens": 47641457.0,
      "step": 21750
    },
    {
      "entropy": 5.638498973846436,
      "epoch": 2.159090909090909,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045362198865029696,
      "loss": 5.0217,
      "mean_token_accuracy": 0.17574773579835892,
      "num_tokens": 47652269.0,
      "step": 21755
    },
    {
      "entropy": 5.603327417373658,
      "epoch": 2.1595871377530766,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045360044123998495,
      "loss": 5.0432,
      "mean_token_accuracy": 0.18111125230789185,
      "num_tokens": 47664601.0,
      "step": 21760
    },
    {
      "entropy": 5.66009750366211,
      "epoch": 2.160083366415244,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045357888940076685,
      "loss": 5.1582,
      "mean_token_accuracy": 0.17769457399845123,
      "num_tokens": 47676187.0,
      "step": 21765
    },
    {
      "entropy": 5.571529769897461,
      "epoch": 2.1605795950774116,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004535573331331771,
      "loss": 4.9632,
      "mean_token_accuracy": 0.1836626872420311,
      "num_tokens": 47687057.0,
      "step": 21770
    },
    {
      "entropy": 5.614437007904053,
      "epoch": 2.161075823739579,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00045353577243775035,
      "loss": 5.014,
      "mean_token_accuracy": 0.18068371266126632,
      "num_tokens": 47699074.0,
      "step": 21775
    },
    {
      "entropy": 5.6305251121521,
      "epoch": 2.1615720524017465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004535142073150211,
      "loss": 4.9991,
      "mean_token_accuracy": 0.18320581167936326,
      "num_tokens": 47710400.0,
      "step": 21780
    },
    {
      "entropy": 5.569340801239013,
      "epoch": 2.162068281063914,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004534926377655241,
      "loss": 4.8985,
      "mean_token_accuracy": 0.1975107192993164,
      "num_tokens": 47720900.0,
      "step": 21785
    },
    {
      "entropy": 5.594466257095337,
      "epoch": 2.162564509726082,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045347106378979435,
      "loss": 4.9214,
      "mean_token_accuracy": 0.18542141169309617,
      "num_tokens": 47730702.0,
      "step": 21790
    },
    {
      "entropy": 5.609736537933349,
      "epoch": 2.1630607383882494,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045344948538836674,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18279297053813934,
      "num_tokens": 47740451.0,
      "step": 21795
    },
    {
      "entropy": 5.61353816986084,
      "epoch": 2.163556967050417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045342790256177646,
      "loss": 5.0683,
      "mean_token_accuracy": 0.1810429573059082,
      "num_tokens": 47752285.0,
      "step": 21800
    },
    {
      "entropy": 5.6516711711883545,
      "epoch": 2.1640531957125844,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004534063153105587,
      "loss": 5.0704,
      "mean_token_accuracy": 0.1787519097328186,
      "num_tokens": 47761895.0,
      "step": 21805
    },
    {
      "entropy": 5.638353872299194,
      "epoch": 2.164549424374752,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004533847236352487,
      "loss": 4.9575,
      "mean_token_accuracy": 0.1917251780629158,
      "num_tokens": 47771718.0,
      "step": 21810
    },
    {
      "entropy": 5.577188158035279,
      "epoch": 2.1650456530369193,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000453363127536382,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18011937588453292,
      "num_tokens": 47782183.0,
      "step": 21815
    },
    {
      "entropy": 5.564975786209106,
      "epoch": 2.165541881699087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045334152701449407,
      "loss": 4.9546,
      "mean_token_accuracy": 0.18786201477050782,
      "num_tokens": 47792190.0,
      "step": 21820
    },
    {
      "entropy": 5.727806138992309,
      "epoch": 2.1660381103612543,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004533199220701205,
      "loss": 5.1672,
      "mean_token_accuracy": 0.17330887764692307,
      "num_tokens": 47803245.0,
      "step": 21825
    },
    {
      "entropy": 5.646506595611572,
      "epoch": 2.166534339023422,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045329831270379714,
      "loss": 5.0996,
      "mean_token_accuracy": 0.17620295733213426,
      "num_tokens": 47814605.0,
      "step": 21830
    },
    {
      "entropy": 5.596370506286621,
      "epoch": 2.1670305676855897,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045327669891605975,
      "loss": 5.0685,
      "mean_token_accuracy": 0.1810953512787819,
      "num_tokens": 47825945.0,
      "step": 21835
    },
    {
      "entropy": 5.627554512023925,
      "epoch": 2.167526796347757,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004532550807074445,
      "loss": 5.0652,
      "mean_token_accuracy": 0.17536602467298507,
      "num_tokens": 47838119.0,
      "step": 21840
    },
    {
      "entropy": 5.691180276870727,
      "epoch": 2.1680230250099246,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004532334580784873,
      "loss": 5.0584,
      "mean_token_accuracy": 0.1787424385547638,
      "num_tokens": 47849560.0,
      "step": 21845
    },
    {
      "entropy": 5.716678762435913,
      "epoch": 2.168519253672092,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004532118310297244,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1713024616241455,
      "num_tokens": 47860036.0,
      "step": 21850
    },
    {
      "entropy": 5.563648796081543,
      "epoch": 2.1690154823342596,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004531901995616921,
      "loss": 4.963,
      "mean_token_accuracy": 0.18195141553878785,
      "num_tokens": 47870826.0,
      "step": 21855
    },
    {
      "entropy": 5.599672317504883,
      "epoch": 2.169511710996427,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004531685636749268,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1726074367761612,
      "num_tokens": 47881415.0,
      "step": 21860
    },
    {
      "entropy": 5.70466160774231,
      "epoch": 2.1700079396585945,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004531469233699651,
      "loss": 5.0855,
      "mean_token_accuracy": 0.1778191551566124,
      "num_tokens": 47893343.0,
      "step": 21865
    },
    {
      "entropy": 5.639388179779052,
      "epoch": 2.170504168320762,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004531252786473436,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17804837971925735,
      "num_tokens": 47904871.0,
      "step": 21870
    },
    {
      "entropy": 5.618955326080322,
      "epoch": 2.17100039698293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004531036295075989,
      "loss": 4.9936,
      "mean_token_accuracy": 0.18124319463968278,
      "num_tokens": 47915367.0,
      "step": 21875
    },
    {
      "entropy": 5.615488529205322,
      "epoch": 2.1714966256450974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045308197595126797,
      "loss": 4.9952,
      "mean_token_accuracy": 0.1786472260951996,
      "num_tokens": 47925987.0,
      "step": 21880
    },
    {
      "entropy": 5.644804525375366,
      "epoch": 2.171992854307265,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004530603179788879,
      "loss": 5.0959,
      "mean_token_accuracy": 0.1770787015557289,
      "num_tokens": 47937688.0,
      "step": 21885
    },
    {
      "entropy": 5.627934837341309,
      "epoch": 2.1724890829694323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004530386555909955,
      "loss": 5.053,
      "mean_token_accuracy": 0.18215274512767793,
      "num_tokens": 47948550.0,
      "step": 21890
    },
    {
      "entropy": 5.632548713684082,
      "epoch": 2.1729853116316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045301698878812813,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18593871742486953,
      "num_tokens": 47957982.0,
      "step": 21895
    },
    {
      "entropy": 5.683117151260376,
      "epoch": 2.1734815402937673,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004529953175708231,
      "loss": 5.0863,
      "mean_token_accuracy": 0.17729239016771317,
      "num_tokens": 47969562.0,
      "step": 21900
    },
    {
      "entropy": 5.661708641052246,
      "epoch": 2.1739777689559348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045297364193961763,
      "loss": 5.0958,
      "mean_token_accuracy": 0.1730342075228691,
      "num_tokens": 47981323.0,
      "step": 21905
    },
    {
      "entropy": 5.626578283309937,
      "epoch": 2.1744739976181022,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045295196189504934,
      "loss": 5.0231,
      "mean_token_accuracy": 0.17872336953878404,
      "num_tokens": 47991454.0,
      "step": 21910
    },
    {
      "entropy": 5.5936431884765625,
      "epoch": 2.17497022628027,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045293027743765595,
      "loss": 4.9241,
      "mean_token_accuracy": 0.18465662449598313,
      "num_tokens": 48001300.0,
      "step": 21915
    },
    {
      "entropy": 5.634315490722656,
      "epoch": 2.1754664549424376,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045290858856797503,
      "loss": 5.023,
      "mean_token_accuracy": 0.18152159750461577,
      "num_tokens": 48012608.0,
      "step": 21920
    },
    {
      "entropy": 5.640278291702271,
      "epoch": 2.175962683604605,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045288689528654444,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18633034974336624,
      "num_tokens": 48022120.0,
      "step": 21925
    },
    {
      "entropy": 5.633399772644043,
      "epoch": 2.1764589122667726,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045286519759390217,
      "loss": 5.0451,
      "mean_token_accuracy": 0.17726479172706605,
      "num_tokens": 48032629.0,
      "step": 21930
    },
    {
      "entropy": 5.637757349014282,
      "epoch": 2.17695514092894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004528434954905863,
      "loss": 4.9861,
      "mean_token_accuracy": 0.18278150111436844,
      "num_tokens": 48042934.0,
      "step": 21935
    },
    {
      "entropy": 5.65530686378479,
      "epoch": 2.1774513695911075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004528217889771349,
      "loss": 5.0459,
      "mean_token_accuracy": 0.18289434164762497,
      "num_tokens": 48052825.0,
      "step": 21940
    },
    {
      "entropy": 5.663210010528564,
      "epoch": 2.177947598253275,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045280007805408636,
      "loss": 5.1191,
      "mean_token_accuracy": 0.16940819770097731,
      "num_tokens": 48064565.0,
      "step": 21945
    },
    {
      "entropy": 5.669862461090088,
      "epoch": 2.1784438269154425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045277836272197895,
      "loss": 5.0917,
      "mean_token_accuracy": 0.18151675164699554,
      "num_tokens": 48076006.0,
      "step": 21950
    },
    {
      "entropy": 5.751352167129516,
      "epoch": 2.17894005557761,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045275664298135134,
      "loss": 5.1898,
      "mean_token_accuracy": 0.16942874491214752,
      "num_tokens": 48087110.0,
      "step": 21955
    },
    {
      "entropy": 5.649253034591675,
      "epoch": 2.179436284239778,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004527349188327419,
      "loss": 5.0774,
      "mean_token_accuracy": 0.1832103595137596,
      "num_tokens": 48098894.0,
      "step": 21960
    },
    {
      "entropy": 5.661367273330688,
      "epoch": 2.1799325129019453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004527131902766895,
      "loss": 5.0814,
      "mean_token_accuracy": 0.17663509249687195,
      "num_tokens": 48109729.0,
      "step": 21965
    },
    {
      "entropy": 5.678110837936401,
      "epoch": 2.180428741564113,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004526914573137329,
      "loss": 5.0795,
      "mean_token_accuracy": 0.1725434109568596,
      "num_tokens": 48120345.0,
      "step": 21970
    },
    {
      "entropy": 5.634387350082397,
      "epoch": 2.1809249702262803,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045266971994441116,
      "loss": 5.0097,
      "mean_token_accuracy": 0.17831757217645644,
      "num_tokens": 48130185.0,
      "step": 21975
    },
    {
      "entropy": 5.58721284866333,
      "epoch": 2.1814211988884478,
      "grad_norm": 1.015625,
      "learning_rate": 0.000452647978169263,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18102657794952393,
      "num_tokens": 48141020.0,
      "step": 21980
    },
    {
      "entropy": 5.628988313674927,
      "epoch": 2.1819174275506152,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045262623198882797,
      "loss": 5.0629,
      "mean_token_accuracy": 0.18416932821273804,
      "num_tokens": 48153361.0,
      "step": 21985
    },
    {
      "entropy": 5.623304986953736,
      "epoch": 2.1824136562127827,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004526044814036451,
      "loss": 5.0348,
      "mean_token_accuracy": 0.18192316144704818,
      "num_tokens": 48163604.0,
      "step": 21990
    },
    {
      "entropy": 5.629463815689087,
      "epoch": 2.18290988487495,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045258272641425377,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1770092576742172,
      "num_tokens": 48174732.0,
      "step": 21995
    },
    {
      "entropy": 5.660993432998657,
      "epoch": 2.183406113537118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004525609670211934,
      "loss": 5.006,
      "mean_token_accuracy": 0.18245165497064592,
      "num_tokens": 48187082.0,
      "step": 22000
    },
    {
      "entropy": 5.627263832092285,
      "epoch": 2.1839023421992856,
      "grad_norm": 1.125,
      "learning_rate": 0.00045253920322500373,
      "loss": 5.0285,
      "mean_token_accuracy": 0.18443397730588912,
      "num_tokens": 48196547.0,
      "step": 22005
    },
    {
      "entropy": 5.610086584091187,
      "epoch": 2.184398570861453,
      "grad_norm": 1.0,
      "learning_rate": 0.00045251743502622445,
      "loss": 5.0767,
      "mean_token_accuracy": 0.179419706761837,
      "num_tokens": 48207404.0,
      "step": 22010
    },
    {
      "entropy": 5.56248893737793,
      "epoch": 2.1848947995236205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004524956624253952,
      "loss": 5.0477,
      "mean_token_accuracy": 0.18717203736305238,
      "num_tokens": 48217093.0,
      "step": 22015
    },
    {
      "entropy": 5.655049610137939,
      "epoch": 2.185391028185788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045247388542305595,
      "loss": 5.0856,
      "mean_token_accuracy": 0.1790393888950348,
      "num_tokens": 48228093.0,
      "step": 22020
    },
    {
      "entropy": 5.621839761734009,
      "epoch": 2.1858872568479555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004524521040197469,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18589906692504882,
      "num_tokens": 48238834.0,
      "step": 22025
    },
    {
      "entropy": 5.697005033493042,
      "epoch": 2.186383485510123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045243031821600794,
      "loss": 5.0834,
      "mean_token_accuracy": 0.18443846702575684,
      "num_tokens": 48250782.0,
      "step": 22030
    },
    {
      "entropy": 5.538712120056152,
      "epoch": 2.1868797141722904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004524085280123795,
      "loss": 4.9697,
      "mean_token_accuracy": 0.17766082733869554,
      "num_tokens": 48261711.0,
      "step": 22035
    },
    {
      "entropy": 5.606232833862305,
      "epoch": 2.187375942834458,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004523867334094018,
      "loss": 5.0843,
      "mean_token_accuracy": 0.17648404687643052,
      "num_tokens": 48273505.0,
      "step": 22040
    },
    {
      "entropy": 5.572363090515137,
      "epoch": 2.187872171496626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004523649344076154,
      "loss": 4.993,
      "mean_token_accuracy": 0.18259650319814683,
      "num_tokens": 48284639.0,
      "step": 22045
    },
    {
      "entropy": 5.638118267059326,
      "epoch": 2.1883684001587933,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045234313100756077,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17662874162197112,
      "num_tokens": 48296275.0,
      "step": 22050
    },
    {
      "entropy": 5.628632640838623,
      "epoch": 2.188864628820961,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045232132320977854,
      "loss": 4.984,
      "mean_token_accuracy": 0.18417198657989503,
      "num_tokens": 48306859.0,
      "step": 22055
    },
    {
      "entropy": 5.664434623718262,
      "epoch": 2.1893608574831283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004522995110148097,
      "loss": 5.1136,
      "mean_token_accuracy": 0.17838209867477417,
      "num_tokens": 48317561.0,
      "step": 22060
    },
    {
      "entropy": 5.636991930007935,
      "epoch": 2.1898570861452957,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000452277694423195,
      "loss": 5.1282,
      "mean_token_accuracy": 0.17036063522100447,
      "num_tokens": 48327814.0,
      "step": 22065
    },
    {
      "entropy": 5.632955598831177,
      "epoch": 2.190353314807463,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004522558734354755,
      "loss": 4.9632,
      "mean_token_accuracy": 0.18936610221862793,
      "num_tokens": 48339658.0,
      "step": 22070
    },
    {
      "entropy": 5.618787956237793,
      "epoch": 2.1908495434696307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045223404805219225,
      "loss": 5.0806,
      "mean_token_accuracy": 0.18029789477586747,
      "num_tokens": 48352115.0,
      "step": 22075
    },
    {
      "entropy": 5.66240668296814,
      "epoch": 2.191345772131798,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045221221827388655,
      "loss": 5.0383,
      "mean_token_accuracy": 0.18301020860671996,
      "num_tokens": 48362378.0,
      "step": 22080
    },
    {
      "entropy": 5.627870559692383,
      "epoch": 2.191842000793966,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004521903841010997,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18828303068876268,
      "num_tokens": 48373035.0,
      "step": 22085
    },
    {
      "entropy": 5.573232507705688,
      "epoch": 2.1923382294561335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521685455343731,
      "loss": 5.0423,
      "mean_token_accuracy": 0.18477026820182801,
      "num_tokens": 48383737.0,
      "step": 22090
    },
    {
      "entropy": 5.61153473854065,
      "epoch": 2.192834458118301,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045214670257424835,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1813200443983078,
      "num_tokens": 48394547.0,
      "step": 22095
    },
    {
      "entropy": 5.672780227661133,
      "epoch": 2.1933306867804685,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004521248552212671,
      "loss": 5.0857,
      "mean_token_accuracy": 0.18232647478580474,
      "num_tokens": 48404906.0,
      "step": 22100
    },
    {
      "entropy": 5.612067413330078,
      "epoch": 2.193826915442636,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004521030034759711,
      "loss": 5.0303,
      "mean_token_accuracy": 0.18416082710027695,
      "num_tokens": 48416425.0,
      "step": 22105
    },
    {
      "entropy": 5.609609985351563,
      "epoch": 2.1943231441048034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004520811473389023,
      "loss": 5.0008,
      "mean_token_accuracy": 0.17902086526155472,
      "num_tokens": 48426967.0,
      "step": 22110
    },
    {
      "entropy": 5.7286333560943605,
      "epoch": 2.194819372766971,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045205928681060255,
      "loss": 5.1247,
      "mean_token_accuracy": 0.18070243746042253,
      "num_tokens": 48439720.0,
      "step": 22115
    },
    {
      "entropy": 5.687529373168945,
      "epoch": 2.1953156014291384,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000452037421891614,
      "loss": 5.0814,
      "mean_token_accuracy": 0.182662533223629,
      "num_tokens": 48449949.0,
      "step": 22120
    },
    {
      "entropy": 5.588335943222046,
      "epoch": 2.195811830091306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045201555258247896,
      "loss": 4.9804,
      "mean_token_accuracy": 0.18498859703540801,
      "num_tokens": 48462515.0,
      "step": 22125
    },
    {
      "entropy": 5.58319058418274,
      "epoch": 2.196308058753474,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004519936788837396,
      "loss": 5.0132,
      "mean_token_accuracy": 0.1808069348335266,
      "num_tokens": 48471984.0,
      "step": 22130
    },
    {
      "entropy": 5.68801326751709,
      "epoch": 2.1968042874156413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00045197180079593836,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17580154687166213,
      "num_tokens": 48482814.0,
      "step": 22135
    },
    {
      "entropy": 5.696237421035766,
      "epoch": 2.1973005160778087,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045194991831961777,
      "loss": 5.0505,
      "mean_token_accuracy": 0.18246175348758698,
      "num_tokens": 48493670.0,
      "step": 22140
    },
    {
      "entropy": 5.629911661148071,
      "epoch": 2.197796744739976,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004519280314553206,
      "loss": 5.0305,
      "mean_token_accuracy": 0.17295643389225007,
      "num_tokens": 48504758.0,
      "step": 22145
    },
    {
      "entropy": 5.672423362731934,
      "epoch": 2.1982929734021437,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004519061402035894,
      "loss": 5.0557,
      "mean_token_accuracy": 0.17217237800359725,
      "num_tokens": 48515846.0,
      "step": 22150
    },
    {
      "entropy": 5.6648975849151615,
      "epoch": 2.198789202064311,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045188424456496723,
      "loss": 5.0439,
      "mean_token_accuracy": 0.18773962408304215,
      "num_tokens": 48526136.0,
      "step": 22155
    },
    {
      "entropy": 5.694012403488159,
      "epoch": 2.1992854307264786,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045186234453999693,
      "loss": 5.153,
      "mean_token_accuracy": 0.17723034918308259,
      "num_tokens": 48536031.0,
      "step": 22160
    },
    {
      "entropy": 5.536454534530639,
      "epoch": 2.199781659388646,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045184044012922154,
      "loss": 4.9822,
      "mean_token_accuracy": 0.18140684813261032,
      "num_tokens": 48546965.0,
      "step": 22165
    },
    {
      "entropy": 5.653983879089355,
      "epoch": 2.200277888050814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045181853133318425,
      "loss": 5.0487,
      "mean_token_accuracy": 0.1789536029100418,
      "num_tokens": 48557637.0,
      "step": 22170
    },
    {
      "entropy": 5.66221661567688,
      "epoch": 2.2007741167129815,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004517966181524285,
      "loss": 5.0909,
      "mean_token_accuracy": 0.18295980989933014,
      "num_tokens": 48568578.0,
      "step": 22175
    },
    {
      "entropy": 5.694706916809082,
      "epoch": 2.201270345375149,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004517747005874975,
      "loss": 5.1478,
      "mean_token_accuracy": 0.1764875441789627,
      "num_tokens": 48581782.0,
      "step": 22180
    },
    {
      "entropy": 5.579651737213135,
      "epoch": 2.2017665740373165,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004517527786389349,
      "loss": 5.0021,
      "mean_token_accuracy": 0.19264275431632996,
      "num_tokens": 48592432.0,
      "step": 22185
    },
    {
      "entropy": 5.596989011764526,
      "epoch": 2.202262802699484,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045173085230728417,
      "loss": 5.0299,
      "mean_token_accuracy": 0.18266889303922654,
      "num_tokens": 48603251.0,
      "step": 22190
    },
    {
      "entropy": 5.678913116455078,
      "epoch": 2.2027590313616514,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004517089215930892,
      "loss": 5.1279,
      "mean_token_accuracy": 0.1754453420639038,
      "num_tokens": 48616118.0,
      "step": 22195
    },
    {
      "entropy": 5.68898572921753,
      "epoch": 2.203255260023819,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045168698649689375,
      "loss": 5.0868,
      "mean_token_accuracy": 0.17709010392427443,
      "num_tokens": 48626903.0,
      "step": 22200
    },
    {
      "entropy": 5.591966676712036,
      "epoch": 2.2037514886859864,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045166504701924176,
      "loss": 4.971,
      "mean_token_accuracy": 0.1863262340426445,
      "num_tokens": 48638186.0,
      "step": 22205
    },
    {
      "entropy": 5.645975494384766,
      "epoch": 2.204247717348154,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00045164310316067727,
      "loss": 5.0884,
      "mean_token_accuracy": 0.17506727427244187,
      "num_tokens": 48649948.0,
      "step": 22210
    },
    {
      "entropy": 5.6352808475494385,
      "epoch": 2.2047439460103218,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004516211549217444,
      "loss": 4.9944,
      "mean_token_accuracy": 0.18381980955600738,
      "num_tokens": 48660869.0,
      "step": 22215
    },
    {
      "entropy": 5.585766839981079,
      "epoch": 2.2052401746724892,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004515992023029875,
      "loss": 5.0353,
      "mean_token_accuracy": 0.17918121367692946,
      "num_tokens": 48673383.0,
      "step": 22220
    },
    {
      "entropy": 5.559322834014893,
      "epoch": 2.2057364033346567,
      "grad_norm": 1.078125,
      "learning_rate": 0.000451577245304951,
      "loss": 4.9842,
      "mean_token_accuracy": 0.18749285489320755,
      "num_tokens": 48684071.0,
      "step": 22225
    },
    {
      "entropy": 5.650371456146241,
      "epoch": 2.206232631996824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004515552839281792,
      "loss": 5.022,
      "mean_token_accuracy": 0.17627158612012864,
      "num_tokens": 48695432.0,
      "step": 22230
    },
    {
      "entropy": 5.64069128036499,
      "epoch": 2.2067288606589917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004515333181732169,
      "loss": 5.0413,
      "mean_token_accuracy": 0.17949969470500945,
      "num_tokens": 48706102.0,
      "step": 22235
    },
    {
      "entropy": 5.6064605712890625,
      "epoch": 2.207225089321159,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045151134804060863,
      "loss": 5.0636,
      "mean_token_accuracy": 0.17625881731510162,
      "num_tokens": 48717748.0,
      "step": 22240
    },
    {
      "entropy": 5.613475799560547,
      "epoch": 2.2077213179833266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004514893735308993,
      "loss": 4.9862,
      "mean_token_accuracy": 0.18516102880239488,
      "num_tokens": 48728941.0,
      "step": 22245
    },
    {
      "entropy": 5.584397411346435,
      "epoch": 2.208217546645494,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004514673946446338,
      "loss": 5.0417,
      "mean_token_accuracy": 0.18154095113277435,
      "num_tokens": 48738565.0,
      "step": 22250
    },
    {
      "entropy": 5.622395038604736,
      "epoch": 2.208713775307662,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004514454113823572,
      "loss": 5.0208,
      "mean_token_accuracy": 0.1778368502855301,
      "num_tokens": 48749537.0,
      "step": 22255
    },
    {
      "entropy": 5.604114294052124,
      "epoch": 2.2092100039698295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004514234237446146,
      "loss": 5.0316,
      "mean_token_accuracy": 0.17617825120687486,
      "num_tokens": 48760388.0,
      "step": 22260
    },
    {
      "entropy": 5.588932800292969,
      "epoch": 2.209706232631997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045140143173195117,
      "loss": 5.0529,
      "mean_token_accuracy": 0.17897360324859618,
      "num_tokens": 48770551.0,
      "step": 22265
    },
    {
      "entropy": 5.5990638256073,
      "epoch": 2.2102024612941644,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004513794353449124,
      "loss": 5.0061,
      "mean_token_accuracy": 0.17961354106664656,
      "num_tokens": 48780739.0,
      "step": 22270
    },
    {
      "entropy": 5.644645118713379,
      "epoch": 2.210698689956332,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004513574345840437,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17622489482164383,
      "num_tokens": 48791478.0,
      "step": 22275
    },
    {
      "entropy": 5.667417860031128,
      "epoch": 2.2111949186184994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513354294498907,
      "loss": 5.0845,
      "mean_token_accuracy": 0.1760719195008278,
      "num_tokens": 48802202.0,
      "step": 22280
    },
    {
      "entropy": 5.627660942077637,
      "epoch": 2.211691147280667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045131341994299895,
      "loss": 5.0561,
      "mean_token_accuracy": 0.18006281852722167,
      "num_tokens": 48813619.0,
      "step": 22285
    },
    {
      "entropy": 5.626554775238037,
      "epoch": 2.2121873759428343,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004512914060639143,
      "loss": 5.0134,
      "mean_token_accuracy": 0.18709145337343216,
      "num_tokens": 48824951.0,
      "step": 22290
    },
    {
      "entropy": 5.629577684402466,
      "epoch": 2.212683604605002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004512693878131827,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18558180183172227,
      "num_tokens": 48834328.0,
      "step": 22295
    },
    {
      "entropy": 5.6095579147338865,
      "epoch": 2.2131798332671697,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045124736519135003,
      "loss": 5.0561,
      "mean_token_accuracy": 0.1786453068256378,
      "num_tokens": 48843880.0,
      "step": 22300
    },
    {
      "entropy": 5.62524938583374,
      "epoch": 2.213676061929337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045122533819896257,
      "loss": 5.0655,
      "mean_token_accuracy": 0.18093697875738143,
      "num_tokens": 48854893.0,
      "step": 22305
    },
    {
      "entropy": 5.665227460861206,
      "epoch": 2.2141722905915047,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004512033068365664,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17631811648607254,
      "num_tokens": 48867308.0,
      "step": 22310
    },
    {
      "entropy": 5.690033864974976,
      "epoch": 2.214668519253672,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004511812711047079,
      "loss": 5.118,
      "mean_token_accuracy": 0.17744719088077546,
      "num_tokens": 48880222.0,
      "step": 22315
    },
    {
      "entropy": 5.682362747192383,
      "epoch": 2.2151647479158396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045115923100393356,
      "loss": 5.0518,
      "mean_token_accuracy": 0.17829504758119583,
      "num_tokens": 48890345.0,
      "step": 22320
    },
    {
      "entropy": 5.601732778549194,
      "epoch": 2.215660976578007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045113718653478984,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18404218703508377,
      "num_tokens": 48901267.0,
      "step": 22325
    },
    {
      "entropy": 5.640412521362305,
      "epoch": 2.2161572052401746,
      "grad_norm": 1.125,
      "learning_rate": 0.0004511151376978234,
      "loss": 5.067,
      "mean_token_accuracy": 0.18079798072576522,
      "num_tokens": 48911739.0,
      "step": 22330
    },
    {
      "entropy": 5.651979207992554,
      "epoch": 2.216653433902342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045109308449358114,
      "loss": 5.006,
      "mean_token_accuracy": 0.18227063864469528,
      "num_tokens": 48922376.0,
      "step": 22335
    },
    {
      "entropy": 5.5875331401824955,
      "epoch": 2.21714966256451,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004510710269226098,
      "loss": 4.9377,
      "mean_token_accuracy": 0.18481998443603515,
      "num_tokens": 48934101.0,
      "step": 22340
    },
    {
      "entropy": 5.572833776473999,
      "epoch": 2.2176458912266774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004510489649854564,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18504538983106614,
      "num_tokens": 48944684.0,
      "step": 22345
    },
    {
      "entropy": 5.649921846389771,
      "epoch": 2.218142119888845,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000451026898682668,
      "loss": 5.0511,
      "mean_token_accuracy": 0.17446654438972473,
      "num_tokens": 48955255.0,
      "step": 22350
    },
    {
      "entropy": 5.60948600769043,
      "epoch": 2.2186383485510124,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004510048280147918,
      "loss": 5.044,
      "mean_token_accuracy": 0.17908626347780227,
      "num_tokens": 48967034.0,
      "step": 22355
    },
    {
      "entropy": 5.6313989639282225,
      "epoch": 2.21913457721318,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004509827529823751,
      "loss": 5.1471,
      "mean_token_accuracy": 0.17481691241264344,
      "num_tokens": 48977903.0,
      "step": 22360
    },
    {
      "entropy": 5.635047769546508,
      "epoch": 2.2196308058753473,
      "grad_norm": 1.0,
      "learning_rate": 0.00045096067358596536,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18255043625831605,
      "num_tokens": 48988889.0,
      "step": 22365
    },
    {
      "entropy": 5.73606767654419,
      "epoch": 2.220127034537515,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045093858982611006,
      "loss": 5.1432,
      "mean_token_accuracy": 0.1664557695388794,
      "num_tokens": 48998946.0,
      "step": 22370
    },
    {
      "entropy": 5.656725025177002,
      "epoch": 2.2206232631996823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004509165017033569,
      "loss": 5.0247,
      "mean_token_accuracy": 0.1868304967880249,
      "num_tokens": 49008141.0,
      "step": 22375
    },
    {
      "entropy": 5.575255870819092,
      "epoch": 2.2211194918618498,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045089440921825354,
      "loss": 4.9386,
      "mean_token_accuracy": 0.189548259973526,
      "num_tokens": 49019387.0,
      "step": 22380
    },
    {
      "entropy": 5.542311811447144,
      "epoch": 2.2216157205240172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045087231237134787,
      "loss": 5.0023,
      "mean_token_accuracy": 0.17907448410987853,
      "num_tokens": 49030280.0,
      "step": 22385
    },
    {
      "entropy": 5.638730573654175,
      "epoch": 2.222111949186185,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045085021116318783,
      "loss": 4.9875,
      "mean_token_accuracy": 0.18551590591669082,
      "num_tokens": 49041766.0,
      "step": 22390
    },
    {
      "entropy": 5.727093267440796,
      "epoch": 2.2226081778483526,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508281055943214,
      "loss": 5.2232,
      "mean_token_accuracy": 0.16851307451725006,
      "num_tokens": 49052931.0,
      "step": 22395
    },
    {
      "entropy": 5.613234663009644,
      "epoch": 2.22310440651052,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004508059956652969,
      "loss": 5.0174,
      "mean_token_accuracy": 0.17703485786914824,
      "num_tokens": 49064153.0,
      "step": 22400
    },
    {
      "entropy": 5.617754602432251,
      "epoch": 2.2236006351726876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004507838813766625,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18403135538101195,
      "num_tokens": 49075857.0,
      "step": 22405
    },
    {
      "entropy": 5.718054151535034,
      "epoch": 2.224096863834855,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004507617627289666,
      "loss": 5.084,
      "mean_token_accuracy": 0.1806526944041252,
      "num_tokens": 49086632.0,
      "step": 22410
    },
    {
      "entropy": 5.638623666763306,
      "epoch": 2.2245930924970225,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004507396397227577,
      "loss": 5.0309,
      "mean_token_accuracy": 0.18039307743310928,
      "num_tokens": 49098086.0,
      "step": 22415
    },
    {
      "entropy": 5.560442876815796,
      "epoch": 2.22508932115919,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045071751235858454,
      "loss": 4.9504,
      "mean_token_accuracy": 0.1862817019224167,
      "num_tokens": 49108857.0,
      "step": 22420
    },
    {
      "entropy": 5.58989839553833,
      "epoch": 2.2255855498213575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004506953806369956,
      "loss": 4.981,
      "mean_token_accuracy": 0.1833336904644966,
      "num_tokens": 49119609.0,
      "step": 22425
    },
    {
      "entropy": 5.637095737457275,
      "epoch": 2.2260817784835254,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004506732445585399,
      "loss": 5.0731,
      "mean_token_accuracy": 0.1796647161245346,
      "num_tokens": 49130047.0,
      "step": 22430
    },
    {
      "entropy": 5.595197296142578,
      "epoch": 2.226578007145693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004506511041237662,
      "loss": 5.0163,
      "mean_token_accuracy": 0.18016178160905838,
      "num_tokens": 49140517.0,
      "step": 22435
    },
    {
      "entropy": 5.645664024353027,
      "epoch": 2.2270742358078603,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045062895933322366,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1824939727783203,
      "num_tokens": 49151497.0,
      "step": 22440
    },
    {
      "entropy": 5.547566699981689,
      "epoch": 2.227570464470028,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004506068101874614,
      "loss": 4.9635,
      "mean_token_accuracy": 0.18072703927755357,
      "num_tokens": 49161568.0,
      "step": 22445
    },
    {
      "entropy": 5.6609416007995605,
      "epoch": 2.2280666931321953,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00045058465668702856,
      "loss": 5.0812,
      "mean_token_accuracy": 0.17823317050933837,
      "num_tokens": 49173554.0,
      "step": 22450
    },
    {
      "entropy": 5.715932893753052,
      "epoch": 2.2285629217943628,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045056249883247457,
      "loss": 5.0644,
      "mean_token_accuracy": 0.17890566885471343,
      "num_tokens": 49185399.0,
      "step": 22455
    },
    {
      "entropy": 5.640054845809937,
      "epoch": 2.2290591504565302,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045054033662434903,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1838201329112053,
      "num_tokens": 49195489.0,
      "step": 22460
    },
    {
      "entropy": 5.540700197219849,
      "epoch": 2.2295553791186977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045051817006320127,
      "loss": 4.9755,
      "mean_token_accuracy": 0.18128081858158113,
      "num_tokens": 49204652.0,
      "step": 22465
    },
    {
      "entropy": 5.579576826095581,
      "epoch": 2.230051607780865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045049599914958124,
      "loss": 5.0567,
      "mean_token_accuracy": 0.17455294877290725,
      "num_tokens": 49215420.0,
      "step": 22470
    },
    {
      "entropy": 5.641563653945923,
      "epoch": 2.230547836443033,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004504738238840385,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18484584987163544,
      "num_tokens": 49225375.0,
      "step": 22475
    },
    {
      "entropy": 5.653909158706665,
      "epoch": 2.2310440651052006,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004504516442671231,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17783950865268708,
      "num_tokens": 49238399.0,
      "step": 22480
    },
    {
      "entropy": 5.62728362083435,
      "epoch": 2.231540293767368,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00045042946029938495,
      "loss": 5.0102,
      "mean_token_accuracy": 0.17875318974256516,
      "num_tokens": 49251069.0,
      "step": 22485
    },
    {
      "entropy": 5.626343154907227,
      "epoch": 2.2320365224295355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045040727198137416,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17983463704586028,
      "num_tokens": 49263859.0,
      "step": 22490
    },
    {
      "entropy": 5.650638580322266,
      "epoch": 2.232532751091703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045038507931364103,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18408472687005997,
      "num_tokens": 49274846.0,
      "step": 22495
    },
    {
      "entropy": 5.5992992401123045,
      "epoch": 2.2330289797538705,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045036288229673583,
      "loss": 4.9788,
      "mean_token_accuracy": 0.1895568326115608,
      "num_tokens": 49285197.0,
      "step": 22500
    },
    {
      "entropy": 5.552833032608032,
      "epoch": 2.233525208416038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045034068093120907,
      "loss": 4.9816,
      "mean_token_accuracy": 0.1787910208106041,
      "num_tokens": 49296238.0,
      "step": 22505
    },
    {
      "entropy": 5.706791925430298,
      "epoch": 2.2340214370782054,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004503184752176112,
      "loss": 5.2155,
      "mean_token_accuracy": 0.17185685485601426,
      "num_tokens": 49308262.0,
      "step": 22510
    },
    {
      "entropy": 5.631393957138061,
      "epoch": 2.2345176657403734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000450296265156493,
      "loss": 4.9867,
      "mean_token_accuracy": 0.18636642545461654,
      "num_tokens": 49318881.0,
      "step": 22515
    },
    {
      "entropy": 5.699501466751099,
      "epoch": 2.235013894402541,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045027405074840506,
      "loss": 5.1327,
      "mean_token_accuracy": 0.17862297594547272,
      "num_tokens": 49329470.0,
      "step": 22520
    },
    {
      "entropy": 5.623202610015869,
      "epoch": 2.2355101230647083,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004502518319938984,
      "loss": 5.0274,
      "mean_token_accuracy": 0.1852976694703102,
      "num_tokens": 49339806.0,
      "step": 22525
    },
    {
      "entropy": 5.748794651031494,
      "epoch": 2.236006351726876,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045022960889352394,
      "loss": 5.1393,
      "mean_token_accuracy": 0.18276335150003434,
      "num_tokens": 49350295.0,
      "step": 22530
    },
    {
      "entropy": 5.644366645812989,
      "epoch": 2.2365025803890433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045020738144783273,
      "loss": 5.0311,
      "mean_token_accuracy": 0.18204993456602098,
      "num_tokens": 49360735.0,
      "step": 22535
    },
    {
      "entropy": 5.595080947875976,
      "epoch": 2.2369988090512107,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004501851496573761,
      "loss": 4.9948,
      "mean_token_accuracy": 0.18389134705066681,
      "num_tokens": 49372278.0,
      "step": 22540
    },
    {
      "entropy": 5.6409159183502195,
      "epoch": 2.237495037713378,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004501629135227051,
      "loss": 5.087,
      "mean_token_accuracy": 0.17414968609809875,
      "num_tokens": 49382842.0,
      "step": 22545
    },
    {
      "entropy": 5.647449588775634,
      "epoch": 2.2379912663755457,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004501406730443714,
      "loss": 5.1006,
      "mean_token_accuracy": 0.1738020345568657,
      "num_tokens": 49394633.0,
      "step": 22550
    },
    {
      "entropy": 5.686812114715576,
      "epoch": 2.238487495037713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045011842822292636,
      "loss": 5.0682,
      "mean_token_accuracy": 0.179960760474205,
      "num_tokens": 49404731.0,
      "step": 22555
    },
    {
      "entropy": 5.6899841785430905,
      "epoch": 2.238983723699881,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004500961790589216,
      "loss": 5.048,
      "mean_token_accuracy": 0.179903781414032,
      "num_tokens": 49415918.0,
      "step": 22560
    },
    {
      "entropy": 5.612211513519287,
      "epoch": 2.2394799523620486,
      "grad_norm": 1.078125,
      "learning_rate": 0.000450073925552909,
      "loss": 4.9917,
      "mean_token_accuracy": 0.18340194076299668,
      "num_tokens": 49427458.0,
      "step": 22565
    },
    {
      "entropy": 5.6077086448669435,
      "epoch": 2.239976181024216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004500516677054403,
      "loss": 5.0215,
      "mean_token_accuracy": 0.18465539067983627,
      "num_tokens": 49437696.0,
      "step": 22570
    },
    {
      "entropy": 5.688495540618897,
      "epoch": 2.2404724096863835,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045002940551706726,
      "loss": 5.229,
      "mean_token_accuracy": 0.17361029833555222,
      "num_tokens": 49449220.0,
      "step": 22575
    },
    {
      "entropy": 5.649671936035157,
      "epoch": 2.240968638348551,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045000713898834225,
      "loss": 5.0527,
      "mean_token_accuracy": 0.1785306990146637,
      "num_tokens": 49459624.0,
      "step": 22580
    },
    {
      "entropy": 5.638653230667114,
      "epoch": 2.2414648670107185,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004499848681198173,
      "loss": 4.975,
      "mean_token_accuracy": 0.18571872413158416,
      "num_tokens": 49471813.0,
      "step": 22585
    },
    {
      "entropy": 5.623488283157348,
      "epoch": 2.241961095672886,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004499625929120447,
      "loss": 5.0484,
      "mean_token_accuracy": 0.18104502707719802,
      "num_tokens": 49482624.0,
      "step": 22590
    },
    {
      "entropy": 5.653807544708252,
      "epoch": 2.2424573243350534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004499403133655768,
      "loss": 5.074,
      "mean_token_accuracy": 0.17454354465007782,
      "num_tokens": 49493791.0,
      "step": 22595
    },
    {
      "entropy": 5.705524063110351,
      "epoch": 2.2429535529972213,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004499180294809661,
      "loss": 5.0997,
      "mean_token_accuracy": 0.17723260670900345,
      "num_tokens": 49504060.0,
      "step": 22600
    },
    {
      "entropy": 5.58730034828186,
      "epoch": 2.243449781659389,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044989574125876516,
      "loss": 4.9337,
      "mean_token_accuracy": 0.1884652242064476,
      "num_tokens": 49515962.0,
      "step": 22605
    },
    {
      "entropy": 5.645301723480225,
      "epoch": 2.2439460103215563,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004498734486995267,
      "loss": 5.0637,
      "mean_token_accuracy": 0.17523577511310579,
      "num_tokens": 49526676.0,
      "step": 22610
    },
    {
      "entropy": 5.635608386993408,
      "epoch": 2.2444422389837237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044985115180380355,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18099260777235032,
      "num_tokens": 49537347.0,
      "step": 22615
    },
    {
      "entropy": 5.638432312011719,
      "epoch": 2.2449384676458912,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044982885057214857,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17568125277757646,
      "num_tokens": 49547422.0,
      "step": 22620
    },
    {
      "entropy": 5.606238269805909,
      "epoch": 2.2454346963080587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044980654500511485,
      "loss": 4.9282,
      "mean_token_accuracy": 0.18857939243316652,
      "num_tokens": 49557250.0,
      "step": 22625
    },
    {
      "entropy": 5.708834266662597,
      "epoch": 2.245930924970226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044978423510325544,
      "loss": 5.1233,
      "mean_token_accuracy": 0.1779326468706131,
      "num_tokens": 49567106.0,
      "step": 22630
    },
    {
      "entropy": 5.6664872646331785,
      "epoch": 2.2464271536323936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004497619208671237,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17968458980321883,
      "num_tokens": 49578978.0,
      "step": 22635
    },
    {
      "entropy": 5.688622665405274,
      "epoch": 2.246923382294561,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497396022972728,
      "loss": 5.1071,
      "mean_token_accuracy": 0.17719468623399734,
      "num_tokens": 49589730.0,
      "step": 22640
    },
    {
      "entropy": 5.635559320449829,
      "epoch": 2.247419610956729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004497172793942564,
      "loss": 5.0193,
      "mean_token_accuracy": 0.1805146560072899,
      "num_tokens": 49600881.0,
      "step": 22645
    },
    {
      "entropy": 5.652336263656617,
      "epoch": 2.2479158396188965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004496949521586279,
      "loss": 5.0696,
      "mean_token_accuracy": 0.1818108692765236,
      "num_tokens": 49611120.0,
      "step": 22650
    },
    {
      "entropy": 5.663575029373169,
      "epoch": 2.248412068281064,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044967262059094097,
      "loss": 5.0867,
      "mean_token_accuracy": 0.17606960386037826,
      "num_tokens": 49622978.0,
      "step": 22655
    },
    {
      "entropy": 5.642402458190918,
      "epoch": 2.2489082969432315,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044965028469174956,
      "loss": 5.0701,
      "mean_token_accuracy": 0.18125383406877518,
      "num_tokens": 49633678.0,
      "step": 22660
    },
    {
      "entropy": 5.643183660507202,
      "epoch": 2.249404525605399,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044962794446160736,
      "loss": 5.0612,
      "mean_token_accuracy": 0.17257893085479736,
      "num_tokens": 49644886.0,
      "step": 22665
    },
    {
      "entropy": 5.666671943664551,
      "epoch": 2.2499007542675664,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044960559990106844,
      "loss": 5.0213,
      "mean_token_accuracy": 0.18839923590421676,
      "num_tokens": 49656790.0,
      "step": 22670
    },
    {
      "entropy": 5.662738227844239,
      "epoch": 2.250396982929734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004495832510106869,
      "loss": 5.0517,
      "mean_token_accuracy": 0.1765791043639183,
      "num_tokens": 49667943.0,
      "step": 22675
    },
    {
      "entropy": 5.621828365325928,
      "epoch": 2.2508932115919014,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004495608977910169,
      "loss": 5.0285,
      "mean_token_accuracy": 0.17645280212163925,
      "num_tokens": 49678652.0,
      "step": 22680
    },
    {
      "entropy": 5.634355735778809,
      "epoch": 2.2513894402540693,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004495385402426128,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18862853795289994,
      "num_tokens": 49688951.0,
      "step": 22685
    },
    {
      "entropy": 5.625527477264404,
      "epoch": 2.2518856689162368,
      "grad_norm": 1.0625,
      "learning_rate": 0.000449516178366029,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17789370119571685,
      "num_tokens": 49698750.0,
      "step": 22690
    },
    {
      "entropy": 5.653861379623413,
      "epoch": 2.2523818975784042,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044949381216182,
      "loss": 5.1435,
      "mean_token_accuracy": 0.17055310010910035,
      "num_tokens": 49709442.0,
      "step": 22695
    },
    {
      "entropy": 5.6716248989105225,
      "epoch": 2.2528781262405717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044947144163054036,
      "loss": 5.0405,
      "mean_token_accuracy": 0.18585855066776275,
      "num_tokens": 49720517.0,
      "step": 22700
    },
    {
      "entropy": 5.665954637527466,
      "epoch": 2.253374354902739,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004494490667727451,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1797198012471199,
      "num_tokens": 49731413.0,
      "step": 22705
    },
    {
      "entropy": 5.717906713485718,
      "epoch": 2.2538705835649067,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044942668758898873,
      "loss": 5.2044,
      "mean_token_accuracy": 0.16919995546340943,
      "num_tokens": 49742492.0,
      "step": 22710
    },
    {
      "entropy": 5.681395196914673,
      "epoch": 2.254366812227074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004494043040798265,
      "loss": 5.0444,
      "mean_token_accuracy": 0.1767629235982895,
      "num_tokens": 49753981.0,
      "step": 22715
    },
    {
      "entropy": 5.651780128479004,
      "epoch": 2.2548630408892416,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004493819162458133,
      "loss": 5.0657,
      "mean_token_accuracy": 0.1767941027879715,
      "num_tokens": 49764718.0,
      "step": 22720
    },
    {
      "entropy": 5.649694204330444,
      "epoch": 2.255359269551409,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004493595240875043,
      "loss": 5.0362,
      "mean_token_accuracy": 0.17882635295391083,
      "num_tokens": 49775361.0,
      "step": 22725
    },
    {
      "entropy": 5.596966075897217,
      "epoch": 2.255855498213577,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044933712760545486,
      "loss": 5.0237,
      "mean_token_accuracy": 0.17738897502422332,
      "num_tokens": 49786670.0,
      "step": 22730
    },
    {
      "entropy": 5.663456296920776,
      "epoch": 2.2563517268757445,
      "grad_norm": 1.125,
      "learning_rate": 0.0004493147268002203,
      "loss": 5.1482,
      "mean_token_accuracy": 0.176592093706131,
      "num_tokens": 49796514.0,
      "step": 22735
    },
    {
      "entropy": 5.5907410144805905,
      "epoch": 2.256847955537912,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044929232167235613,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1880721405148506,
      "num_tokens": 49807042.0,
      "step": 22740
    },
    {
      "entropy": 5.643150186538696,
      "epoch": 2.2573441842000794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004492699122224179,
      "loss": 5.0674,
      "mean_token_accuracy": 0.1774784505367279,
      "num_tokens": 49818779.0,
      "step": 22745
    },
    {
      "entropy": 5.644111061096192,
      "epoch": 2.257840412862247,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044924749845096144,
      "loss": 5.0724,
      "mean_token_accuracy": 0.179672472178936,
      "num_tokens": 49829901.0,
      "step": 22750
    },
    {
      "entropy": 5.607132196426392,
      "epoch": 2.2583366415244144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004492250803585425,
      "loss": 5.0566,
      "mean_token_accuracy": 0.18401563167572021,
      "num_tokens": 49840837.0,
      "step": 22755
    },
    {
      "entropy": 5.648698902130127,
      "epoch": 2.258832870186582,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004492026579457169,
      "loss": 5.0218,
      "mean_token_accuracy": 0.17677749544382096,
      "num_tokens": 49853052.0,
      "step": 22760
    },
    {
      "entropy": 5.6285943508148195,
      "epoch": 2.2593290988487493,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004491802312130408,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17682956159114838,
      "num_tokens": 49864210.0,
      "step": 22765
    },
    {
      "entropy": 5.595581007003784,
      "epoch": 2.2598253275109172,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004491578001610703,
      "loss": 5.0016,
      "mean_token_accuracy": 0.1850048139691353,
      "num_tokens": 49874995.0,
      "step": 22770
    },
    {
      "entropy": 5.542532825469971,
      "epoch": 2.2603215561730847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044913536479036163,
      "loss": 5.0053,
      "mean_token_accuracy": 0.18445322215557097,
      "num_tokens": 49886240.0,
      "step": 22775
    },
    {
      "entropy": 5.632331800460816,
      "epoch": 2.260817784835252,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044911292510147117,
      "loss": 5.012,
      "mean_token_accuracy": 0.18260860592126846,
      "num_tokens": 49897580.0,
      "step": 22780
    },
    {
      "entropy": 5.623622083663941,
      "epoch": 2.2613140134974197,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004490904810949553,
      "loss": 5.0586,
      "mean_token_accuracy": 0.17609767913818358,
      "num_tokens": 49907280.0,
      "step": 22785
    },
    {
      "entropy": 5.616292190551758,
      "epoch": 2.261810242159587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044906803277137064,
      "loss": 5.0718,
      "mean_token_accuracy": 0.18542475551366805,
      "num_tokens": 49917552.0,
      "step": 22790
    },
    {
      "entropy": 5.627035999298096,
      "epoch": 2.2623064708217546,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004490455801312738,
      "loss": 5.0419,
      "mean_token_accuracy": 0.17584687769412993,
      "num_tokens": 49930610.0,
      "step": 22795
    },
    {
      "entropy": 5.637661170959473,
      "epoch": 2.262802699483922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004490231231752217,
      "loss": 5.08,
      "mean_token_accuracy": 0.1752068430185318,
      "num_tokens": 49941054.0,
      "step": 22800
    },
    {
      "entropy": 5.5380415439605715,
      "epoch": 2.2632989281460896,
      "grad_norm": 1.0,
      "learning_rate": 0.0004490006619037711,
      "loss": 4.9411,
      "mean_token_accuracy": 0.18913636058568956,
      "num_tokens": 49950885.0,
      "step": 22805
    },
    {
      "entropy": 5.605648326873779,
      "epoch": 2.263795156808257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044897819631747903,
      "loss": 5.0467,
      "mean_token_accuracy": 0.1791937232017517,
      "num_tokens": 49961723.0,
      "step": 22810
    },
    {
      "entropy": 5.597976875305176,
      "epoch": 2.2642913854704245,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044895572641690256,
      "loss": 5.0117,
      "mean_token_accuracy": 0.17698830962181092,
      "num_tokens": 49974124.0,
      "step": 22815
    },
    {
      "entropy": 5.642538022994995,
      "epoch": 2.2647876141325924,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000448933252202599,
      "loss": 5.0477,
      "mean_token_accuracy": 0.17989098876714707,
      "num_tokens": 49985792.0,
      "step": 22820
    },
    {
      "entropy": 5.597882890701294,
      "epoch": 2.26528384279476,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044891077367512544,
      "loss": 4.9828,
      "mean_token_accuracy": 0.1851254805922508,
      "num_tokens": 49996935.0,
      "step": 22825
    },
    {
      "entropy": 5.741730117797852,
      "epoch": 2.2657800714569274,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004488882908350396,
      "loss": 5.1524,
      "mean_token_accuracy": 0.1694575607776642,
      "num_tokens": 50007545.0,
      "step": 22830
    },
    {
      "entropy": 5.694405889511108,
      "epoch": 2.266276300119095,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004488658036828987,
      "loss": 5.0573,
      "mean_token_accuracy": 0.1819647118449211,
      "num_tokens": 50018797.0,
      "step": 22835
    },
    {
      "entropy": 5.620260381698609,
      "epoch": 2.2667725287812623,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488433122192606,
      "loss": 5.0117,
      "mean_token_accuracy": 0.1850498214364052,
      "num_tokens": 50030884.0,
      "step": 22840
    },
    {
      "entropy": 5.704599761962891,
      "epoch": 2.26726875744343,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044882081644468286,
      "loss": 5.1686,
      "mean_token_accuracy": 0.17058954536914825,
      "num_tokens": 50042327.0,
      "step": 22845
    },
    {
      "entropy": 5.681212759017944,
      "epoch": 2.2677649861055973,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044879831635972345,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1823091596364975,
      "num_tokens": 50053733.0,
      "step": 22850
    },
    {
      "entropy": 5.6102375984191895,
      "epoch": 2.268261214767765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004487758119649403,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18487163186073302,
      "num_tokens": 50064153.0,
      "step": 22855
    },
    {
      "entropy": 5.53473916053772,
      "epoch": 2.2687574434299327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044875330326089153,
      "loss": 4.9767,
      "mean_token_accuracy": 0.1909520998597145,
      "num_tokens": 50074397.0,
      "step": 22860
    },
    {
      "entropy": 5.561961603164673,
      "epoch": 2.2692536720921,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004487307902481352,
      "loss": 4.9718,
      "mean_token_accuracy": 0.1824539929628372,
      "num_tokens": 50084074.0,
      "step": 22865
    },
    {
      "entropy": 5.646994352340698,
      "epoch": 2.2697499007542676,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044870827292722967,
      "loss": 5.0,
      "mean_token_accuracy": 0.17835124284029008,
      "num_tokens": 50094535.0,
      "step": 22870
    },
    {
      "entropy": 5.673174524307251,
      "epoch": 2.270246129416435,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004486857512987333,
      "loss": 5.0775,
      "mean_token_accuracy": 0.1745072364807129,
      "num_tokens": 50105907.0,
      "step": 22875
    },
    {
      "entropy": 5.628068304061889,
      "epoch": 2.2707423580786026,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044866322536320447,
      "loss": 5.0432,
      "mean_token_accuracy": 0.1821552813053131,
      "num_tokens": 50117361.0,
      "step": 22880
    },
    {
      "entropy": 5.683154201507568,
      "epoch": 2.27123858674077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044864069512120185,
      "loss": 5.055,
      "mean_token_accuracy": 0.18515052348375322,
      "num_tokens": 50128598.0,
      "step": 22885
    },
    {
      "entropy": 5.666386508941651,
      "epoch": 2.2717348154029375,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044861816057328424,
      "loss": 5.055,
      "mean_token_accuracy": 0.17711730152368546,
      "num_tokens": 50140470.0,
      "step": 22890
    },
    {
      "entropy": 5.63054347038269,
      "epoch": 2.272231044065105,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044859562172001036,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17877207100391387,
      "num_tokens": 50150950.0,
      "step": 22895
    },
    {
      "entropy": 5.573122262954712,
      "epoch": 2.2727272727272725,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004485730785619391,
      "loss": 4.9616,
      "mean_token_accuracy": 0.1877134382724762,
      "num_tokens": 50161173.0,
      "step": 22900
    },
    {
      "entropy": 5.701330184936523,
      "epoch": 2.2732235013894404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004485505310996295,
      "loss": 5.1698,
      "mean_token_accuracy": 0.17315588742494584,
      "num_tokens": 50173874.0,
      "step": 22905
    },
    {
      "entropy": 5.75177493095398,
      "epoch": 2.273719730051608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485279793336406,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17068931162357331,
      "num_tokens": 50185123.0,
      "step": 22910
    },
    {
      "entropy": 5.656184339523316,
      "epoch": 2.2742159587137754,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004485054232645319,
      "loss": 5.0751,
      "mean_token_accuracy": 0.17948651164770127,
      "num_tokens": 50195699.0,
      "step": 22915
    },
    {
      "entropy": 5.703495979309082,
      "epoch": 2.274712187375943,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004484828628928625,
      "loss": 5.0915,
      "mean_token_accuracy": 0.1805381089448929,
      "num_tokens": 50206623.0,
      "step": 22920
    },
    {
      "entropy": 5.694164800643921,
      "epoch": 2.2752084160381103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044846029821919193,
      "loss": 5.1422,
      "mean_token_accuracy": 0.17442648559808732,
      "num_tokens": 50216086.0,
      "step": 22925
    },
    {
      "entropy": 5.592977142333984,
      "epoch": 2.275704644700278,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004484377292440797,
      "loss": 5.0412,
      "mean_token_accuracy": 0.17746899425983428,
      "num_tokens": 50228472.0,
      "step": 22930
    },
    {
      "entropy": 5.64102463722229,
      "epoch": 2.2762008733624453,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044841515596808553,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1758620873093605,
      "num_tokens": 50238607.0,
      "step": 22935
    },
    {
      "entropy": 5.646271371841431,
      "epoch": 2.276697102024613,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044839257839176914,
      "loss": 5.0275,
      "mean_token_accuracy": 0.1818509206175804,
      "num_tokens": 50249890.0,
      "step": 22940
    },
    {
      "entropy": 5.638804483413696,
      "epoch": 2.2771933306867806,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044836999651569055,
      "loss": 5.0244,
      "mean_token_accuracy": 0.18315569758415223,
      "num_tokens": 50260234.0,
      "step": 22945
    },
    {
      "entropy": 5.63285641670227,
      "epoch": 2.277689559348948,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004483474103404095,
      "loss": 5.0829,
      "mean_token_accuracy": 0.1786001667380333,
      "num_tokens": 50271605.0,
      "step": 22950
    },
    {
      "entropy": 5.634801769256592,
      "epoch": 2.2781857880111156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044832481986648633,
      "loss": 4.9485,
      "mean_token_accuracy": 0.18257802873849868,
      "num_tokens": 50281837.0,
      "step": 22955
    },
    {
      "entropy": 5.607528781890869,
      "epoch": 2.278682016673283,
      "grad_norm": 1.09375,
      "learning_rate": 0.000448302225094481,
      "loss": 5.0042,
      "mean_token_accuracy": 0.1833006203174591,
      "num_tokens": 50292524.0,
      "step": 22960
    },
    {
      "entropy": 5.647265672683716,
      "epoch": 2.2791782453354505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044827962602495393,
      "loss": 5.009,
      "mean_token_accuracy": 0.1810841903090477,
      "num_tokens": 50304009.0,
      "step": 22965
    },
    {
      "entropy": 5.670378017425537,
      "epoch": 2.279674473997618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044825702265846557,
      "loss": 5.0782,
      "mean_token_accuracy": 0.17964017242193223,
      "num_tokens": 50316103.0,
      "step": 22970
    },
    {
      "entropy": 5.653800582885742,
      "epoch": 2.2801707026597855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004482344149955763,
      "loss": 5.089,
      "mean_token_accuracy": 0.17728061228990555,
      "num_tokens": 50326127.0,
      "step": 22975
    },
    {
      "entropy": 5.573613119125366,
      "epoch": 2.280666931321953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004482118030368469,
      "loss": 4.9808,
      "mean_token_accuracy": 0.187232506275177,
      "num_tokens": 50337044.0,
      "step": 22980
    },
    {
      "entropy": 5.656872844696045,
      "epoch": 2.2811631599841204,
      "grad_norm": 0.984375,
      "learning_rate": 0.000448189186782838,
      "loss": 5.1806,
      "mean_token_accuracy": 0.17237078994512559,
      "num_tokens": 50347796.0,
      "step": 22985
    },
    {
      "entropy": 5.619082069396972,
      "epoch": 2.2816593886462884,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044816656623411044,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1801391839981079,
      "num_tokens": 50358454.0,
      "step": 22990
    },
    {
      "entropy": 5.57950382232666,
      "epoch": 2.282155617308456,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004481439413912252,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1814690425992012,
      "num_tokens": 50369415.0,
      "step": 22995
    },
    {
      "entropy": 5.6117839336395265,
      "epoch": 2.2826518459706233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044812131225474333,
      "loss": 5.0624,
      "mean_token_accuracy": 0.17429002821445466,
      "num_tokens": 50379977.0,
      "step": 23000
    },
    {
      "entropy": 5.66528730392456,
      "epoch": 2.283148074632791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004480986788252259,
      "loss": 5.061,
      "mean_token_accuracy": 0.182868193089962,
      "num_tokens": 50390465.0,
      "step": 23005
    },
    {
      "entropy": 5.675006008148193,
      "epoch": 2.2836443032949583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004480760411032343,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17680550664663314,
      "num_tokens": 50402230.0,
      "step": 23010
    },
    {
      "entropy": 5.579122591018677,
      "epoch": 2.2841405319571257,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004480533990893298,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18718125224113463,
      "num_tokens": 50413819.0,
      "step": 23015
    },
    {
      "entropy": 5.779780578613281,
      "epoch": 2.284636760619293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004480307527840739,
      "loss": 5.236,
      "mean_token_accuracy": 0.17099516987800598,
      "num_tokens": 50424977.0,
      "step": 23020
    },
    {
      "entropy": 5.690476799011231,
      "epoch": 2.285132989281461,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004480081021880282,
      "loss": 5.0592,
      "mean_token_accuracy": 0.17775694727897645,
      "num_tokens": 50438357.0,
      "step": 23025
    },
    {
      "entropy": 5.6462665557861325,
      "epoch": 2.2856292179436286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004479854473017543,
      "loss": 5.0684,
      "mean_token_accuracy": 0.1778028890490532,
      "num_tokens": 50449081.0,
      "step": 23030
    },
    {
      "entropy": 5.554302787780761,
      "epoch": 2.286125446605796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044796278812581417,
      "loss": 4.9608,
      "mean_token_accuracy": 0.187720587849617,
      "num_tokens": 50460358.0,
      "step": 23035
    },
    {
      "entropy": 5.653097200393677,
      "epoch": 2.2866216752679636,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004479401246607695,
      "loss": 5.0454,
      "mean_token_accuracy": 0.1797054886817932,
      "num_tokens": 50470955.0,
      "step": 23040
    },
    {
      "entropy": 5.650041723251343,
      "epoch": 2.287117903930131,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004479174569071823,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18445621728897094,
      "num_tokens": 50482250.0,
      "step": 23045
    },
    {
      "entropy": 5.601606559753418,
      "epoch": 2.2876141325922985,
      "grad_norm": 1.125,
      "learning_rate": 0.0004478947848656149,
      "loss": 5.0483,
      "mean_token_accuracy": 0.18256603926420212,
      "num_tokens": 50493868.0,
      "step": 23050
    },
    {
      "entropy": 5.638349723815918,
      "epoch": 2.288110361254466,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044787210853662943,
      "loss": 5.0801,
      "mean_token_accuracy": 0.18316202610731125,
      "num_tokens": 50504012.0,
      "step": 23055
    },
    {
      "entropy": 5.666625070571899,
      "epoch": 2.2886065899166335,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004478494279207881,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18483115136623382,
      "num_tokens": 50515503.0,
      "step": 23060
    },
    {
      "entropy": 5.6030525207519535,
      "epoch": 2.289102818578801,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044782674301865346,
      "loss": 5.0445,
      "mean_token_accuracy": 0.18222351372241974,
      "num_tokens": 50526180.0,
      "step": 23065
    },
    {
      "entropy": 5.63634934425354,
      "epoch": 2.2895990472409684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004478040538307879,
      "loss": 5.0555,
      "mean_token_accuracy": 0.17107984125614167,
      "num_tokens": 50536711.0,
      "step": 23070
    },
    {
      "entropy": 5.698263502120971,
      "epoch": 2.2900952759031363,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004477813603577543,
      "loss": 5.1159,
      "mean_token_accuracy": 0.17280070930719377,
      "num_tokens": 50547736.0,
      "step": 23075
    },
    {
      "entropy": 5.6405110359191895,
      "epoch": 2.290591504565304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044775866260011516,
      "loss": 5.0249,
      "mean_token_accuracy": 0.18725291937589644,
      "num_tokens": 50558173.0,
      "step": 23080
    },
    {
      "entropy": 5.657270860671997,
      "epoch": 2.2910877332274713,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004477359605584335,
      "loss": 5.0463,
      "mean_token_accuracy": 0.17784691751003265,
      "num_tokens": 50568421.0,
      "step": 23085
    },
    {
      "entropy": 5.628287982940674,
      "epoch": 2.2915839618896388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044771325423327227,
      "loss": 5.0183,
      "mean_token_accuracy": 0.17980681210756302,
      "num_tokens": 50579182.0,
      "step": 23090
    },
    {
      "entropy": 5.643021488189698,
      "epoch": 2.2920801905518062,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004476905436251945,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18011582940816878,
      "num_tokens": 50590916.0,
      "step": 23095
    },
    {
      "entropy": 5.522994470596314,
      "epoch": 2.2925764192139737,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044766782873476333,
      "loss": 4.934,
      "mean_token_accuracy": 0.18718986213207245,
      "num_tokens": 50600682.0,
      "step": 23100
    },
    {
      "entropy": 5.711855554580689,
      "epoch": 2.293072647876141,
      "grad_norm": 1.015625,
      "learning_rate": 0.000447645109562542,
      "loss": 5.1893,
      "mean_token_accuracy": 0.1685623586177826,
      "num_tokens": 50612103.0,
      "step": 23105
    },
    {
      "entropy": 5.71285605430603,
      "epoch": 2.293568876538309,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044762238610909414,
      "loss": 5.1263,
      "mean_token_accuracy": 0.17579413205385208,
      "num_tokens": 50623894.0,
      "step": 23110
    },
    {
      "entropy": 5.723560571670532,
      "epoch": 2.2940651052004766,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004475996583749829,
      "loss": 5.0965,
      "mean_token_accuracy": 0.17755795121192933,
      "num_tokens": 50634495.0,
      "step": 23115
    },
    {
      "entropy": 5.696341562271118,
      "epoch": 2.294561333862644,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044757692636077215,
      "loss": 5.2227,
      "mean_token_accuracy": 0.17110807150602342,
      "num_tokens": 50645921.0,
      "step": 23120
    },
    {
      "entropy": 5.692655086517334,
      "epoch": 2.2950575625248115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044755419006702545,
      "loss": 5.0579,
      "mean_token_accuracy": 0.17497556358575822,
      "num_tokens": 50657047.0,
      "step": 23125
    },
    {
      "entropy": 5.722243452072144,
      "epoch": 2.295553791186979,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044753144949430667,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1799513429403305,
      "num_tokens": 50668271.0,
      "step": 23130
    },
    {
      "entropy": 5.594829559326172,
      "epoch": 2.2960500198491465,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004475087046431798,
      "loss": 5.0013,
      "mean_token_accuracy": 0.18612300753593444,
      "num_tokens": 50678870.0,
      "step": 23135
    },
    {
      "entropy": 5.610577535629273,
      "epoch": 2.296546248511314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004474859555142087,
      "loss": 5.0515,
      "mean_token_accuracy": 0.186531962454319,
      "num_tokens": 50690322.0,
      "step": 23140
    },
    {
      "entropy": 5.615605354309082,
      "epoch": 2.2970424771734814,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044746320210795753,
      "loss": 5.002,
      "mean_token_accuracy": 0.17781273275613785,
      "num_tokens": 50701958.0,
      "step": 23145
    },
    {
      "entropy": 5.637621545791626,
      "epoch": 2.297538705835649,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004474404444249906,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1812346950173378,
      "num_tokens": 50712962.0,
      "step": 23150
    },
    {
      "entropy": 5.610480928421021,
      "epoch": 2.2980349344978164,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044741768246587226,
      "loss": 5.0919,
      "mean_token_accuracy": 0.18343938738107682,
      "num_tokens": 50723442.0,
      "step": 23155
    },
    {
      "entropy": 5.593527460098267,
      "epoch": 2.2985311631599843,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004473949162311669,
      "loss": 4.9681,
      "mean_token_accuracy": 0.1834363728761673,
      "num_tokens": 50733188.0,
      "step": 23160
    },
    {
      "entropy": 5.6660497188568115,
      "epoch": 2.2990273918221518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473721457214391,
      "loss": 5.1066,
      "mean_token_accuracy": 0.17390185743570327,
      "num_tokens": 50744023.0,
      "step": 23165
    },
    {
      "entropy": 5.74712233543396,
      "epoch": 2.2995236204843192,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473493709372535,
      "loss": 5.268,
      "mean_token_accuracy": 0.17860615253448486,
      "num_tokens": 50756807.0,
      "step": 23170
    },
    {
      "entropy": 5.7035431385040285,
      "epoch": 2.3000198491464867,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004473265918791749,
      "loss": 5.1563,
      "mean_token_accuracy": 0.1772083654999733,
      "num_tokens": 50768981.0,
      "step": 23175
    },
    {
      "entropy": 5.72165813446045,
      "epoch": 2.300516077808654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004473038085477682,
      "loss": 5.0585,
      "mean_token_accuracy": 0.1852637618780136,
      "num_tokens": 50780111.0,
      "step": 23180
    },
    {
      "entropy": 5.679054307937622,
      "epoch": 2.3010123064708217,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044728102094359826,
      "loss": 5.0961,
      "mean_token_accuracy": 0.17677570432424544,
      "num_tokens": 50790318.0,
      "step": 23185
    },
    {
      "entropy": 5.649969625473022,
      "epoch": 2.301508535132989,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004472582290672303,
      "loss": 5.1148,
      "mean_token_accuracy": 0.1793668821454048,
      "num_tokens": 50800734.0,
      "step": 23190
    },
    {
      "entropy": 5.57722315788269,
      "epoch": 2.3020047637951566,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044723543291922936,
      "loss": 5.0189,
      "mean_token_accuracy": 0.18673930317163467,
      "num_tokens": 50811236.0,
      "step": 23195
    },
    {
      "entropy": 5.585062503814697,
      "epoch": 2.3025009924573245,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044721263250016086,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18358812034130095,
      "num_tokens": 50821776.0,
      "step": 23200
    },
    {
      "entropy": 5.679781579971314,
      "epoch": 2.302997221119492,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044718982781059017,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1722346156835556,
      "num_tokens": 50834156.0,
      "step": 23205
    },
    {
      "entropy": 5.656577730178833,
      "epoch": 2.3034934497816595,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044716701885108275,
      "loss": 5.0284,
      "mean_token_accuracy": 0.1768994852900505,
      "num_tokens": 50845086.0,
      "step": 23210
    },
    {
      "entropy": 5.682739400863648,
      "epoch": 2.303989678443827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044714420562220423,
      "loss": 5.0467,
      "mean_token_accuracy": 0.18259551525115966,
      "num_tokens": 50858637.0,
      "step": 23215
    },
    {
      "entropy": 5.654863166809082,
      "epoch": 2.3044859071059944,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004471213881245204,
      "loss": 5.1573,
      "mean_token_accuracy": 0.1752387210726738,
      "num_tokens": 50871159.0,
      "step": 23220
    },
    {
      "entropy": 5.638574171066284,
      "epoch": 2.304982135768162,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000447098566358597,
      "loss": 5.0442,
      "mean_token_accuracy": 0.18241848200559616,
      "num_tokens": 50884072.0,
      "step": 23225
    },
    {
      "entropy": 5.6359607696533205,
      "epoch": 2.3054783644303294,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044707574032500005,
      "loss": 5.0455,
      "mean_token_accuracy": 0.1834094136953354,
      "num_tokens": 50896627.0,
      "step": 23230
    },
    {
      "entropy": 5.601624250411987,
      "epoch": 2.305974593092497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004470529100242955,
      "loss": 4.9634,
      "mean_token_accuracy": 0.18956994414329528,
      "num_tokens": 50907476.0,
      "step": 23235
    },
    {
      "entropy": 5.611721992492676,
      "epoch": 2.3064708217546643,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470300754570495,
      "loss": 5.0541,
      "mean_token_accuracy": 0.1789078712463379,
      "num_tokens": 50918104.0,
      "step": 23240
    },
    {
      "entropy": 5.613946962356567,
      "epoch": 2.3069670504168323,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004470072366238284,
      "loss": 5.0177,
      "mean_token_accuracy": 0.1800752893090248,
      "num_tokens": 50930575.0,
      "step": 23245
    },
    {
      "entropy": 5.630888271331787,
      "epoch": 2.3074632790789997,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044698439352519836,
      "loss": 4.9778,
      "mean_token_accuracy": 0.18117154091596605,
      "num_tokens": 50940592.0,
      "step": 23250
    },
    {
      "entropy": 5.623869800567627,
      "epoch": 2.307959507741167,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044696154616172604,
      "loss": 5.057,
      "mean_token_accuracy": 0.17613251358270646,
      "num_tokens": 50952134.0,
      "step": 23255
    },
    {
      "entropy": 5.641175079345703,
      "epoch": 2.3084557364033347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004469386945339779,
      "loss": 5.0981,
      "mean_token_accuracy": 0.17629600316286087,
      "num_tokens": 50962721.0,
      "step": 23260
    },
    {
      "entropy": 5.606606578826904,
      "epoch": 2.308951965065502,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004469158386425206,
      "loss": 5.0384,
      "mean_token_accuracy": 0.1785607874393463,
      "num_tokens": 50974893.0,
      "step": 23265
    },
    {
      "entropy": 5.655833196640015,
      "epoch": 2.3094481937276696,
      "grad_norm": 1.0,
      "learning_rate": 0.000446892978487921,
      "loss": 5.0085,
      "mean_token_accuracy": 0.17684974223375322,
      "num_tokens": 50986225.0,
      "step": 23270
    },
    {
      "entropy": 5.636183357238769,
      "epoch": 2.309944422389837,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044687011407074583,
      "loss": 5.0945,
      "mean_token_accuracy": 0.16760538518428802,
      "num_tokens": 50997924.0,
      "step": 23275
    },
    {
      "entropy": 5.6262876987457275,
      "epoch": 2.3104406510520046,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004468472453915623,
      "loss": 4.9914,
      "mean_token_accuracy": 0.1796787515282631,
      "num_tokens": 51008179.0,
      "step": 23280
    },
    {
      "entropy": 5.572395372390747,
      "epoch": 2.3109368797141725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004468243724509373,
      "loss": 5.007,
      "mean_token_accuracy": 0.18695622235536574,
      "num_tokens": 51018462.0,
      "step": 23285
    },
    {
      "entropy": 5.625366640090943,
      "epoch": 2.31143310837634,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004468014952494382,
      "loss": 5.0925,
      "mean_token_accuracy": 0.17605340629816055,
      "num_tokens": 51029769.0,
      "step": 23290
    },
    {
      "entropy": 5.667421007156372,
      "epoch": 2.3119293370385074,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044677861378763205,
      "loss": 5.0836,
      "mean_token_accuracy": 0.18066641539335251,
      "num_tokens": 51041058.0,
      "step": 23295
    },
    {
      "entropy": 5.650592517852783,
      "epoch": 2.312425565700675,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044675572806608655,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18077875673770905,
      "num_tokens": 51050958.0,
      "step": 23300
    },
    {
      "entropy": 5.621256256103516,
      "epoch": 2.3129217943628424,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044673283808536915,
      "loss": 5.0243,
      "mean_token_accuracy": 0.18562982529401778,
      "num_tokens": 51062459.0,
      "step": 23305
    },
    {
      "entropy": 5.7118219375610355,
      "epoch": 2.31341802302501,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004467099438460474,
      "loss": 5.2061,
      "mean_token_accuracy": 0.17626275569200517,
      "num_tokens": 51074596.0,
      "step": 23310
    },
    {
      "entropy": 5.697895193099976,
      "epoch": 2.3139142516871773,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044668704534868886,
      "loss": 5.08,
      "mean_token_accuracy": 0.17445161938667297,
      "num_tokens": 51085265.0,
      "step": 23315
    },
    {
      "entropy": 5.641272020339966,
      "epoch": 2.314410480349345,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044666414259386173,
      "loss": 5.0254,
      "mean_token_accuracy": 0.17729086428880692,
      "num_tokens": 51095986.0,
      "step": 23320
    },
    {
      "entropy": 5.6688135147094725,
      "epoch": 2.3149067090115123,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044664123558213367,
      "loss": 5.0415,
      "mean_token_accuracy": 0.17985792756080626,
      "num_tokens": 51107518.0,
      "step": 23325
    },
    {
      "entropy": 5.615531969070434,
      "epoch": 2.31540293767368,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004466183243140729,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18830569237470626,
      "num_tokens": 51119715.0,
      "step": 23330
    },
    {
      "entropy": 5.557191038131714,
      "epoch": 2.3158991663358477,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044659540879024746,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18636624962091447,
      "num_tokens": 51130801.0,
      "step": 23335
    },
    {
      "entropy": 5.592831087112427,
      "epoch": 2.316395394998015,
      "grad_norm": 1.0,
      "learning_rate": 0.0004465724890112256,
      "loss": 5.039,
      "mean_token_accuracy": 0.178148253262043,
      "num_tokens": 51142509.0,
      "step": 23340
    },
    {
      "entropy": 5.681394767761231,
      "epoch": 2.3168916236601826,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044654956497757576,
      "loss": 5.0777,
      "mean_token_accuracy": 0.18371689021587373,
      "num_tokens": 51154839.0,
      "step": 23345
    },
    {
      "entropy": 5.610626363754273,
      "epoch": 2.31738785232235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044652663668986634,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18541369438171387,
      "num_tokens": 51165293.0,
      "step": 23350
    },
    {
      "entropy": 5.609228801727295,
      "epoch": 2.3178840809845176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000446503704148666,
      "loss": 5.0958,
      "mean_token_accuracy": 0.17452152967453002,
      "num_tokens": 51176371.0,
      "step": 23355
    },
    {
      "entropy": 5.630225706100464,
      "epoch": 2.318380309646685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004464807673545433,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18457587957382202,
      "num_tokens": 51187240.0,
      "step": 23360
    },
    {
      "entropy": 5.732465744018555,
      "epoch": 2.3188765383088525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004464578263080671,
      "loss": 5.1067,
      "mean_token_accuracy": 0.17762676924467086,
      "num_tokens": 51199084.0,
      "step": 23365
    },
    {
      "entropy": 5.623921012878418,
      "epoch": 2.3193727669710205,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044643488100980623,
      "loss": 4.97,
      "mean_token_accuracy": 0.1862728327512741,
      "num_tokens": 51209885.0,
      "step": 23370
    },
    {
      "entropy": 5.626689529418945,
      "epoch": 2.319868995633188,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464119314603298,
      "loss": 5.0325,
      "mean_token_accuracy": 0.1868664488196373,
      "num_tokens": 51220644.0,
      "step": 23375
    },
    {
      "entropy": 5.6117236614227295,
      "epoch": 2.3203652242953554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044638897766020673,
      "loss": 5.0136,
      "mean_token_accuracy": 0.17931249290704726,
      "num_tokens": 51231342.0,
      "step": 23380
    },
    {
      "entropy": 5.590123844146729,
      "epoch": 2.320861452957523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004463660196100064,
      "loss": 4.9351,
      "mean_token_accuracy": 0.18839796632528305,
      "num_tokens": 51241368.0,
      "step": 23385
    },
    {
      "entropy": 5.6091389656066895,
      "epoch": 2.3213576816196904,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044634305731029803,
      "loss": 5.1317,
      "mean_token_accuracy": 0.17857041358947753,
      "num_tokens": 51252784.0,
      "step": 23390
    },
    {
      "entropy": 5.6178709983825685,
      "epoch": 2.321853910281858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044632009076165096,
      "loss": 5.0101,
      "mean_token_accuracy": 0.182150898873806,
      "num_tokens": 51264246.0,
      "step": 23395
    },
    {
      "entropy": 5.613768625259399,
      "epoch": 2.3223501389440253,
      "grad_norm": 1.125,
      "learning_rate": 0.000446297119964635,
      "loss": 4.9788,
      "mean_token_accuracy": 0.18998976349830626,
      "num_tokens": 51274575.0,
      "step": 23400
    },
    {
      "entropy": 5.6574615955352785,
      "epoch": 2.322846367606193,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004462741449198194,
      "loss": 5.003,
      "mean_token_accuracy": 0.17834293693304062,
      "num_tokens": 51285384.0,
      "step": 23405
    },
    {
      "entropy": 5.700213384628296,
      "epoch": 2.3233425962683603,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044625116562777413,
      "loss": 5.0825,
      "mean_token_accuracy": 0.18402335792779922,
      "num_tokens": 51297983.0,
      "step": 23410
    },
    {
      "entropy": 5.549267721176148,
      "epoch": 2.323838824930528,
      "grad_norm": 1.0,
      "learning_rate": 0.0004462281820890689,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18745585680007934,
      "num_tokens": 51308795.0,
      "step": 23415
    },
    {
      "entropy": 5.569281387329101,
      "epoch": 2.3243350535926957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044620519430427376,
      "loss": 5.0045,
      "mean_token_accuracy": 0.1849617376923561,
      "num_tokens": 51318597.0,
      "step": 23420
    },
    {
      "entropy": 5.553705883026123,
      "epoch": 2.324831282254863,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004461822022739588,
      "loss": 4.9487,
      "mean_token_accuracy": 0.18325890004634857,
      "num_tokens": 51329238.0,
      "step": 23425
    },
    {
      "entropy": 5.66072359085083,
      "epoch": 2.3253275109170306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000446159205998694,
      "loss": 5.0667,
      "mean_token_accuracy": 0.18175757080316543,
      "num_tokens": 51340295.0,
      "step": 23430
    },
    {
      "entropy": 5.622609806060791,
      "epoch": 2.325823739579198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004461362054790497,
      "loss": 5.028,
      "mean_token_accuracy": 0.1823334887623787,
      "num_tokens": 51350702.0,
      "step": 23435
    },
    {
      "entropy": 5.627181053161621,
      "epoch": 2.3263199682413656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044611320071559637,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18232324868440627,
      "num_tokens": 51359794.0,
      "step": 23440
    },
    {
      "entropy": 5.60952730178833,
      "epoch": 2.326816196903533,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044609019170890424,
      "loss": 5.0165,
      "mean_token_accuracy": 0.17850434631109238,
      "num_tokens": 51368932.0,
      "step": 23445
    },
    {
      "entropy": 5.65521469116211,
      "epoch": 2.3273124255657005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044606717845954415,
      "loss": 5.0788,
      "mean_token_accuracy": 0.1799841672182083,
      "num_tokens": 51381111.0,
      "step": 23450
    },
    {
      "entropy": 5.589181709289551,
      "epoch": 2.3278086542278684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044604416096808656,
      "loss": 4.9757,
      "mean_token_accuracy": 0.19691843986511232,
      "num_tokens": 51392976.0,
      "step": 23455
    },
    {
      "entropy": 5.6526398181915285,
      "epoch": 2.328304882890036,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044602113923510235,
      "loss": 5.014,
      "mean_token_accuracy": 0.18076668232679366,
      "num_tokens": 51403586.0,
      "step": 23460
    },
    {
      "entropy": 5.65369348526001,
      "epoch": 2.3288011115522034,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044599811326116246,
      "loss": 5.0727,
      "mean_token_accuracy": 0.1774643123149872,
      "num_tokens": 51414035.0,
      "step": 23465
    },
    {
      "entropy": 5.635379028320313,
      "epoch": 2.329297340214371,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044597508304683784,
      "loss": 5.0707,
      "mean_token_accuracy": 0.1768762782216072,
      "num_tokens": 51424111.0,
      "step": 23470
    },
    {
      "entropy": 5.60809006690979,
      "epoch": 2.3297935688765383,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004459520485926995,
      "loss": 4.9795,
      "mean_token_accuracy": 0.18724362403154374,
      "num_tokens": 51434566.0,
      "step": 23475
    },
    {
      "entropy": 5.662055635452271,
      "epoch": 2.330289797538706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004459290098993188,
      "loss": 5.0406,
      "mean_token_accuracy": 0.18471300303936006,
      "num_tokens": 51445568.0,
      "step": 23480
    },
    {
      "entropy": 5.668969535827637,
      "epoch": 2.3307860262008733,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044590596696726686,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18569877445697786,
      "num_tokens": 51456482.0,
      "step": 23485
    },
    {
      "entropy": 5.651195383071899,
      "epoch": 2.3312822548630407,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004458829197971153,
      "loss": 5.113,
      "mean_token_accuracy": 0.17083147466182708,
      "num_tokens": 51467809.0,
      "step": 23490
    },
    {
      "entropy": 5.650652599334717,
      "epoch": 2.3317784835252082,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044585986838943556,
      "loss": 5.0121,
      "mean_token_accuracy": 0.17862578630447387,
      "num_tokens": 51477927.0,
      "step": 23495
    },
    {
      "entropy": 5.647022294998169,
      "epoch": 2.3322747121873757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044583681274479927,
      "loss": 5.0775,
      "mean_token_accuracy": 0.18149694353342055,
      "num_tokens": 51488432.0,
      "step": 23500
    },
    {
      "entropy": 5.6672656536102295,
      "epoch": 2.3327709408495436,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004458137528637781,
      "loss": 5.0445,
      "mean_token_accuracy": 0.18019863963127136,
      "num_tokens": 51499149.0,
      "step": 23505
    },
    {
      "entropy": 5.704043245315551,
      "epoch": 2.333267169511711,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044579068874694386,
      "loss": 5.0718,
      "mean_token_accuracy": 0.1782624065876007,
      "num_tokens": 51510672.0,
      "step": 23510
    },
    {
      "entropy": 5.672218084335327,
      "epoch": 2.3337633981738786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004457676203948687,
      "loss": 5.0678,
      "mean_token_accuracy": 0.17459011673927308,
      "num_tokens": 51521855.0,
      "step": 23515
    },
    {
      "entropy": 5.683179187774658,
      "epoch": 2.334259626836046,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004457445478081244,
      "loss": 5.0493,
      "mean_token_accuracy": 0.1815553456544876,
      "num_tokens": 51531970.0,
      "step": 23520
    },
    {
      "entropy": 5.662017679214477,
      "epoch": 2.3347558554982135,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044572147098728335,
      "loss": 5.1123,
      "mean_token_accuracy": 0.17739087492227554,
      "num_tokens": 51543110.0,
      "step": 23525
    },
    {
      "entropy": 5.626097106933594,
      "epoch": 2.335252084160381,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044569838993291757,
      "loss": 5.0467,
      "mean_token_accuracy": 0.17825039476156235,
      "num_tokens": 51552991.0,
      "step": 23530
    },
    {
      "entropy": 5.649916172027588,
      "epoch": 2.3357483128225485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044567530464559963,
      "loss": 5.0791,
      "mean_token_accuracy": 0.18475452214479446,
      "num_tokens": 51564472.0,
      "step": 23535
    },
    {
      "entropy": 5.601472949981689,
      "epoch": 2.3362445414847164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004456522151259019,
      "loss": 5.0916,
      "mean_token_accuracy": 0.1814884588122368,
      "num_tokens": 51576051.0,
      "step": 23540
    },
    {
      "entropy": 5.7131517887115475,
      "epoch": 2.336740770146884,
      "grad_norm": 0.953125,
      "learning_rate": 0.000445629121374397,
      "loss": 5.0698,
      "mean_token_accuracy": 0.17629393190145493,
      "num_tokens": 51586658.0,
      "step": 23545
    },
    {
      "entropy": 5.641535472869873,
      "epoch": 2.3372369988090513,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044560602339165746,
      "loss": 5.0387,
      "mean_token_accuracy": 0.17588323652744292,
      "num_tokens": 51596614.0,
      "step": 23550
    },
    {
      "entropy": 5.622547578811646,
      "epoch": 2.337733227471219,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004455829211782562,
      "loss": 5.0693,
      "mean_token_accuracy": 0.1811390310525894,
      "num_tokens": 51607102.0,
      "step": 23555
    },
    {
      "entropy": 5.737768030166626,
      "epoch": 2.3382294561333863,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004455598147347661,
      "loss": 5.1205,
      "mean_token_accuracy": 0.17105642557144166,
      "num_tokens": 51619079.0,
      "step": 23560
    },
    {
      "entropy": 5.62809739112854,
      "epoch": 2.3387256847955538,
      "grad_norm": 1.0,
      "learning_rate": 0.0004455367040617601,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18722463846206666,
      "num_tokens": 51630273.0,
      "step": 23565
    },
    {
      "entropy": 5.628103160858155,
      "epoch": 2.3392219134577212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044551358915981135,
      "loss": 5.0395,
      "mean_token_accuracy": 0.1833304926753044,
      "num_tokens": 51640480.0,
      "step": 23570
    },
    {
      "entropy": 5.6585184097290036,
      "epoch": 2.3397181421198887,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004454904700294929,
      "loss": 5.0843,
      "mean_token_accuracy": 0.1783070296049118,
      "num_tokens": 51651831.0,
      "step": 23575
    },
    {
      "entropy": 5.643916034698487,
      "epoch": 2.340214370782056,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004454673466713783,
      "loss": 5.0765,
      "mean_token_accuracy": 0.1808541461825371,
      "num_tokens": 51664061.0,
      "step": 23580
    },
    {
      "entropy": 5.599853277206421,
      "epoch": 2.3407105994442237,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044544421908604073,
      "loss": 5.0371,
      "mean_token_accuracy": 0.18233652263879777,
      "num_tokens": 51675081.0,
      "step": 23585
    },
    {
      "entropy": 5.592111349105835,
      "epoch": 2.3412068281063916,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044542108727405386,
      "loss": 4.9513,
      "mean_token_accuracy": 0.18998512625694275,
      "num_tokens": 51684477.0,
      "step": 23590
    },
    {
      "entropy": 5.729834651947021,
      "epoch": 2.341703056768559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044539795123599117,
      "loss": 5.0978,
      "mean_token_accuracy": 0.18341927975416183,
      "num_tokens": 51694420.0,
      "step": 23595
    },
    {
      "entropy": 5.5231029987335205,
      "epoch": 2.3421992854307265,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044537481097242654,
      "loss": 4.9244,
      "mean_token_accuracy": 0.19367887377738952,
      "num_tokens": 51704635.0,
      "step": 23600
    },
    {
      "entropy": 5.658256340026855,
      "epoch": 2.342695514092894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044535166648393377,
      "loss": 5.1306,
      "mean_token_accuracy": 0.17694143652915956,
      "num_tokens": 51716149.0,
      "step": 23605
    },
    {
      "entropy": 5.569886636734009,
      "epoch": 2.3431917427550615,
      "grad_norm": 1.0,
      "learning_rate": 0.00044532851777108663,
      "loss": 4.9866,
      "mean_token_accuracy": 0.17891329973936082,
      "num_tokens": 51726861.0,
      "step": 23610
    },
    {
      "entropy": 5.695457887649536,
      "epoch": 2.343687971417229,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004453053648344593,
      "loss": 5.2327,
      "mean_token_accuracy": 0.17405472993850707,
      "num_tokens": 51738111.0,
      "step": 23615
    },
    {
      "entropy": 5.620181035995484,
      "epoch": 2.3441842000793964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044528220767462595,
      "loss": 5.0553,
      "mean_token_accuracy": 0.1799207091331482,
      "num_tokens": 51749004.0,
      "step": 23620
    },
    {
      "entropy": 5.63593864440918,
      "epoch": 2.3446804287415643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004452590462921607,
      "loss": 5.0372,
      "mean_token_accuracy": 0.17395465075969696,
      "num_tokens": 51759557.0,
      "step": 23625
    },
    {
      "entropy": 5.695861721038819,
      "epoch": 2.345176657403732,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044523588068763804,
      "loss": 5.1042,
      "mean_token_accuracy": 0.17778241038322448,
      "num_tokens": 51769921.0,
      "step": 23630
    },
    {
      "entropy": 5.596221780776977,
      "epoch": 2.3456728860658993,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004452127108616324,
      "loss": 4.9566,
      "mean_token_accuracy": 0.18623185008764268,
      "num_tokens": 51780158.0,
      "step": 23635
    },
    {
      "entropy": 5.662840747833252,
      "epoch": 2.3461691147280668,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044518953681471824,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1832994446158409,
      "num_tokens": 51790502.0,
      "step": 23640
    },
    {
      "entropy": 5.667146968841553,
      "epoch": 2.3466653433902342,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004451663585474703,
      "loss": 5.0927,
      "mean_token_accuracy": 0.17286817282438277,
      "num_tokens": 51801420.0,
      "step": 23645
    },
    {
      "entropy": 5.643948888778686,
      "epoch": 2.3471615720524017,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044514317606046336,
      "loss": 5.0763,
      "mean_token_accuracy": 0.18606505990028382,
      "num_tokens": 51812399.0,
      "step": 23650
    },
    {
      "entropy": 5.6943072319030765,
      "epoch": 2.347657800714569,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004451199893542723,
      "loss": 5.14,
      "mean_token_accuracy": 0.17186671793460845,
      "num_tokens": 51824268.0,
      "step": 23655
    },
    {
      "entropy": 5.600913715362549,
      "epoch": 2.3481540293767367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004450967984294721,
      "loss": 5.0506,
      "mean_token_accuracy": 0.17744104117155074,
      "num_tokens": 51835842.0,
      "step": 23660
    },
    {
      "entropy": 5.628518915176391,
      "epoch": 2.348650258038904,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004450736032866377,
      "loss": 4.9919,
      "mean_token_accuracy": 0.1836063399910927,
      "num_tokens": 51847501.0,
      "step": 23665
    },
    {
      "entropy": 5.633373212814331,
      "epoch": 2.3491464867010716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044505040392634447,
      "loss": 5.0396,
      "mean_token_accuracy": 0.1902475908398628,
      "num_tokens": 51858056.0,
      "step": 23670
    },
    {
      "entropy": 5.662948703765869,
      "epoch": 2.3496427153632395,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004450272003491677,
      "loss": 5.0065,
      "mean_token_accuracy": 0.189346481859684,
      "num_tokens": 51869248.0,
      "step": 23675
    },
    {
      "entropy": 5.629848861694336,
      "epoch": 2.350138944025407,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044500399255568274,
      "loss": 5.0994,
      "mean_token_accuracy": 0.17558325082063675,
      "num_tokens": 51881496.0,
      "step": 23680
    },
    {
      "entropy": 5.621562480926514,
      "epoch": 2.3506351726875745,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044498078054646507,
      "loss": 5.0546,
      "mean_token_accuracy": 0.17686007916927338,
      "num_tokens": 51891851.0,
      "step": 23685
    },
    {
      "entropy": 5.689729881286621,
      "epoch": 2.351131401349742,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044495756432209025,
      "loss": 5.1303,
      "mean_token_accuracy": 0.1773071214556694,
      "num_tokens": 51903581.0,
      "step": 23690
    },
    {
      "entropy": 5.608774185180664,
      "epoch": 2.3516276300119094,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044493434388313423,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18456615954637529,
      "num_tokens": 51913650.0,
      "step": 23695
    },
    {
      "entropy": 5.568599891662598,
      "epoch": 2.352123858674077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044491111923017254,
      "loss": 4.9842,
      "mean_token_accuracy": 0.18863641619682311,
      "num_tokens": 51924704.0,
      "step": 23700
    },
    {
      "entropy": 5.721727991104126,
      "epoch": 2.3526200873362444,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044488789036378117,
      "loss": 5.1157,
      "mean_token_accuracy": 0.180325248837471,
      "num_tokens": 51937576.0,
      "step": 23705
    },
    {
      "entropy": 5.644149732589722,
      "epoch": 2.3531163159984123,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004448646572845363,
      "loss": 4.9563,
      "mean_token_accuracy": 0.1822962537407875,
      "num_tokens": 51947359.0,
      "step": 23710
    },
    {
      "entropy": 5.675649356842041,
      "epoch": 2.35361254466058,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004448414199930139,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17245987206697463,
      "num_tokens": 51958016.0,
      "step": 23715
    },
    {
      "entropy": 5.688798666000366,
      "epoch": 2.3541087733227473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044481817848979026,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1772113636136055,
      "num_tokens": 51968770.0,
      "step": 23720
    },
    {
      "entropy": 5.646723842620849,
      "epoch": 2.3546050019849147,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004447949327754418,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1922467365860939,
      "num_tokens": 51980052.0,
      "step": 23725
    },
    {
      "entropy": 5.676547193527222,
      "epoch": 2.355101230647082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004447716828505447,
      "loss": 5.1319,
      "mean_token_accuracy": 0.17731730341911317,
      "num_tokens": 51992080.0,
      "step": 23730
    },
    {
      "entropy": 5.621990203857422,
      "epoch": 2.3555974593092497,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004447484287156758,
      "loss": 5.0589,
      "mean_token_accuracy": 0.1865004450082779,
      "num_tokens": 52004313.0,
      "step": 23735
    },
    {
      "entropy": 5.704454231262207,
      "epoch": 2.356093687971417,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004447251703714117,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1788470059633255,
      "num_tokens": 52015207.0,
      "step": 23740
    },
    {
      "entropy": 5.684885787963867,
      "epoch": 2.3565899166335846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000444701907818329,
      "loss": 5.0306,
      "mean_token_accuracy": 0.18112519830465318,
      "num_tokens": 52026565.0,
      "step": 23745
    },
    {
      "entropy": 5.6554468154907225,
      "epoch": 2.357086145295752,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004446786410570048,
      "loss": 5.0665,
      "mean_token_accuracy": 0.17708853483200074,
      "num_tokens": 52037654.0,
      "step": 23750
    },
    {
      "entropy": 5.621186542510986,
      "epoch": 2.3575823739579196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044465537008801575,
      "loss": 4.9824,
      "mean_token_accuracy": 0.18335627615451813,
      "num_tokens": 52049772.0,
      "step": 23755
    },
    {
      "entropy": 5.661295795440674,
      "epoch": 2.3580786026200875,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004446320949119392,
      "loss": 5.111,
      "mean_token_accuracy": 0.1751111775636673,
      "num_tokens": 52061363.0,
      "step": 23760
    },
    {
      "entropy": 5.6570087432861325,
      "epoch": 2.358574831282255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044460881552935215,
      "loss": 4.9815,
      "mean_token_accuracy": 0.18458664566278457,
      "num_tokens": 52073153.0,
      "step": 23765
    },
    {
      "entropy": 5.583527660369873,
      "epoch": 2.3590710599444225,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044458553194083204,
      "loss": 5.0078,
      "mean_token_accuracy": 0.1868895873427391,
      "num_tokens": 52083369.0,
      "step": 23770
    },
    {
      "entropy": 5.537939548492432,
      "epoch": 2.35956728860659,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004445622441469562,
      "loss": 4.9796,
      "mean_token_accuracy": 0.1886815086007118,
      "num_tokens": 52093701.0,
      "step": 23775
    },
    {
      "entropy": 5.623079776763916,
      "epoch": 2.3600635172687574,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044453895214830195,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18425245732069015,
      "num_tokens": 52105863.0,
      "step": 23780
    },
    {
      "entropy": 5.6159303188323975,
      "epoch": 2.360559745930925,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004445156559454471,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18544960916042327,
      "num_tokens": 52116476.0,
      "step": 23785
    },
    {
      "entropy": 5.7118010997772215,
      "epoch": 2.3610559745930924,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044449235553896914,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17367386370897292,
      "num_tokens": 52127262.0,
      "step": 23790
    },
    {
      "entropy": 5.711814451217651,
      "epoch": 2.36155220325526,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004444690509294461,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17675951421260833,
      "num_tokens": 52138575.0,
      "step": 23795
    },
    {
      "entropy": 5.653647136688233,
      "epoch": 2.3620484319174277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004444457421174558,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18533073663711547,
      "num_tokens": 52149620.0,
      "step": 23800
    },
    {
      "entropy": 5.688618421554565,
      "epoch": 2.3625446605795952,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004444224291035761,
      "loss": 5.0812,
      "mean_token_accuracy": 0.17791416198015214,
      "num_tokens": 52161929.0,
      "step": 23805
    },
    {
      "entropy": 5.56142897605896,
      "epoch": 2.3630408892417627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044439911188838533,
      "loss": 4.9877,
      "mean_token_accuracy": 0.18844375759363174,
      "num_tokens": 52172283.0,
      "step": 23810
    },
    {
      "entropy": 5.627657890319824,
      "epoch": 2.36353711790393,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004443757904724615,
      "loss": 5.0734,
      "mean_token_accuracy": 0.1819676473736763,
      "num_tokens": 52184130.0,
      "step": 23815
    },
    {
      "entropy": 5.640659141540527,
      "epoch": 2.3640333465660976,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00044435246485638317,
      "loss": 5.087,
      "mean_token_accuracy": 0.180560339987278,
      "num_tokens": 52195358.0,
      "step": 23820
    },
    {
      "entropy": 5.629545068740844,
      "epoch": 2.364529575228265,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004443291350407285,
      "loss": 4.9917,
      "mean_token_accuracy": 0.18432238548994065,
      "num_tokens": 52207070.0,
      "step": 23825
    },
    {
      "entropy": 5.618776845932007,
      "epoch": 2.3650258038904326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004443058010260762,
      "loss": 5.1142,
      "mean_token_accuracy": 0.17500463575124742,
      "num_tokens": 52217658.0,
      "step": 23830
    },
    {
      "entropy": 5.5908438205719,
      "epoch": 2.3655220325526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044428246281300493,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18428633958101273,
      "num_tokens": 52229350.0,
      "step": 23835
    },
    {
      "entropy": 5.7200891971588135,
      "epoch": 2.3660182612147675,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044425912040209326,
      "loss": 5.1542,
      "mean_token_accuracy": 0.17437507063150406,
      "num_tokens": 52241260.0,
      "step": 23840
    },
    {
      "entropy": 5.732102298736573,
      "epoch": 2.3665144898769355,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044423577379392023,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18032486438751222,
      "num_tokens": 52253049.0,
      "step": 23845
    },
    {
      "entropy": 5.634666872024536,
      "epoch": 2.367010718539103,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004442124229890646,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1834353968501091,
      "num_tokens": 52263405.0,
      "step": 23850
    },
    {
      "entropy": 5.628441905975341,
      "epoch": 2.3675069472012704,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00044418906798810554,
      "loss": 5.0566,
      "mean_token_accuracy": 0.18022844791412354,
      "num_tokens": 52273075.0,
      "step": 23855
    },
    {
      "entropy": 5.750368070602417,
      "epoch": 2.368003175863438,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00044416570879162213,
      "loss": 5.232,
      "mean_token_accuracy": 0.17635360807180406,
      "num_tokens": 52284275.0,
      "step": 23860
    },
    {
      "entropy": 5.626895236968994,
      "epoch": 2.3684994045256054,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441423454001936,
      "loss": 5.0528,
      "mean_token_accuracy": 0.17310045063495635,
      "num_tokens": 52295084.0,
      "step": 23865
    },
    {
      "entropy": 5.645531225204468,
      "epoch": 2.368995633187773,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044411897781439946,
      "loss": 5.0305,
      "mean_token_accuracy": 0.18111558854579926,
      "num_tokens": 52304172.0,
      "step": 23870
    },
    {
      "entropy": 5.576230859756469,
      "epoch": 2.3694918618499403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004440956060348191,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18759787678718567,
      "num_tokens": 52315370.0,
      "step": 23875
    },
    {
      "entropy": 5.617411041259766,
      "epoch": 2.369988090512108,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004440722300620321,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18349443823099137,
      "num_tokens": 52326517.0,
      "step": 23880
    },
    {
      "entropy": 5.66649603843689,
      "epoch": 2.3704843191742757,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044404884989661804,
      "loss": 5.09,
      "mean_token_accuracy": 0.18432708233594894,
      "num_tokens": 52337110.0,
      "step": 23885
    },
    {
      "entropy": 5.630427837371826,
      "epoch": 2.370980547836443,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004440254655391568,
      "loss": 5.0107,
      "mean_token_accuracy": 0.1851227328181267,
      "num_tokens": 52347509.0,
      "step": 23890
    },
    {
      "entropy": 5.758263826370239,
      "epoch": 2.3714767764986107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004440020769902282,
      "loss": 5.1345,
      "mean_token_accuracy": 0.17552033960819244,
      "num_tokens": 52357920.0,
      "step": 23895
    },
    {
      "entropy": 5.6873376846313475,
      "epoch": 2.371973005160778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044397868425041234,
      "loss": 5.1129,
      "mean_token_accuracy": 0.17488602846860885,
      "num_tokens": 52368138.0,
      "step": 23900
    },
    {
      "entropy": 5.694574022293091,
      "epoch": 2.3724692338229456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004439552873202891,
      "loss": 5.0741,
      "mean_token_accuracy": 0.18021257668733598,
      "num_tokens": 52380503.0,
      "step": 23905
    },
    {
      "entropy": 5.63793044090271,
      "epoch": 2.372965462485113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044393188620043893,
      "loss": 5.0517,
      "mean_token_accuracy": 0.18047939538955687,
      "num_tokens": 52390850.0,
      "step": 23910
    },
    {
      "entropy": 5.580529022216797,
      "epoch": 2.3734616911472806,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004439084808914419,
      "loss": 5.0027,
      "mean_token_accuracy": 0.1850647285580635,
      "num_tokens": 52400829.0,
      "step": 23915
    },
    {
      "entropy": 5.67503981590271,
      "epoch": 2.373957919809448,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044388507139387864,
      "loss": 5.0819,
      "mean_token_accuracy": 0.18077048808336257,
      "num_tokens": 52412892.0,
      "step": 23920
    },
    {
      "entropy": 5.618984603881836,
      "epoch": 2.3744541484716155,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004438616577083293,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18640097230672836,
      "num_tokens": 52423662.0,
      "step": 23925
    },
    {
      "entropy": 5.621635007858276,
      "epoch": 2.3749503771337834,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000443838239835375,
      "loss": 5.0491,
      "mean_token_accuracy": 0.18637704253196716,
      "num_tokens": 52433541.0,
      "step": 23930
    },
    {
      "entropy": 5.665837049484253,
      "epoch": 2.375446605795951,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004438148177755959,
      "loss": 5.1063,
      "mean_token_accuracy": 0.17957591861486435,
      "num_tokens": 52445504.0,
      "step": 23935
    },
    {
      "entropy": 5.6917163848876955,
      "epoch": 2.3759428344581184,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044379139152957327,
      "loss": 5.0799,
      "mean_token_accuracy": 0.17381191551685332,
      "num_tokens": 52456116.0,
      "step": 23940
    },
    {
      "entropy": 5.62945237159729,
      "epoch": 2.376439063120286,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044376796109788777,
      "loss": 4.9503,
      "mean_token_accuracy": 0.18958612829446791,
      "num_tokens": 52465462.0,
      "step": 23945
    },
    {
      "entropy": 5.6204001903533936,
      "epoch": 2.3769352917824533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044374452648112046,
      "loss": 5.0968,
      "mean_token_accuracy": 0.17679387480020523,
      "num_tokens": 52477213.0,
      "step": 23950
    },
    {
      "entropy": 5.669510173797607,
      "epoch": 2.377431520444621,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044372108767985253,
      "loss": 5.0986,
      "mean_token_accuracy": 0.17699605524539946,
      "num_tokens": 52488095.0,
      "step": 23955
    },
    {
      "entropy": 5.702954626083374,
      "epoch": 2.3779277491067883,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004436976446946652,
      "loss": 5.153,
      "mean_token_accuracy": 0.1727050945162773,
      "num_tokens": 52499078.0,
      "step": 23960
    },
    {
      "entropy": 5.6857081890106205,
      "epoch": 2.3784239777689558,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044367419752613983,
      "loss": 5.0352,
      "mean_token_accuracy": 0.184202179312706,
      "num_tokens": 52510829.0,
      "step": 23965
    },
    {
      "entropy": 5.7086358070373535,
      "epoch": 2.3789202064311237,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004436507461748577,
      "loss": 5.0349,
      "mean_token_accuracy": 0.17900011241436004,
      "num_tokens": 52521466.0,
      "step": 23970
    },
    {
      "entropy": 5.6412547588348385,
      "epoch": 2.379416435093291,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004436272906414007,
      "loss": 4.9774,
      "mean_token_accuracy": 0.18889648765325545,
      "num_tokens": 52532617.0,
      "step": 23975
    },
    {
      "entropy": 5.533463001251221,
      "epoch": 2.3799126637554586,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004436038309263501,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18432613462209702,
      "num_tokens": 52543826.0,
      "step": 23980
    },
    {
      "entropy": 5.605553817749024,
      "epoch": 2.380408892417626,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044358036703028773,
      "loss": 5.0611,
      "mean_token_accuracy": 0.17872810065746308,
      "num_tokens": 52556752.0,
      "step": 23985
    },
    {
      "entropy": 5.662611961364746,
      "epoch": 2.3809051210797936,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044355689895379563,
      "loss": 5.129,
      "mean_token_accuracy": 0.17374031543731688,
      "num_tokens": 52567268.0,
      "step": 23990
    },
    {
      "entropy": 5.677391529083252,
      "epoch": 2.381401349741961,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044353342669745566,
      "loss": 5.0573,
      "mean_token_accuracy": 0.18625050485134126,
      "num_tokens": 52576894.0,
      "step": 23995
    },
    {
      "entropy": 5.6303037166595455,
      "epoch": 2.3818975784041285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044350995026184976,
      "loss": 5.0397,
      "mean_token_accuracy": 0.18005985468626023,
      "num_tokens": 52587434.0,
      "step": 24000
    },
    {
      "epoch": 2.3818975784041285,
      "eval_entropy": 5.443730418771031,
      "eval_loss": 5.159703731536865,
      "eval_mean_token_accuracy": 0.18192305230179545,
      "eval_num_tokens": 52587434.0,
      "eval_runtime": 20.4157,
      "eval_samples_per_second": 1649.17,
      "eval_steps_per_second": 206.165,
      "step": 24000
    },
    {
      "entropy": 5.708327865600586,
      "epoch": 2.382393807066296,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044348646964756036,
      "loss": 5.1166,
      "mean_token_accuracy": 0.18086824417114258,
      "num_tokens": 52599455.0,
      "step": 24005
    },
    {
      "entropy": 5.692405366897583,
      "epoch": 2.3828900357284635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004434629848551694,
      "loss": 5.0891,
      "mean_token_accuracy": 0.17840274572372436,
      "num_tokens": 52611020.0,
      "step": 24010
    },
    {
      "entropy": 5.60822377204895,
      "epoch": 2.3833862643906314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004434394958852595,
      "loss": 5.036,
      "mean_token_accuracy": 0.1809602916240692,
      "num_tokens": 52621743.0,
      "step": 24015
    },
    {
      "entropy": 5.645718574523926,
      "epoch": 2.383882493052799,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044341600273841304,
      "loss": 5.0983,
      "mean_token_accuracy": 0.18243863433599472,
      "num_tokens": 52633437.0,
      "step": 24020
    },
    {
      "entropy": 5.623079395294189,
      "epoch": 2.3843787217149663,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044339250541521264,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18354335874319078,
      "num_tokens": 52643849.0,
      "step": 24025
    },
    {
      "entropy": 5.789769506454467,
      "epoch": 2.384874950377134,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044336900391624096,
      "loss": 5.1809,
      "mean_token_accuracy": 0.17380232661962508,
      "num_tokens": 52656417.0,
      "step": 24030
    },
    {
      "entropy": 5.605811834335327,
      "epoch": 2.3853711790393013,
      "grad_norm": 1.125,
      "learning_rate": 0.0004433454982420808,
      "loss": 4.9481,
      "mean_token_accuracy": 0.1875785395503044,
      "num_tokens": 52668491.0,
      "step": 24035
    },
    {
      "entropy": 5.599684619903565,
      "epoch": 2.3858674077014688,
      "grad_norm": 1.15625,
      "learning_rate": 0.000443321988393315,
      "loss": 5.099,
      "mean_token_accuracy": 0.1794159531593323,
      "num_tokens": 52679532.0,
      "step": 24040
    },
    {
      "entropy": 5.619836139678955,
      "epoch": 2.3863636363636362,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004432984743705265,
      "loss": 4.9853,
      "mean_token_accuracy": 0.17723403573036195,
      "num_tokens": 52689850.0,
      "step": 24045
    },
    {
      "entropy": 5.566048812866211,
      "epoch": 2.3868598650258037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044327495617429866,
      "loss": 4.9675,
      "mean_token_accuracy": 0.1856963098049164,
      "num_tokens": 52699529.0,
      "step": 24050
    },
    {
      "entropy": 5.572166156768799,
      "epoch": 2.3873560936879716,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004432514338052144,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1902967244386673,
      "num_tokens": 52711449.0,
      "step": 24055
    },
    {
      "entropy": 5.726732444763184,
      "epoch": 2.387852322350139,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004432279072638571,
      "loss": 5.1247,
      "mean_token_accuracy": 0.16821037381887435,
      "num_tokens": 52721528.0,
      "step": 24060
    },
    {
      "entropy": 5.667291307449341,
      "epoch": 2.3883485510123066,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044320437655081033,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18384081572294236,
      "num_tokens": 52731367.0,
      "step": 24065
    },
    {
      "entropy": 5.573930263519287,
      "epoch": 2.388844779674474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044318084166665737,
      "loss": 4.9329,
      "mean_token_accuracy": 0.18844350576400756,
      "num_tokens": 52740822.0,
      "step": 24070
    },
    {
      "entropy": 5.534777593612671,
      "epoch": 2.3893410083366415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000443157302611982,
      "loss": 5.0231,
      "mean_token_accuracy": 0.17690811455249786,
      "num_tokens": 52751609.0,
      "step": 24075
    },
    {
      "entropy": 5.6604328632354735,
      "epoch": 2.389837236998809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004431337593873678,
      "loss": 5.0759,
      "mean_token_accuracy": 0.1796773836016655,
      "num_tokens": 52762914.0,
      "step": 24080
    },
    {
      "entropy": 5.757787990570068,
      "epoch": 2.3903334656609765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004431102119933987,
      "loss": 5.1127,
      "mean_token_accuracy": 0.17688402086496352,
      "num_tokens": 52773894.0,
      "step": 24085
    },
    {
      "entropy": 5.623678159713745,
      "epoch": 2.390829694323144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004430866604306586,
      "loss": 5.0404,
      "mean_token_accuracy": 0.18464304953813554,
      "num_tokens": 52784606.0,
      "step": 24090
    },
    {
      "entropy": 5.625314044952392,
      "epoch": 2.3913259229853114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044306310469973147,
      "loss": 5.0174,
      "mean_token_accuracy": 0.17817746102809906,
      "num_tokens": 52795275.0,
      "step": 24095
    },
    {
      "entropy": 5.690853595733643,
      "epoch": 2.391822151647479,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044303954480120155,
      "loss": 5.0971,
      "mean_token_accuracy": 0.18176882714033127,
      "num_tokens": 52805058.0,
      "step": 24100
    },
    {
      "entropy": 5.652186107635498,
      "epoch": 2.392318380309647,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044301598073565295,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1771961510181427,
      "num_tokens": 52815486.0,
      "step": 24105
    },
    {
      "entropy": 5.640275239944458,
      "epoch": 2.3928146089718143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004429924125036702,
      "loss": 5.0462,
      "mean_token_accuracy": 0.1822650298476219,
      "num_tokens": 52827039.0,
      "step": 24110
    },
    {
      "entropy": 5.626214790344238,
      "epoch": 2.393310837633982,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004429688401058375,
      "loss": 5.0592,
      "mean_token_accuracy": 0.18034036010503768,
      "num_tokens": 52837712.0,
      "step": 24115
    },
    {
      "entropy": 5.626235628128052,
      "epoch": 2.3938070662961493,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004429452635427394,
      "loss": 4.943,
      "mean_token_accuracy": 0.1870670273900032,
      "num_tokens": 52846321.0,
      "step": 24120
    },
    {
      "entropy": 5.6432067394256595,
      "epoch": 2.3943032949583167,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044292168281496086,
      "loss": 5.0025,
      "mean_token_accuracy": 0.18454407751560212,
      "num_tokens": 52856342.0,
      "step": 24125
    },
    {
      "entropy": 5.642481708526612,
      "epoch": 2.394799523620484,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004428980979230863,
      "loss": 5.1292,
      "mean_token_accuracy": 0.17938793450593948,
      "num_tokens": 52866602.0,
      "step": 24130
    },
    {
      "entropy": 5.585146760940551,
      "epoch": 2.3952957522826517,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044287450886770075,
      "loss": 4.9446,
      "mean_token_accuracy": 0.18915708214044571,
      "num_tokens": 52877066.0,
      "step": 24135
    },
    {
      "entropy": 5.592752742767334,
      "epoch": 2.3957919809448196,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004428509156493891,
      "loss": 4.9738,
      "mean_token_accuracy": 0.19005665183067322,
      "num_tokens": 52887907.0,
      "step": 24140
    },
    {
      "entropy": 5.712968826293945,
      "epoch": 2.396288209606987,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004428273182687364,
      "loss": 5.1249,
      "mean_token_accuracy": 0.1778334692120552,
      "num_tokens": 52899308.0,
      "step": 24145
    },
    {
      "entropy": 5.616712856292724,
      "epoch": 2.3967844382691545,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428037167263279,
      "loss": 4.9553,
      "mean_token_accuracy": 0.18336808532476426,
      "num_tokens": 52909127.0,
      "step": 24150
    },
    {
      "entropy": 5.608267784118652,
      "epoch": 2.397280666931322,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044278011102274887,
      "loss": 4.9828,
      "mean_token_accuracy": 0.18786257803440093,
      "num_tokens": 52920503.0,
      "step": 24155
    },
    {
      "entropy": 5.636089944839478,
      "epoch": 2.3977768955934895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044275650115858445,
      "loss": 5.0189,
      "mean_token_accuracy": 0.17983858585357665,
      "num_tokens": 52931904.0,
      "step": 24160
    },
    {
      "entropy": 5.669622182846069,
      "epoch": 2.398273124255657,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004427328871344204,
      "loss": 5.1581,
      "mean_token_accuracy": 0.16919738054275513,
      "num_tokens": 52945101.0,
      "step": 24165
    },
    {
      "entropy": 5.631828784942627,
      "epoch": 2.3987693529178244,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004427092689508422,
      "loss": 5.0359,
      "mean_token_accuracy": 0.1810856744647026,
      "num_tokens": 52955999.0,
      "step": 24170
    },
    {
      "entropy": 5.62688684463501,
      "epoch": 2.399265581579992,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004426856466084354,
      "loss": 5.0582,
      "mean_token_accuracy": 0.17957428693771363,
      "num_tokens": 52967273.0,
      "step": 24175
    },
    {
      "entropy": 5.706053733825684,
      "epoch": 2.3997618102421594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000442662020107786,
      "loss": 5.1188,
      "mean_token_accuracy": 0.18159503638744354,
      "num_tokens": 52977415.0,
      "step": 24180
    },
    {
      "entropy": 5.631115436553955,
      "epoch": 2.400258038904327,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004426383894494797,
      "loss": 5.0126,
      "mean_token_accuracy": 0.18202046900987626,
      "num_tokens": 52990071.0,
      "step": 24185
    },
    {
      "entropy": 5.577979898452758,
      "epoch": 2.400754267566495,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004426147546341027,
      "loss": 4.9759,
      "mean_token_accuracy": 0.1836872696876526,
      "num_tokens": 52999874.0,
      "step": 24190
    },
    {
      "entropy": 5.6714677810668945,
      "epoch": 2.4012504962286623,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044259111566224084,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18011655658483505,
      "num_tokens": 53011349.0,
      "step": 24195
    },
    {
      "entropy": 5.614288711547852,
      "epoch": 2.4017467248908297,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044256747253448045,
      "loss": 4.9888,
      "mean_token_accuracy": 0.19050996005535126,
      "num_tokens": 53024256.0,
      "step": 24200
    },
    {
      "entropy": 5.603956317901611,
      "epoch": 2.402242953552997,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044254382525140784,
      "loss": 4.9605,
      "mean_token_accuracy": 0.19191956371068955,
      "num_tokens": 53035617.0,
      "step": 24205
    },
    {
      "entropy": 5.6317425727844235,
      "epoch": 2.4027391822151647,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044252017381360935,
      "loss": 4.992,
      "mean_token_accuracy": 0.19016102850437164,
      "num_tokens": 53046219.0,
      "step": 24210
    },
    {
      "entropy": 5.718701982498169,
      "epoch": 2.403235410877332,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004424965182216716,
      "loss": 5.11,
      "mean_token_accuracy": 0.1707630142569542,
      "num_tokens": 53057114.0,
      "step": 24215
    },
    {
      "entropy": 5.672120523452759,
      "epoch": 2.4037316395394996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044247285847618105,
      "loss": 5.056,
      "mean_token_accuracy": 0.18217649608850478,
      "num_tokens": 53067481.0,
      "step": 24220
    },
    {
      "entropy": 5.677069807052613,
      "epoch": 2.4042278682016676,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004424491945777245,
      "loss": 5.1322,
      "mean_token_accuracy": 0.18052324503660203,
      "num_tokens": 53078129.0,
      "step": 24225
    },
    {
      "entropy": 5.622431898117066,
      "epoch": 2.404724096863835,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044242552652688885,
      "loss": 4.9934,
      "mean_token_accuracy": 0.18576168566942214,
      "num_tokens": 53088919.0,
      "step": 24230
    },
    {
      "entropy": 5.662460613250732,
      "epoch": 2.4052203255260025,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004424018543242608,
      "loss": 5.0548,
      "mean_token_accuracy": 0.18582193553447723,
      "num_tokens": 53099716.0,
      "step": 24235
    },
    {
      "entropy": 5.589839029312134,
      "epoch": 2.40571655418817,
      "grad_norm": 1.125,
      "learning_rate": 0.0004423781779704275,
      "loss": 4.9668,
      "mean_token_accuracy": 0.19022763669490814,
      "num_tokens": 53108798.0,
      "step": 24240
    },
    {
      "entropy": 5.638107490539551,
      "epoch": 2.4062127828503375,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004423544974659761,
      "loss": 5.0741,
      "mean_token_accuracy": 0.18092873245477675,
      "num_tokens": 53120655.0,
      "step": 24245
    },
    {
      "entropy": 5.59249906539917,
      "epoch": 2.406709011512505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044233081281149373,
      "loss": 4.9352,
      "mean_token_accuracy": 0.1887353926897049,
      "num_tokens": 53131760.0,
      "step": 24250
    },
    {
      "entropy": 5.6478838443756105,
      "epoch": 2.4072052401746724,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044230712400756774,
      "loss": 5.01,
      "mean_token_accuracy": 0.18298080712556838,
      "num_tokens": 53142347.0,
      "step": 24255
    },
    {
      "entropy": 5.641078233718872,
      "epoch": 2.40770146883684,
      "grad_norm": 1.125,
      "learning_rate": 0.00044228343105478564,
      "loss": 5.0312,
      "mean_token_accuracy": 0.18337930887937545,
      "num_tokens": 53153678.0,
      "step": 24260
    },
    {
      "entropy": 5.644010400772094,
      "epoch": 2.4081976974990074,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004422597339537349,
      "loss": 5.0683,
      "mean_token_accuracy": 0.18348855525255203,
      "num_tokens": 53165089.0,
      "step": 24265
    },
    {
      "entropy": 5.647004842758179,
      "epoch": 2.408693926161175,
      "grad_norm": 1.125,
      "learning_rate": 0.0004422360327050032,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1730660542845726,
      "num_tokens": 53175614.0,
      "step": 24270
    },
    {
      "entropy": 5.666976737976074,
      "epoch": 2.4091901548233428,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044221232730917823,
      "loss": 4.9996,
      "mean_token_accuracy": 0.18577104508876802,
      "num_tokens": 53186306.0,
      "step": 24275
    },
    {
      "entropy": 5.647832298278809,
      "epoch": 2.4096863834855102,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044218861776684776,
      "loss": 5.071,
      "mean_token_accuracy": 0.1822887569665909,
      "num_tokens": 53197730.0,
      "step": 24280
    },
    {
      "entropy": 5.597526788711548,
      "epoch": 2.4101826121476777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044216490407859995,
      "loss": 4.9774,
      "mean_token_accuracy": 0.18554596304893495,
      "num_tokens": 53208486.0,
      "step": 24285
    },
    {
      "entropy": 5.684983110427856,
      "epoch": 2.410678840809845,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044214118624502267,
      "loss": 5.0434,
      "mean_token_accuracy": 0.18432655036449433,
      "num_tokens": 53220045.0,
      "step": 24290
    },
    {
      "entropy": 5.690744543075562,
      "epoch": 2.4111750694720127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004421174642667042,
      "loss": 5.134,
      "mean_token_accuracy": 0.17714149951934816,
      "num_tokens": 53231385.0,
      "step": 24295
    },
    {
      "entropy": 5.715699911117554,
      "epoch": 2.41167129813418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044209373814423265,
      "loss": 5.1054,
      "mean_token_accuracy": 0.18099375814199448,
      "num_tokens": 53243038.0,
      "step": 24300
    },
    {
      "entropy": 5.67984185218811,
      "epoch": 2.4121675267963476,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004420700078781965,
      "loss": 5.0769,
      "mean_token_accuracy": 0.1770028829574585,
      "num_tokens": 53252959.0,
      "step": 24305
    },
    {
      "entropy": 5.571003437042236,
      "epoch": 2.4126637554585155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044204627346918407,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18523896485567093,
      "num_tokens": 53263542.0,
      "step": 24310
    },
    {
      "entropy": 5.636269903182983,
      "epoch": 2.413159984120683,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044202253491778404,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18651655167341233,
      "num_tokens": 53273511.0,
      "step": 24315
    },
    {
      "entropy": 5.688987064361572,
      "epoch": 2.4136562127828505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004419987922245852,
      "loss": 5.1209,
      "mean_token_accuracy": 0.17763165831565858,
      "num_tokens": 53286173.0,
      "step": 24320
    },
    {
      "entropy": 5.651490116119385,
      "epoch": 2.414152441445018,
      "grad_norm": 1.15625,
      "learning_rate": 0.000441975045390176,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18540438115596772,
      "num_tokens": 53296784.0,
      "step": 24325
    },
    {
      "entropy": 5.685799884796142,
      "epoch": 2.4146486701071854,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044195129441514547,
      "loss": 5.038,
      "mean_token_accuracy": 0.1850449874997139,
      "num_tokens": 53308044.0,
      "step": 24330
    },
    {
      "entropy": 5.583820009231568,
      "epoch": 2.415144898769353,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044192753930008267,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18840708285570146,
      "num_tokens": 53319161.0,
      "step": 24335
    },
    {
      "entropy": 5.584305953979492,
      "epoch": 2.4156411274315204,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004419037800455765,
      "loss": 4.9614,
      "mean_token_accuracy": 0.19145094156265258,
      "num_tokens": 53329766.0,
      "step": 24340
    },
    {
      "entropy": 5.678786516189575,
      "epoch": 2.416137356093688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004418800166522163,
      "loss": 5.0569,
      "mean_token_accuracy": 0.18134965002536774,
      "num_tokens": 53340329.0,
      "step": 24345
    },
    {
      "entropy": 5.672094917297363,
      "epoch": 2.4166335847558553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044185624912059125,
      "loss": 5.0557,
      "mean_token_accuracy": 0.18445801436901094,
      "num_tokens": 53352592.0,
      "step": 24350
    },
    {
      "entropy": 5.689843368530274,
      "epoch": 2.417129813418023,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004418324774512908,
      "loss": 5.0345,
      "mean_token_accuracy": 0.18165375739336015,
      "num_tokens": 53362877.0,
      "step": 24355
    },
    {
      "entropy": 5.638301944732666,
      "epoch": 2.4176260420801907,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004418087016449043,
      "loss": 5.0474,
      "mean_token_accuracy": 0.1809063360095024,
      "num_tokens": 53374612.0,
      "step": 24360
    },
    {
      "entropy": 5.659292411804199,
      "epoch": 2.418122270742358,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044178492170202153,
      "loss": 5.0733,
      "mean_token_accuracy": 0.17481256872415543,
      "num_tokens": 53385129.0,
      "step": 24365
    },
    {
      "entropy": 5.684609651565552,
      "epoch": 2.4186184994045257,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044176113762323213,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17351271212100983,
      "num_tokens": 53395403.0,
      "step": 24370
    },
    {
      "entropy": 5.68206558227539,
      "epoch": 2.419114728066693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044173734940912586,
      "loss": 5.1045,
      "mean_token_accuracy": 0.16889066994190216,
      "num_tokens": 53406684.0,
      "step": 24375
    },
    {
      "entropy": 5.644416427612304,
      "epoch": 2.4196109567288606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044171355706029247,
      "loss": 4.9778,
      "mean_token_accuracy": 0.18677250444889068,
      "num_tokens": 53417256.0,
      "step": 24380
    },
    {
      "entropy": 5.59112229347229,
      "epoch": 2.420107185391028,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044168976057732225,
      "loss": 4.9833,
      "mean_token_accuracy": 0.1771117627620697,
      "num_tokens": 53428305.0,
      "step": 24385
    },
    {
      "entropy": 5.561171579360962,
      "epoch": 2.4206034140531956,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004416659599608051,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18021249622106553,
      "num_tokens": 53439301.0,
      "step": 24390
    },
    {
      "entropy": 5.698903369903564,
      "epoch": 2.4210996427153635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004416421552113312,
      "loss": 5.1413,
      "mean_token_accuracy": 0.17375903129577636,
      "num_tokens": 53450478.0,
      "step": 24395
    },
    {
      "entropy": 5.677814483642578,
      "epoch": 2.421595871377531,
      "grad_norm": 1.0,
      "learning_rate": 0.000441618346329491,
      "loss": 5.0611,
      "mean_token_accuracy": 0.17757441848516464,
      "num_tokens": 53461293.0,
      "step": 24400
    },
    {
      "entropy": 5.587476110458374,
      "epoch": 2.4220921000396984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044159453331587493,
      "loss": 5.0238,
      "mean_token_accuracy": 0.18812523037195206,
      "num_tokens": 53472014.0,
      "step": 24405
    },
    {
      "entropy": 5.6487095832824705,
      "epoch": 2.422588328701866,
      "grad_norm": 1.125,
      "learning_rate": 0.00044157071617107325,
      "loss": 5.0767,
      "mean_token_accuracy": 0.1817982316017151,
      "num_tokens": 53483010.0,
      "step": 24410
    },
    {
      "entropy": 5.686507320404052,
      "epoch": 2.4230845573640334,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044154689489567684,
      "loss": 5.0734,
      "mean_token_accuracy": 0.17969647347927092,
      "num_tokens": 53494151.0,
      "step": 24415
    },
    {
      "entropy": 5.645055913925171,
      "epoch": 2.423580786026201,
      "grad_norm": 1.125,
      "learning_rate": 0.00044152306949027625,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18423163294792175,
      "num_tokens": 53504462.0,
      "step": 24420
    },
    {
      "entropy": 5.617379570007325,
      "epoch": 2.4240770146883683,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044149923995546244,
      "loss": 5.0254,
      "mean_token_accuracy": 0.18824912011623382,
      "num_tokens": 53514896.0,
      "step": 24425
    },
    {
      "entropy": 5.612107515335083,
      "epoch": 2.424573243350536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004414754062918262,
      "loss": 4.996,
      "mean_token_accuracy": 0.18772246092557907,
      "num_tokens": 53524656.0,
      "step": 24430
    },
    {
      "entropy": 5.6672813415527346,
      "epoch": 2.4250694720127033,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004414515684999587,
      "loss": 5.0546,
      "mean_token_accuracy": 0.1877248615026474,
      "num_tokens": 53534896.0,
      "step": 24435
    },
    {
      "entropy": 5.603263664245605,
      "epoch": 2.4255657006748708,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004414277265804508,
      "loss": 4.9923,
      "mean_token_accuracy": 0.18739176988601686,
      "num_tokens": 53546108.0,
      "step": 24440
    },
    {
      "entropy": 5.606229829788208,
      "epoch": 2.4260619293370387,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000441403880533894,
      "loss": 5.0105,
      "mean_token_accuracy": 0.18166758865118027,
      "num_tokens": 53557325.0,
      "step": 24445
    },
    {
      "entropy": 5.5573036670684814,
      "epoch": 2.426558157999206,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044138003036087956,
      "loss": 4.943,
      "mean_token_accuracy": 0.18657089322805404,
      "num_tokens": 53568737.0,
      "step": 24450
    },
    {
      "entropy": 5.649560594558716,
      "epoch": 2.4270543866613736,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004413561760619988,
      "loss": 5.0666,
      "mean_token_accuracy": 0.174499973654747,
      "num_tokens": 53579312.0,
      "step": 24455
    },
    {
      "entropy": 5.601145887374878,
      "epoch": 2.427550615323541,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004413323176378434,
      "loss": 4.9569,
      "mean_token_accuracy": 0.19068650901317596,
      "num_tokens": 53590476.0,
      "step": 24460
    },
    {
      "entropy": 5.649459171295166,
      "epoch": 2.4280468439857086,
      "grad_norm": 1.0,
      "learning_rate": 0.0004413084550890049,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1736850082874298,
      "num_tokens": 53601284.0,
      "step": 24465
    },
    {
      "entropy": 5.647459983825684,
      "epoch": 2.428543072647876,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004412845884160751,
      "loss": 5.0724,
      "mean_token_accuracy": 0.17417675703763963,
      "num_tokens": 53612590.0,
      "step": 24470
    },
    {
      "entropy": 5.662662172317505,
      "epoch": 2.4290393013100435,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004412607176196458,
      "loss": 5.0942,
      "mean_token_accuracy": 0.17566999346017836,
      "num_tokens": 53623049.0,
      "step": 24475
    },
    {
      "entropy": 5.625589752197266,
      "epoch": 2.429535529972211,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044123684270030907,
      "loss": 5.0195,
      "mean_token_accuracy": 0.18437493741512298,
      "num_tokens": 53633961.0,
      "step": 24480
    },
    {
      "entropy": 5.634482717514038,
      "epoch": 2.430031758634379,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044121296365865674,
      "loss": 5.0688,
      "mean_token_accuracy": 0.17547668218612672,
      "num_tokens": 53644634.0,
      "step": 24485
    },
    {
      "entropy": 5.623129367828369,
      "epoch": 2.4305279872965464,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044118908049528105,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18519652038812637,
      "num_tokens": 53655982.0,
      "step": 24490
    },
    {
      "entropy": 5.701758670806885,
      "epoch": 2.431024215958714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044116519321077435,
      "loss": 5.1018,
      "mean_token_accuracy": 0.17538531869649887,
      "num_tokens": 53667367.0,
      "step": 24495
    },
    {
      "entropy": 5.618528938293457,
      "epoch": 2.4315204446208813,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044114130180572883,
      "loss": 4.9819,
      "mean_token_accuracy": 0.18445915132761,
      "num_tokens": 53677191.0,
      "step": 24500
    },
    {
      "entropy": 5.6278563976287845,
      "epoch": 2.432016673283049,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004411174062807371,
      "loss": 5.0606,
      "mean_token_accuracy": 0.18296241760253906,
      "num_tokens": 53688636.0,
      "step": 24505
    },
    {
      "entropy": 5.646000337600708,
      "epoch": 2.4325129019452163,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004410935066363916,
      "loss": 5.0193,
      "mean_token_accuracy": 0.17996798306703568,
      "num_tokens": 53699588.0,
      "step": 24510
    },
    {
      "entropy": 5.68361463546753,
      "epoch": 2.4330091306073838,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044106960287328503,
      "loss": 5.1355,
      "mean_token_accuracy": 0.17193110138177872,
      "num_tokens": 53710427.0,
      "step": 24515
    },
    {
      "entropy": 5.702255725860596,
      "epoch": 2.4335053592695512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004410456949920102,
      "loss": 5.0919,
      "mean_token_accuracy": 0.1709837421774864,
      "num_tokens": 53722319.0,
      "step": 24520
    },
    {
      "entropy": 5.648608636856079,
      "epoch": 2.4340015879317187,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044102178299315994,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17703719884157182,
      "num_tokens": 53734059.0,
      "step": 24525
    },
    {
      "entropy": 5.6585616111755375,
      "epoch": 2.4344978165938866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044099786687732713,
      "loss": 5.0689,
      "mean_token_accuracy": 0.17957741171121597,
      "num_tokens": 53745795.0,
      "step": 24530
    },
    {
      "entropy": 5.652870178222656,
      "epoch": 2.434994045256054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004409739466451049,
      "loss": 5.0761,
      "mean_token_accuracy": 0.17751525491476058,
      "num_tokens": 53756781.0,
      "step": 24535
    },
    {
      "entropy": 5.7313707828521725,
      "epoch": 2.4354902739182216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044095002229708655,
      "loss": 5.127,
      "mean_token_accuracy": 0.17795209884643554,
      "num_tokens": 53769725.0,
      "step": 24540
    },
    {
      "entropy": 5.6448540687561035,
      "epoch": 2.435986502580389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044092609383386517,
      "loss": 4.9996,
      "mean_token_accuracy": 0.19067037105560303,
      "num_tokens": 53781217.0,
      "step": 24545
    },
    {
      "entropy": 5.658794116973877,
      "epoch": 2.4364827312425565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044090216125603426,
      "loss": 5.0867,
      "mean_token_accuracy": 0.18763491958379747,
      "num_tokens": 53791478.0,
      "step": 24550
    },
    {
      "entropy": 5.5969305515289305,
      "epoch": 2.436978959904724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004408782245641872,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18784477412700654,
      "num_tokens": 53802844.0,
      "step": 24555
    },
    {
      "entropy": 5.6631875991821286,
      "epoch": 2.4374751885668915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004408542837589176,
      "loss": 5.0632,
      "mean_token_accuracy": 0.1799333617091179,
      "num_tokens": 53812938.0,
      "step": 24560
    },
    {
      "entropy": 5.660980987548828,
      "epoch": 2.437971417229059,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044083033884081914,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18094178140163422,
      "num_tokens": 53824105.0,
      "step": 24565
    },
    {
      "entropy": 5.718647384643555,
      "epoch": 2.438467645891227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004408063898104856,
      "loss": 5.1698,
      "mean_token_accuracy": 0.17348246276378632,
      "num_tokens": 53834956.0,
      "step": 24570
    },
    {
      "entropy": 5.668689012527466,
      "epoch": 2.4389638745533944,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044078243666851094,
      "loss": 5.0669,
      "mean_token_accuracy": 0.17825323045253755,
      "num_tokens": 53846257.0,
      "step": 24575
    },
    {
      "entropy": 5.62367696762085,
      "epoch": 2.439460103215562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044075847941548905,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18446790874004365,
      "num_tokens": 53857107.0,
      "step": 24580
    },
    {
      "entropy": 5.643803358078003,
      "epoch": 2.4399563318777293,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044073451805201396,
      "loss": 5.0478,
      "mean_token_accuracy": 0.17916251122951507,
      "num_tokens": 53867779.0,
      "step": 24585
    },
    {
      "entropy": 5.658034133911133,
      "epoch": 2.440452560539897,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044071055257868,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18055958896875382,
      "num_tokens": 53878661.0,
      "step": 24590
    },
    {
      "entropy": 5.738666772842407,
      "epoch": 2.4409487892020643,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044068658299608145,
      "loss": 5.1242,
      "mean_token_accuracy": 0.18005000352859496,
      "num_tokens": 53891300.0,
      "step": 24595
    },
    {
      "entropy": 5.676792335510254,
      "epoch": 2.4414450178642317,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044066260930481264,
      "loss": 4.9905,
      "mean_token_accuracy": 0.183909772336483,
      "num_tokens": 53901086.0,
      "step": 24600
    },
    {
      "entropy": 5.653081178665161,
      "epoch": 2.441941246526399,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004406386315054681,
      "loss": 5.0814,
      "mean_token_accuracy": 0.17449379414319993,
      "num_tokens": 53912599.0,
      "step": 24605
    },
    {
      "entropy": 5.585606241226197,
      "epoch": 2.4424374751885667,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044061464959864243,
      "loss": 5.0671,
      "mean_token_accuracy": 0.17792797982692718,
      "num_tokens": 53924293.0,
      "step": 24610
    },
    {
      "entropy": 5.669604635238647,
      "epoch": 2.4429337038507346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004405906635849302,
      "loss": 5.1226,
      "mean_token_accuracy": 0.17749904990196227,
      "num_tokens": 53935074.0,
      "step": 24615
    },
    {
      "entropy": 5.669009256362915,
      "epoch": 2.443429932512902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004405666734649265,
      "loss": 5.0546,
      "mean_token_accuracy": 0.18085940927267075,
      "num_tokens": 53946770.0,
      "step": 24620
    },
    {
      "entropy": 5.662995052337647,
      "epoch": 2.4439261611750696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044054267923922594,
      "loss": 5.0225,
      "mean_token_accuracy": 0.1811560198664665,
      "num_tokens": 53957077.0,
      "step": 24625
    },
    {
      "entropy": 5.646111536026001,
      "epoch": 2.444422389837237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044051868090842376,
      "loss": 5.0562,
      "mean_token_accuracy": 0.18507880866527557,
      "num_tokens": 53968455.0,
      "step": 24630
    },
    {
      "entropy": 5.562753677368164,
      "epoch": 2.4449186184994045,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044049467847311486,
      "loss": 5.0216,
      "mean_token_accuracy": 0.1786852225661278,
      "num_tokens": 53980429.0,
      "step": 24635
    },
    {
      "entropy": 5.603952646255493,
      "epoch": 2.445414847161572,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004404706719338946,
      "loss": 5.0187,
      "mean_token_accuracy": 0.18867385983467103,
      "num_tokens": 53990937.0,
      "step": 24640
    },
    {
      "entropy": 5.652561616897583,
      "epoch": 2.4459110758237395,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044044666129135825,
      "loss": 5.0185,
      "mean_token_accuracy": 0.18719376027584075,
      "num_tokens": 54003526.0,
      "step": 24645
    },
    {
      "entropy": 5.620787477493286,
      "epoch": 2.446407304485907,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044042264654610115,
      "loss": 4.9824,
      "mean_token_accuracy": 0.19089700579643248,
      "num_tokens": 54013875.0,
      "step": 24650
    },
    {
      "entropy": 5.678896474838257,
      "epoch": 2.446903533148075,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044039862769871884,
      "loss": 5.1577,
      "mean_token_accuracy": 0.17602687180042267,
      "num_tokens": 54024103.0,
      "step": 24655
    },
    {
      "entropy": 5.679017496109009,
      "epoch": 2.4473997618102423,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044037460474980703,
      "loss": 5.1159,
      "mean_token_accuracy": 0.17360696345567703,
      "num_tokens": 54035417.0,
      "step": 24660
    },
    {
      "entropy": 5.6864423751831055,
      "epoch": 2.44789599047241,
      "grad_norm": 1.125,
      "learning_rate": 0.0004403505776999614,
      "loss": 5.1139,
      "mean_token_accuracy": 0.16950680166482926,
      "num_tokens": 54045180.0,
      "step": 24665
    },
    {
      "entropy": 5.683794260025024,
      "epoch": 2.4483922191345773,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044032654654977766,
      "loss": 4.9984,
      "mean_token_accuracy": 0.18990812301635743,
      "num_tokens": 54056493.0,
      "step": 24670
    },
    {
      "entropy": 5.604797506332398,
      "epoch": 2.4488884477967447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004403025112998519,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18386187106370927,
      "num_tokens": 54067445.0,
      "step": 24675
    },
    {
      "entropy": 5.6507385730743405,
      "epoch": 2.4493846764589122,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004402784719507799,
      "loss": 5.0282,
      "mean_token_accuracy": 0.18702514320611954,
      "num_tokens": 54076937.0,
      "step": 24680
    },
    {
      "entropy": 5.685776090621948,
      "epoch": 2.4498809051210797,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004402544285031581,
      "loss": 5.0731,
      "mean_token_accuracy": 0.18503032624721527,
      "num_tokens": 54088014.0,
      "step": 24685
    },
    {
      "entropy": 5.681982040405273,
      "epoch": 2.450377133783247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004402303809575825,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18604982197284697,
      "num_tokens": 54098984.0,
      "step": 24690
    },
    {
      "entropy": 5.7052772521972654,
      "epoch": 2.4508733624454146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004402063293146495,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17853373736143113,
      "num_tokens": 54110490.0,
      "step": 24695
    },
    {
      "entropy": 5.6699601173400875,
      "epoch": 2.451369591107582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401822735749555,
      "loss": 5.0536,
      "mean_token_accuracy": 0.17823436260223388,
      "num_tokens": 54120785.0,
      "step": 24700
    },
    {
      "entropy": 5.660261583328247,
      "epoch": 2.45186581976975,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044015821373909707,
      "loss": 5.0048,
      "mean_token_accuracy": 0.1809498205780983,
      "num_tokens": 54131339.0,
      "step": 24705
    },
    {
      "entropy": 5.672362327575684,
      "epoch": 2.4523620484319175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044013414980767084,
      "loss": 5.073,
      "mean_token_accuracy": 0.17514344304800034,
      "num_tokens": 54142270.0,
      "step": 24710
    },
    {
      "entropy": 5.664493608474731,
      "epoch": 2.452858277094085,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004401100817812735,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18149049580097198,
      "num_tokens": 54153045.0,
      "step": 24715
    },
    {
      "entropy": 5.614759254455566,
      "epoch": 2.4533545057562525,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044008600966050196,
      "loss": 4.9836,
      "mean_token_accuracy": 0.18450899571180343,
      "num_tokens": 54164069.0,
      "step": 24720
    },
    {
      "entropy": 5.669102191925049,
      "epoch": 2.45385073441842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004400619334459531,
      "loss": 5.0668,
      "mean_token_accuracy": 0.17783805578947068,
      "num_tokens": 54176902.0,
      "step": 24725
    },
    {
      "entropy": 5.600476884841919,
      "epoch": 2.4543469630805874,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044003785313822386,
      "loss": 5.0127,
      "mean_token_accuracy": 0.1915619194507599,
      "num_tokens": 54188515.0,
      "step": 24730
    },
    {
      "entropy": 5.688239812850952,
      "epoch": 2.454843191742755,
      "grad_norm": 1.0,
      "learning_rate": 0.00044001376873791166,
      "loss": 5.0784,
      "mean_token_accuracy": 0.1804449662566185,
      "num_tokens": 54199077.0,
      "step": 24735
    },
    {
      "entropy": 5.747378730773926,
      "epoch": 2.455339420404923,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00043998968024561337,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18354761004447936,
      "num_tokens": 54210216.0,
      "step": 24740
    },
    {
      "entropy": 5.6147809505462645,
      "epoch": 2.4558356490670903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004399655876619267,
      "loss": 5.031,
      "mean_token_accuracy": 0.18267270624637605,
      "num_tokens": 54221143.0,
      "step": 24745
    },
    {
      "entropy": 5.678212881088257,
      "epoch": 2.4563318777292578,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043994149098744884,
      "loss": 5.0873,
      "mean_token_accuracy": 0.18121718168258666,
      "num_tokens": 54233070.0,
      "step": 24750
    },
    {
      "entropy": 5.652835559844971,
      "epoch": 2.4568281063914252,
      "grad_norm": 1.0,
      "learning_rate": 0.0004399173902227775,
      "loss": 5.0212,
      "mean_token_accuracy": 0.183928781747818,
      "num_tokens": 54243793.0,
      "step": 24755
    },
    {
      "entropy": 5.606001901626587,
      "epoch": 2.4573243350535927,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043989328536851014,
      "loss": 5.037,
      "mean_token_accuracy": 0.18383980840444564,
      "num_tokens": 54256983.0,
      "step": 24760
    },
    {
      "entropy": 5.631918334960938,
      "epoch": 2.45782056371576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043986917642524465,
      "loss": 5.0183,
      "mean_token_accuracy": 0.1847965106368065,
      "num_tokens": 54268789.0,
      "step": 24765
    },
    {
      "entropy": 5.758157110214233,
      "epoch": 2.4583167923779277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043984506339357887,
      "loss": 5.0503,
      "mean_token_accuracy": 0.17345488518476487,
      "num_tokens": 54279517.0,
      "step": 24770
    },
    {
      "entropy": 5.668013000488282,
      "epoch": 2.458813021040095,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004398209462741107,
      "loss": 5.0437,
      "mean_token_accuracy": 0.1808445706963539,
      "num_tokens": 54290865.0,
      "step": 24775
    },
    {
      "entropy": 5.763889646530151,
      "epoch": 2.4593092497022626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397968250674383,
      "loss": 5.0804,
      "mean_token_accuracy": 0.18832856118679048,
      "num_tokens": 54302658.0,
      "step": 24780
    },
    {
      "entropy": 5.644113779067993,
      "epoch": 2.45980547836443,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004397726997741597,
      "loss": 5.0725,
      "mean_token_accuracy": 0.1841200903058052,
      "num_tokens": 54314000.0,
      "step": 24785
    },
    {
      "entropy": 5.582438802719116,
      "epoch": 2.460301707026598,
      "grad_norm": 1.0,
      "learning_rate": 0.0004397485703948732,
      "loss": 5.0262,
      "mean_token_accuracy": 0.1878236010670662,
      "num_tokens": 54325952.0,
      "step": 24790
    },
    {
      "entropy": 5.683558511734009,
      "epoch": 2.4607979356887655,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004397244369301771,
      "loss": 5.1246,
      "mean_token_accuracy": 0.17412910163402556,
      "num_tokens": 54338582.0,
      "step": 24795
    },
    {
      "entropy": 5.691143131256103,
      "epoch": 2.461294164350933,
      "grad_norm": 1.25,
      "learning_rate": 0.00043970029938066995,
      "loss": 5.0881,
      "mean_token_accuracy": 0.17502160370349884,
      "num_tokens": 54349513.0,
      "step": 24800
    },
    {
      "entropy": 5.66074538230896,
      "epoch": 2.4617903930131004,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004396761577469503,
      "loss": 5.03,
      "mean_token_accuracy": 0.1782473012804985,
      "num_tokens": 54360264.0,
      "step": 24805
    },
    {
      "entropy": 5.666816520690918,
      "epoch": 2.462286621675268,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043965201202961673,
      "loss": 5.0908,
      "mean_token_accuracy": 0.1754157468676567,
      "num_tokens": 54371379.0,
      "step": 24810
    },
    {
      "entropy": 5.697214555740357,
      "epoch": 2.4627828503374354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043962786222926807,
      "loss": 5.0604,
      "mean_token_accuracy": 0.18566280603408813,
      "num_tokens": 54382844.0,
      "step": 24815
    },
    {
      "entropy": 5.637014579772949,
      "epoch": 2.463279078999603,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043960370834650325,
      "loss": 5.017,
      "mean_token_accuracy": 0.1849992573261261,
      "num_tokens": 54392807.0,
      "step": 24820
    },
    {
      "entropy": 5.579197645187378,
      "epoch": 2.4637753076617708,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004395795503819211,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18532495945692062,
      "num_tokens": 54404060.0,
      "step": 24825
    },
    {
      "entropy": 5.636567401885986,
      "epoch": 2.4642715363239382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004395553883361207,
      "loss": 4.9595,
      "mean_token_accuracy": 0.1819923460483551,
      "num_tokens": 54414636.0,
      "step": 24830
    },
    {
      "entropy": 5.640518760681152,
      "epoch": 2.4647677649861057,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043953122220970127,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18435487598180772,
      "num_tokens": 54426185.0,
      "step": 24835
    },
    {
      "entropy": 5.636225795745849,
      "epoch": 2.465263993648273,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043950705200326205,
      "loss": 5.0266,
      "mean_token_accuracy": 0.1785135105252266,
      "num_tokens": 54436468.0,
      "step": 24840
    },
    {
      "entropy": 5.621134996414185,
      "epoch": 2.4657602223104407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004394828777174025,
      "loss": 5.0344,
      "mean_token_accuracy": 0.18397989124059677,
      "num_tokens": 54446228.0,
      "step": 24845
    },
    {
      "entropy": 5.605884456634522,
      "epoch": 2.466256450972608,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043945869935272194,
      "loss": 5.0758,
      "mean_token_accuracy": 0.17334654182195663,
      "num_tokens": 54456704.0,
      "step": 24850
    },
    {
      "entropy": 5.675971937179566,
      "epoch": 2.4667526796347756,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043943451690982006,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18875830322504045,
      "num_tokens": 54467780.0,
      "step": 24855
    },
    {
      "entropy": 5.750992679595948,
      "epoch": 2.467248908296943,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004394103303892964,
      "loss": 5.1564,
      "mean_token_accuracy": 0.17702420502901078,
      "num_tokens": 54478428.0,
      "step": 24860
    },
    {
      "entropy": 5.590981483459473,
      "epoch": 2.4677451369591106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043938613979175087,
      "loss": 4.9442,
      "mean_token_accuracy": 0.18632734566926956,
      "num_tokens": 54489544.0,
      "step": 24865
    },
    {
      "entropy": 5.6061936378479,
      "epoch": 2.468241365621278,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004393619451177834,
      "loss": 5.0292,
      "mean_token_accuracy": 0.1852469727396965,
      "num_tokens": 54501438.0,
      "step": 24870
    },
    {
      "entropy": 5.572884845733642,
      "epoch": 2.468737594283446,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043933774636799365,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18791565597057341,
      "num_tokens": 54511746.0,
      "step": 24875
    },
    {
      "entropy": 5.679198837280273,
      "epoch": 2.4692338229456134,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043931354354298204,
      "loss": 5.0702,
      "mean_token_accuracy": 0.17872852832078934,
      "num_tokens": 54523141.0,
      "step": 24880
    },
    {
      "entropy": 5.785866355895996,
      "epoch": 2.469730051607781,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004392893366433486,
      "loss": 5.1928,
      "mean_token_accuracy": 0.1736557051539421,
      "num_tokens": 54533275.0,
      "step": 24885
    },
    {
      "entropy": 5.710293006896973,
      "epoch": 2.4702262802699484,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043926512566969366,
      "loss": 5.027,
      "mean_token_accuracy": 0.1813216581940651,
      "num_tokens": 54545309.0,
      "step": 24890
    },
    {
      "entropy": 5.592803716659546,
      "epoch": 2.470722508932116,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043924091062261743,
      "loss": 4.9919,
      "mean_token_accuracy": 0.19042752236127852,
      "num_tokens": 54554905.0,
      "step": 24895
    },
    {
      "entropy": 5.6495462417602536,
      "epoch": 2.4712187375942833,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004392166915027206,
      "loss": 5.0408,
      "mean_token_accuracy": 0.1759948745369911,
      "num_tokens": 54566050.0,
      "step": 24900
    },
    {
      "entropy": 5.679964303970337,
      "epoch": 2.471714966256451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004391924683106037,
      "loss": 5.1036,
      "mean_token_accuracy": 0.1773971885442734,
      "num_tokens": 54577791.0,
      "step": 24905
    },
    {
      "entropy": 5.662644004821777,
      "epoch": 2.4722111949186187,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004391682410468674,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18441083580255507,
      "num_tokens": 54589183.0,
      "step": 24910
    },
    {
      "entropy": 5.607084703445435,
      "epoch": 2.472707423580786,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043914400971211255,
      "loss": 5.0331,
      "mean_token_accuracy": 0.17492370009422303,
      "num_tokens": 54600776.0,
      "step": 24915
    },
    {
      "entropy": 5.623539161682129,
      "epoch": 2.4732036522429537,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004391197743069398,
      "loss": 4.9955,
      "mean_token_accuracy": 0.19291624128818513,
      "num_tokens": 54611167.0,
      "step": 24920
    },
    {
      "entropy": 5.637652826309204,
      "epoch": 2.473699880905121,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004390955348319503,
      "loss": 5.0399,
      "mean_token_accuracy": 0.18451584577560426,
      "num_tokens": 54622811.0,
      "step": 24925
    },
    {
      "entropy": 5.6436079978942875,
      "epoch": 2.4741961095672886,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043907129128774525,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18180340826511382,
      "num_tokens": 54635109.0,
      "step": 24930
    },
    {
      "entropy": 5.704784917831421,
      "epoch": 2.474692338229456,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043904704367492563,
      "loss": 5.0829,
      "mean_token_accuracy": 0.18493657410144806,
      "num_tokens": 54644260.0,
      "step": 24935
    },
    {
      "entropy": 5.628142976760865,
      "epoch": 2.4751885668916236,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004390227919940928,
      "loss": 4.9754,
      "mean_token_accuracy": 0.18396759331226348,
      "num_tokens": 54654433.0,
      "step": 24940
    },
    {
      "entropy": 5.669334125518799,
      "epoch": 2.475684795553791,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043899853624584825,
      "loss": 5.0983,
      "mean_token_accuracy": 0.17520237267017363,
      "num_tokens": 54666582.0,
      "step": 24945
    },
    {
      "entropy": 5.6449743747711185,
      "epoch": 2.4761810242159585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043897427643079336,
      "loss": 4.9966,
      "mean_token_accuracy": 0.18667120337486268,
      "num_tokens": 54677247.0,
      "step": 24950
    },
    {
      "entropy": 5.656476831436157,
      "epoch": 2.476677252878126,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043895001254952963,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17914730682969093,
      "num_tokens": 54688423.0,
      "step": 24955
    },
    {
      "entropy": 5.653966617584229,
      "epoch": 2.477173481540294,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000438925744602659,
      "loss": 5.0594,
      "mean_token_accuracy": 0.18538887649774552,
      "num_tokens": 54698517.0,
      "step": 24960
    },
    {
      "entropy": 5.734698390960693,
      "epoch": 2.4776697102024614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043890147259078314,
      "loss": 5.0939,
      "mean_token_accuracy": 0.17857514768838884,
      "num_tokens": 54711635.0,
      "step": 24965
    },
    {
      "entropy": 5.682784986495972,
      "epoch": 2.478165938864629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043887719651450384,
      "loss": 5.0567,
      "mean_token_accuracy": 0.18409755378961562,
      "num_tokens": 54723051.0,
      "step": 24970
    },
    {
      "entropy": 5.741815662384033,
      "epoch": 2.4786621675267964,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004388529163744233,
      "loss": 5.0491,
      "mean_token_accuracy": 0.18228800743818283,
      "num_tokens": 54734191.0,
      "step": 24975
    },
    {
      "entropy": 5.636474227905273,
      "epoch": 2.479158396188964,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004388286321711435,
      "loss": 5.0166,
      "mean_token_accuracy": 0.17897817641496658,
      "num_tokens": 54745632.0,
      "step": 24980
    },
    {
      "entropy": 5.620825624465942,
      "epoch": 2.4796546248511313,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004388043439052665,
      "loss": 5.0772,
      "mean_token_accuracy": 0.17310031801462172,
      "num_tokens": 54755871.0,
      "step": 24985
    },
    {
      "entropy": 5.605988168716431,
      "epoch": 2.480150853513299,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043878005157739486,
      "loss": 4.9798,
      "mean_token_accuracy": 0.18682535737752914,
      "num_tokens": 54768036.0,
      "step": 24990
    },
    {
      "entropy": 5.664324617385864,
      "epoch": 2.4806470821754667,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004387557551881308,
      "loss": 5.0072,
      "mean_token_accuracy": 0.18111298382282257,
      "num_tokens": 54778518.0,
      "step": 24995
    },
    {
      "entropy": 5.618691158294678,
      "epoch": 2.481143310837634,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000438731454738077,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18805916607379913,
      "num_tokens": 54790786.0,
      "step": 25000
    },
    {
      "entropy": 5.602381658554077,
      "epoch": 2.4816395394998016,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043870715022783575,
      "loss": 4.9988,
      "mean_token_accuracy": 0.1821173384785652,
      "num_tokens": 54800063.0,
      "step": 25005
    },
    {
      "entropy": 5.6233960628509525,
      "epoch": 2.482135768161969,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004386828416580101,
      "loss": 5.0743,
      "mean_token_accuracy": 0.17406624555587769,
      "num_tokens": 54810605.0,
      "step": 25010
    },
    {
      "entropy": 5.653490114212036,
      "epoch": 2.4826319968241366,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004386585290292026,
      "loss": 5.0005,
      "mean_token_accuracy": 0.1810569405555725,
      "num_tokens": 54820626.0,
      "step": 25015
    },
    {
      "entropy": 5.615113878250122,
      "epoch": 2.483128225486304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004386342123420161,
      "loss": 4.992,
      "mean_token_accuracy": 0.1807320237159729,
      "num_tokens": 54831395.0,
      "step": 25020
    },
    {
      "entropy": 5.583057022094726,
      "epoch": 2.4836244541484715,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043860989159705396,
      "loss": 5.0103,
      "mean_token_accuracy": 0.1792276009917259,
      "num_tokens": 54841238.0,
      "step": 25025
    },
    {
      "entropy": 5.6807506561279295,
      "epoch": 2.484120682810639,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043858556679491896,
      "loss": 5.089,
      "mean_token_accuracy": 0.18474095314741135,
      "num_tokens": 54851301.0,
      "step": 25030
    },
    {
      "entropy": 5.665053653717041,
      "epoch": 2.4846169114728065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043856123793621437,
      "loss": 5.0002,
      "mean_token_accuracy": 0.1878359615802765,
      "num_tokens": 54863135.0,
      "step": 25035
    },
    {
      "entropy": 5.646624135971069,
      "epoch": 2.485113140134974,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043853690502154355,
      "loss": 5.0813,
      "mean_token_accuracy": 0.179325470328331,
      "num_tokens": 54874780.0,
      "step": 25040
    },
    {
      "entropy": 5.669253015518189,
      "epoch": 2.485609368797142,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004385125680515099,
      "loss": 5.0667,
      "mean_token_accuracy": 0.1844715505838394,
      "num_tokens": 54885237.0,
      "step": 25045
    },
    {
      "entropy": 5.669059181213379,
      "epoch": 2.4861055974593094,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004384882270267168,
      "loss": 5.0763,
      "mean_token_accuracy": 0.1839492216706276,
      "num_tokens": 54896801.0,
      "step": 25050
    },
    {
      "entropy": 5.621484470367432,
      "epoch": 2.486601826121477,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004384638819477681,
      "loss": 5.027,
      "mean_token_accuracy": 0.17712628543376924,
      "num_tokens": 54908427.0,
      "step": 25055
    },
    {
      "entropy": 5.532269239425659,
      "epoch": 2.4870980547836443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043843953281526727,
      "loss": 4.9862,
      "mean_token_accuracy": 0.19014714360237123,
      "num_tokens": 54918394.0,
      "step": 25060
    },
    {
      "entropy": 5.630861711502075,
      "epoch": 2.487594283445812,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043841517962981823,
      "loss": 5.0182,
      "mean_token_accuracy": 0.18236323595046997,
      "num_tokens": 54928613.0,
      "step": 25065
    },
    {
      "entropy": 5.67335147857666,
      "epoch": 2.4880905121079793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043839082239202496,
      "loss": 4.9784,
      "mean_token_accuracy": 0.1838345631957054,
      "num_tokens": 54938938.0,
      "step": 25070
    },
    {
      "entropy": 5.704059600830078,
      "epoch": 2.4885867407701467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004383664611024912,
      "loss": 5.1154,
      "mean_token_accuracy": 0.18143160939216613,
      "num_tokens": 54950152.0,
      "step": 25075
    },
    {
      "entropy": 5.676141738891602,
      "epoch": 2.489082969432314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004383420957618213,
      "loss": 5.1459,
      "mean_token_accuracy": 0.17239236384630202,
      "num_tokens": 54962191.0,
      "step": 25080
    },
    {
      "entropy": 5.6278821468353275,
      "epoch": 2.489579198094482,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043831772637061944,
      "loss": 4.9524,
      "mean_token_accuracy": 0.18533722162246705,
      "num_tokens": 54972532.0,
      "step": 25085
    },
    {
      "entropy": 5.71053466796875,
      "epoch": 2.4900754267566496,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043829335292948986,
      "loss": 5.0984,
      "mean_token_accuracy": 0.17520108073949814,
      "num_tokens": 54983212.0,
      "step": 25090
    },
    {
      "entropy": 5.578927612304687,
      "epoch": 2.490571655418817,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043826897543903696,
      "loss": 4.9903,
      "mean_token_accuracy": 0.18798344284296037,
      "num_tokens": 54994094.0,
      "step": 25095
    },
    {
      "entropy": 5.620845174789428,
      "epoch": 2.4910678840809846,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043824459389986534,
      "loss": 5.069,
      "mean_token_accuracy": 0.1816760092973709,
      "num_tokens": 55004098.0,
      "step": 25100
    },
    {
      "entropy": 5.620698928833008,
      "epoch": 2.491564112743152,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004382202083125795,
      "loss": 4.9562,
      "mean_token_accuracy": 0.18661869019269944,
      "num_tokens": 55013680.0,
      "step": 25105
    },
    {
      "entropy": 5.651741456985474,
      "epoch": 2.4920603414053195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004381958186777842,
      "loss": 4.9419,
      "mean_token_accuracy": 0.18977354019880294,
      "num_tokens": 55024311.0,
      "step": 25110
    },
    {
      "entropy": 5.638939476013183,
      "epoch": 2.492556570067487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004381714249960843,
      "loss": 5.1925,
      "mean_token_accuracy": 0.17365105003118514,
      "num_tokens": 55036215.0,
      "step": 25115
    },
    {
      "entropy": 5.66772985458374,
      "epoch": 2.4930527987296545,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004381470272680846,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1778908148407936,
      "num_tokens": 55046402.0,
      "step": 25120
    },
    {
      "entropy": 5.6605494022369385,
      "epoch": 2.493549027391822,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004381226254943902,
      "loss": 4.997,
      "mean_token_accuracy": 0.17967742532491685,
      "num_tokens": 55056307.0,
      "step": 25125
    },
    {
      "entropy": 5.6321813583374025,
      "epoch": 2.49404525605399,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004380982196756062,
      "loss": 5.0118,
      "mean_token_accuracy": 0.18277247101068497,
      "num_tokens": 55066918.0,
      "step": 25130
    },
    {
      "entropy": 5.626764822006225,
      "epoch": 2.4945414847161573,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043807380981233777,
      "loss": 5.0354,
      "mean_token_accuracy": 0.1802588328719139,
      "num_tokens": 55078166.0,
      "step": 25135
    },
    {
      "entropy": 5.645209455490113,
      "epoch": 2.495037713378325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004380493959051903,
      "loss": 5.0532,
      "mean_token_accuracy": 0.18467887192964555,
      "num_tokens": 55090919.0,
      "step": 25140
    },
    {
      "entropy": 5.694576215744019,
      "epoch": 2.4955339420404923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043802497795476916,
      "loss": 5.0774,
      "mean_token_accuracy": 0.18157800734043122,
      "num_tokens": 55100512.0,
      "step": 25145
    },
    {
      "entropy": 5.633235025405884,
      "epoch": 2.4960301707026598,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004380005559616798,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18734667003154754,
      "num_tokens": 55111845.0,
      "step": 25150
    },
    {
      "entropy": 5.690884590148926,
      "epoch": 2.4965263993648272,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004379761299265279,
      "loss": 5.153,
      "mean_token_accuracy": 0.17313326448202132,
      "num_tokens": 55123744.0,
      "step": 25155
    },
    {
      "entropy": 5.732018899917603,
      "epoch": 2.4970226280269947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004379516998499192,
      "loss": 5.0493,
      "mean_token_accuracy": 0.18919702023267745,
      "num_tokens": 55133789.0,
      "step": 25160
    },
    {
      "entropy": 5.662260389328003,
      "epoch": 2.497518856689162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004379272657324594,
      "loss": 5.0504,
      "mean_token_accuracy": 0.1834808275103569,
      "num_tokens": 55145325.0,
      "step": 25165
    },
    {
      "entropy": 5.65447154045105,
      "epoch": 2.49801508535133,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004379028275747545,
      "loss": 5.0132,
      "mean_token_accuracy": 0.18130835741758347,
      "num_tokens": 55157639.0,
      "step": 25170
    },
    {
      "entropy": 5.613567352294922,
      "epoch": 2.4985113140134976,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043787838537741054,
      "loss": 4.9687,
      "mean_token_accuracy": 0.18429943174123764,
      "num_tokens": 55167796.0,
      "step": 25175
    },
    {
      "entropy": 5.628173303604126,
      "epoch": 2.499007542675665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004378539391410336,
      "loss": 5.083,
      "mean_token_accuracy": 0.17921543270349502,
      "num_tokens": 55178384.0,
      "step": 25180
    },
    {
      "entropy": 5.730166625976563,
      "epoch": 2.4995037713378325,
      "grad_norm": 1.125,
      "learning_rate": 0.00043782948886622987,
      "loss": 5.1092,
      "mean_token_accuracy": 0.17823060601949692,
      "num_tokens": 55188363.0,
      "step": 25185
    },
    {
      "entropy": 5.717347574234009,
      "epoch": 2.5,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004378050345536057,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17933911979198455,
      "num_tokens": 55198293.0,
      "step": 25190
    },
    {
      "entropy": 5.651530981063843,
      "epoch": 2.5004962286621675,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043778057620376745,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18227852284908294,
      "num_tokens": 55208368.0,
      "step": 25195
    },
    {
      "entropy": 5.645267057418823,
      "epoch": 2.500992457324335,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043775611381732164,
      "loss": 5.0821,
      "mean_token_accuracy": 0.17820404171943666,
      "num_tokens": 55220075.0,
      "step": 25200
    },
    {
      "entropy": 5.626022052764893,
      "epoch": 2.5014886859865024,
      "grad_norm": 1.078125,
      "learning_rate": 0.000437731647394875,
      "loss": 5.0385,
      "mean_token_accuracy": 0.18249255418777466,
      "num_tokens": 55231060.0,
      "step": 25205
    },
    {
      "entropy": 5.681679677963257,
      "epoch": 2.50198491464867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004377071769370341,
      "loss": 5.0602,
      "mean_token_accuracy": 0.1823606237769127,
      "num_tokens": 55242443.0,
      "step": 25210
    },
    {
      "entropy": 5.53555703163147,
      "epoch": 2.5024811433108374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043768270244440587,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1902638703584671,
      "num_tokens": 55253285.0,
      "step": 25215
    },
    {
      "entropy": 5.593061351776123,
      "epoch": 2.5029773719730053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043765822391759714,
      "loss": 5.0018,
      "mean_token_accuracy": 0.18366309404373168,
      "num_tokens": 55264694.0,
      "step": 25220
    },
    {
      "entropy": 5.631147575378418,
      "epoch": 2.5034736006351728,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004376337413572149,
      "loss": 5.0293,
      "mean_token_accuracy": 0.1839797616004944,
      "num_tokens": 55275429.0,
      "step": 25225
    },
    {
      "entropy": 5.716000890731811,
      "epoch": 2.5039698292973402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043760925476386635,
      "loss": 5.0837,
      "mean_token_accuracy": 0.1790689378976822,
      "num_tokens": 55286764.0,
      "step": 25230
    },
    {
      "entropy": 5.543856573104859,
      "epoch": 2.5044660579595077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004375847641381587,
      "loss": 4.9298,
      "mean_token_accuracy": 0.18994757980108262,
      "num_tokens": 55296284.0,
      "step": 25235
    },
    {
      "entropy": 5.631589603424072,
      "epoch": 2.504962286621675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004375602694806993,
      "loss": 5.0331,
      "mean_token_accuracy": 0.18408808261156082,
      "num_tokens": 55307684.0,
      "step": 25240
    },
    {
      "entropy": 5.6350733757019045,
      "epoch": 2.5054585152838427,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043753577079209537,
      "loss": 5.0071,
      "mean_token_accuracy": 0.1779598131775856,
      "num_tokens": 55318202.0,
      "step": 25245
    },
    {
      "entropy": 5.67785964012146,
      "epoch": 2.5059547439460106,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043751126807295464,
      "loss": 5.0969,
      "mean_token_accuracy": 0.1783499836921692,
      "num_tokens": 55328827.0,
      "step": 25250
    },
    {
      "entropy": 5.686655902862549,
      "epoch": 2.506450972608178,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043748676132388464,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17331475913524627,
      "num_tokens": 55338171.0,
      "step": 25255
    },
    {
      "entropy": 5.6297996044158936,
      "epoch": 2.5069472012703455,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043746225054549316,
      "loss": 5.0412,
      "mean_token_accuracy": 0.18067149966955184,
      "num_tokens": 55349081.0,
      "step": 25260
    },
    {
      "entropy": 5.641663122177124,
      "epoch": 2.507443429932513,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043743773573838785,
      "loss": 5.0198,
      "mean_token_accuracy": 0.18432980328798293,
      "num_tokens": 55359344.0,
      "step": 25265
    },
    {
      "entropy": 5.62427659034729,
      "epoch": 2.5079396585946805,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043741321690317673,
      "loss": 4.9408,
      "mean_token_accuracy": 0.19213340878486634,
      "num_tokens": 55369643.0,
      "step": 25270
    },
    {
      "entropy": 5.619438314437867,
      "epoch": 2.508435887256848,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004373886940404677,
      "loss": 4.9519,
      "mean_token_accuracy": 0.19202728867530822,
      "num_tokens": 55380812.0,
      "step": 25275
    },
    {
      "entropy": 5.553179311752319,
      "epoch": 2.5089321159190154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004373641671508691,
      "loss": 5.0164,
      "mean_token_accuracy": 0.18903560638427735,
      "num_tokens": 55393136.0,
      "step": 25280
    },
    {
      "entropy": 5.5516037940979,
      "epoch": 2.509428344581183,
      "grad_norm": 1.0,
      "learning_rate": 0.000437339636234989,
      "loss": 4.941,
      "mean_token_accuracy": 0.1816619172692299,
      "num_tokens": 55404317.0,
      "step": 25285
    },
    {
      "entropy": 5.657345724105835,
      "epoch": 2.5099245732433504,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043731510129343574,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17792997062206267,
      "num_tokens": 55415812.0,
      "step": 25290
    },
    {
      "entropy": 5.701036500930786,
      "epoch": 2.510420801905518,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043729056232681766,
      "loss": 5.083,
      "mean_token_accuracy": 0.1814524456858635,
      "num_tokens": 55426402.0,
      "step": 25295
    },
    {
      "entropy": 5.6772315979003904,
      "epoch": 2.5109170305676853,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004372660193357434,
      "loss": 5.0835,
      "mean_token_accuracy": 0.18338994979858397,
      "num_tokens": 55437474.0,
      "step": 25300
    },
    {
      "entropy": 5.6318652629852295,
      "epoch": 2.5114132592298533,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004372414723208215,
      "loss": 5.0334,
      "mean_token_accuracy": 0.17707351446151734,
      "num_tokens": 55447587.0,
      "step": 25305
    },
    {
      "entropy": 5.737020492553711,
      "epoch": 2.5119094878920207,
      "grad_norm": 1.0,
      "learning_rate": 0.0004372169212826606,
      "loss": 5.1673,
      "mean_token_accuracy": 0.17581838071346284,
      "num_tokens": 55460587.0,
      "step": 25310
    },
    {
      "entropy": 5.651350593566894,
      "epoch": 2.512405716554188,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043719236622186976,
      "loss": 5.0319,
      "mean_token_accuracy": 0.18704639971256257,
      "num_tokens": 55471150.0,
      "step": 25315
    },
    {
      "entropy": 5.608349800109863,
      "epoch": 2.5129019452163557,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004371678071390576,
      "loss": 5.0191,
      "mean_token_accuracy": 0.1881071373820305,
      "num_tokens": 55480368.0,
      "step": 25320
    },
    {
      "entropy": 5.685173225402832,
      "epoch": 2.513398173878523,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043714324403483333,
      "loss": 5.0854,
      "mean_token_accuracy": 0.1787121996283531,
      "num_tokens": 55491954.0,
      "step": 25325
    },
    {
      "entropy": 5.675013065338135,
      "epoch": 2.5138944025406906,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043711867690980605,
      "loss": 5.0509,
      "mean_token_accuracy": 0.17514970451593398,
      "num_tokens": 55502560.0,
      "step": 25330
    },
    {
      "entropy": 5.612113761901855,
      "epoch": 2.5143906312028586,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043709410576458477,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18160830736160277,
      "num_tokens": 55514571.0,
      "step": 25335
    },
    {
      "entropy": 5.644307661056518,
      "epoch": 2.514886859865026,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043706953059977907,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18187057375907897,
      "num_tokens": 55525206.0,
      "step": 25340
    },
    {
      "entropy": 5.629896926879883,
      "epoch": 2.5153830885271935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043704495141599816,
      "loss": 5.0667,
      "mean_token_accuracy": 0.17357901036739348,
      "num_tokens": 55536551.0,
      "step": 25345
    },
    {
      "entropy": 5.65068941116333,
      "epoch": 2.515879317189361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004370203682138517,
      "loss": 5.0269,
      "mean_token_accuracy": 0.1759581372141838,
      "num_tokens": 55547842.0,
      "step": 25350
    },
    {
      "entropy": 5.72446928024292,
      "epoch": 2.5163755458515285,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004369957809939492,
      "loss": 4.9921,
      "mean_token_accuracy": 0.19718352258205413,
      "num_tokens": 55558248.0,
      "step": 25355
    },
    {
      "entropy": 5.620606136322022,
      "epoch": 2.516871774513696,
      "grad_norm": 1.0,
      "learning_rate": 0.0004369711897569004,
      "loss": 4.9854,
      "mean_token_accuracy": 0.17792647927999497,
      "num_tokens": 55569195.0,
      "step": 25360
    },
    {
      "entropy": 5.620577764511109,
      "epoch": 2.5173680031758634,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004369465945033151,
      "loss": 4.9638,
      "mean_token_accuracy": 0.19340836107730866,
      "num_tokens": 55579649.0,
      "step": 25365
    },
    {
      "entropy": 5.5511860847473145,
      "epoch": 2.517864231838031,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043692199523380325,
      "loss": 4.8807,
      "mean_token_accuracy": 0.19513580650091172,
      "num_tokens": 55589550.0,
      "step": 25370
    },
    {
      "entropy": 5.606029415130616,
      "epoch": 2.5183604605001984,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004368973919489748,
      "loss": 5.0233,
      "mean_token_accuracy": 0.18145364820957183,
      "num_tokens": 55600673.0,
      "step": 25375
    },
    {
      "entropy": 5.5400291919708256,
      "epoch": 2.518856689162366,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043687278464943986,
      "loss": 4.9859,
      "mean_token_accuracy": 0.18085746616125106,
      "num_tokens": 55612045.0,
      "step": 25380
    },
    {
      "entropy": 5.632215070724487,
      "epoch": 2.5193529178245333,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004368481733358088,
      "loss": 5.0346,
      "mean_token_accuracy": 0.18282643109560012,
      "num_tokens": 55623666.0,
      "step": 25385
    },
    {
      "entropy": 5.614487648010254,
      "epoch": 2.519849146486701,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004368235580086917,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1844155579805374,
      "num_tokens": 55634236.0,
      "step": 25390
    },
    {
      "entropy": 5.5563719272613525,
      "epoch": 2.5203453751488687,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043679893866869904,
      "loss": 4.9588,
      "mean_token_accuracy": 0.18647626489400865,
      "num_tokens": 55645412.0,
      "step": 25395
    },
    {
      "entropy": 5.605341815948487,
      "epoch": 2.520841603811036,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004367743153164414,
      "loss": 4.9959,
      "mean_token_accuracy": 0.18101792335510253,
      "num_tokens": 55655861.0,
      "step": 25400
    },
    {
      "entropy": 5.609581565856933,
      "epoch": 2.5213378324732036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043674968795252926,
      "loss": 4.9707,
      "mean_token_accuracy": 0.1866985574364662,
      "num_tokens": 55665227.0,
      "step": 25405
    },
    {
      "entropy": 5.625945711135865,
      "epoch": 2.521834061135371,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004367250565775734,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18944838494062424,
      "num_tokens": 55676104.0,
      "step": 25410
    },
    {
      "entropy": 5.632349348068237,
      "epoch": 2.5223302897975386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004367004211921847,
      "loss": 5.097,
      "mean_token_accuracy": 0.17473617792129517,
      "num_tokens": 55688222.0,
      "step": 25415
    },
    {
      "entropy": 5.636612319946289,
      "epoch": 2.522826518459706,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043667578179697387,
      "loss": 5.0387,
      "mean_token_accuracy": 0.18256115317344665,
      "num_tokens": 55700138.0,
      "step": 25420
    },
    {
      "entropy": 5.6416120529174805,
      "epoch": 2.523322747121874,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004366511383925521,
      "loss": 4.9889,
      "mean_token_accuracy": 0.18769195973873137,
      "num_tokens": 55709444.0,
      "step": 25425
    },
    {
      "entropy": 5.719867753982544,
      "epoch": 2.5238189757840415,
      "grad_norm": 1.25,
      "learning_rate": 0.0004366264909795305,
      "loss": 5.1296,
      "mean_token_accuracy": 0.1819441571831703,
      "num_tokens": 55720327.0,
      "step": 25430
    },
    {
      "entropy": 5.558931303024292,
      "epoch": 2.524315204446209,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004366018395585201,
      "loss": 4.9117,
      "mean_token_accuracy": 0.18804533779621124,
      "num_tokens": 55730727.0,
      "step": 25435
    },
    {
      "entropy": 5.669615983963013,
      "epoch": 2.5248114331083764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004365771841301323,
      "loss": 5.0664,
      "mean_token_accuracy": 0.18514263331890107,
      "num_tokens": 55741568.0,
      "step": 25440
    },
    {
      "entropy": 5.585062456130982,
      "epoch": 2.525307661770544,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004365525246949786,
      "loss": 5.014,
      "mean_token_accuracy": 0.18487314134836197,
      "num_tokens": 55752511.0,
      "step": 25445
    },
    {
      "entropy": 5.637103843688965,
      "epoch": 2.5258038904327114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043652786125367035,
      "loss": 4.9886,
      "mean_token_accuracy": 0.18213440030813216,
      "num_tokens": 55763004.0,
      "step": 25450
    },
    {
      "entropy": 5.648900032043457,
      "epoch": 2.526300119094879,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043650319380681915,
      "loss": 5.0585,
      "mean_token_accuracy": 0.18351855576038362,
      "num_tokens": 55773276.0,
      "step": 25455
    },
    {
      "entropy": 5.683533239364624,
      "epoch": 2.5267963477570463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004364785223550368,
      "loss": 5.0596,
      "mean_token_accuracy": 0.18014211356639862,
      "num_tokens": 55784607.0,
      "step": 25460
    },
    {
      "entropy": 5.6929590702056885,
      "epoch": 2.527292576419214,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043645384689893504,
      "loss": 5.0745,
      "mean_token_accuracy": 0.17320135235786438,
      "num_tokens": 55796730.0,
      "step": 25465
    },
    {
      "entropy": 5.700607347488403,
      "epoch": 2.5277888050813813,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004364291674391258,
      "loss": 5.0714,
      "mean_token_accuracy": 0.176780167222023,
      "num_tokens": 55806940.0,
      "step": 25470
    },
    {
      "entropy": 5.609504079818725,
      "epoch": 2.528285033743549,
      "grad_norm": 1.046875,
      "learning_rate": 0.000436404483976221,
      "loss": 5.0011,
      "mean_token_accuracy": 0.18431168049573898,
      "num_tokens": 55818229.0,
      "step": 25475
    },
    {
      "entropy": 5.622170114517212,
      "epoch": 2.5287812624057167,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043637979651083285,
      "loss": 5.0896,
      "mean_token_accuracy": 0.18114383071660994,
      "num_tokens": 55831100.0,
      "step": 25480
    },
    {
      "entropy": 5.6600799560546875,
      "epoch": 2.529277491067884,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043635510504357343,
      "loss": 5.0299,
      "mean_token_accuracy": 0.17969617992639542,
      "num_tokens": 55841713.0,
      "step": 25485
    },
    {
      "entropy": 5.620818996429444,
      "epoch": 2.5297737197300516,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043633040957505515,
      "loss": 5.0115,
      "mean_token_accuracy": 0.17211297452449797,
      "num_tokens": 55852737.0,
      "step": 25490
    },
    {
      "entropy": 5.675095939636231,
      "epoch": 2.530269948392219,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004363057101058903,
      "loss": 5.0989,
      "mean_token_accuracy": 0.1824176698923111,
      "num_tokens": 55863471.0,
      "step": 25495
    },
    {
      "entropy": 5.5972353458404545,
      "epoch": 2.5307661770543866,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004362810066366915,
      "loss": 4.9751,
      "mean_token_accuracy": 0.18499550968408585,
      "num_tokens": 55874305.0,
      "step": 25500
    },
    {
      "entropy": 5.677726888656617,
      "epoch": 2.531262405716554,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043625629916807117,
      "loss": 5.1124,
      "mean_token_accuracy": 0.17984167039394378,
      "num_tokens": 55887400.0,
      "step": 25505
    },
    {
      "entropy": 5.591338205337524,
      "epoch": 2.531758634378722,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00043623158770064214,
      "loss": 5.0454,
      "mean_token_accuracy": 0.18106629848480224,
      "num_tokens": 55900052.0,
      "step": 25510
    },
    {
      "entropy": 5.767742681503296,
      "epoch": 2.5322548630408894,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004362068722350172,
      "loss": 5.1698,
      "mean_token_accuracy": 0.172431767731905,
      "num_tokens": 55911609.0,
      "step": 25515
    },
    {
      "entropy": 5.657179737091065,
      "epoch": 2.532751091703057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043618215277180917,
      "loss": 4.9839,
      "mean_token_accuracy": 0.1858830615878105,
      "num_tokens": 55922873.0,
      "step": 25520
    },
    {
      "entropy": 5.638728189468384,
      "epoch": 2.5332473203652244,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043615742931163106,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18053706884384155,
      "num_tokens": 55932161.0,
      "step": 25525
    },
    {
      "entropy": 5.6656616687774655,
      "epoch": 2.533743549027392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000436132701855096,
      "loss": 5.1575,
      "mean_token_accuracy": 0.1713552549481392,
      "num_tokens": 55943601.0,
      "step": 25530
    },
    {
      "entropy": 5.581442499160767,
      "epoch": 2.5342397776895593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043610797040281705,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18206181079149247,
      "num_tokens": 55954981.0,
      "step": 25535
    },
    {
      "entropy": 5.644953107833862,
      "epoch": 2.534736006351727,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004360832349554077,
      "loss": 5.0888,
      "mean_token_accuracy": 0.17933254688978195,
      "num_tokens": 55964407.0,
      "step": 25540
    },
    {
      "entropy": 5.59496955871582,
      "epoch": 2.5352322350138943,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004360584955134811,
      "loss": 4.9851,
      "mean_token_accuracy": 0.1791578233242035,
      "num_tokens": 55974634.0,
      "step": 25545
    },
    {
      "entropy": 5.672387552261353,
      "epoch": 2.5357284636760618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043603375207765103,
      "loss": 5.0466,
      "mean_token_accuracy": 0.17299692928791047,
      "num_tokens": 55986753.0,
      "step": 25550
    },
    {
      "entropy": 5.65263819694519,
      "epoch": 2.5362246923382292,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004360090046485308,
      "loss": 5.0251,
      "mean_token_accuracy": 0.18356794863939285,
      "num_tokens": 55997839.0,
      "step": 25555
    },
    {
      "entropy": 5.643859720230102,
      "epoch": 2.536720921000397,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004359842532267342,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18795090019702912,
      "num_tokens": 56008765.0,
      "step": 25560
    },
    {
      "entropy": 5.619660425186157,
      "epoch": 2.5372171496625646,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043595949781287505,
      "loss": 5.035,
      "mean_token_accuracy": 0.18721595853567125,
      "num_tokens": 56020077.0,
      "step": 25565
    },
    {
      "entropy": 5.64134669303894,
      "epoch": 2.537713378324732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043593473840756715,
      "loss": 5.0233,
      "mean_token_accuracy": 0.1883018732070923,
      "num_tokens": 56031298.0,
      "step": 25570
    },
    {
      "entropy": 5.701592874526978,
      "epoch": 2.5382096069868996,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004359099750114245,
      "loss": 5.0729,
      "mean_token_accuracy": 0.17969880998134613,
      "num_tokens": 56043955.0,
      "step": 25575
    },
    {
      "entropy": 5.64795937538147,
      "epoch": 2.538705835649067,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004358852076250613,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18509970903396605,
      "num_tokens": 56055734.0,
      "step": 25580
    },
    {
      "entropy": 5.699804210662842,
      "epoch": 2.5392020643112345,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043586043624909153,
      "loss": 5.0508,
      "mean_token_accuracy": 0.18420817852020263,
      "num_tokens": 56065472.0,
      "step": 25585
    },
    {
      "entropy": 5.745367479324341,
      "epoch": 2.539698292973402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004358356608841296,
      "loss": 5.0625,
      "mean_token_accuracy": 0.1836368754506111,
      "num_tokens": 56075691.0,
      "step": 25590
    },
    {
      "entropy": 5.631707429885864,
      "epoch": 2.54019452163557,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00043581088153078984,
      "loss": 5.028,
      "mean_token_accuracy": 0.18636296838521957,
      "num_tokens": 56085900.0,
      "step": 25595
    },
    {
      "entropy": 5.636427307128907,
      "epoch": 2.5406907502977374,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043578609818968676,
      "loss": 5.0228,
      "mean_token_accuracy": 0.1841539815068245,
      "num_tokens": 56097675.0,
      "step": 25600
    },
    {
      "entropy": 5.7002862930297855,
      "epoch": 2.541186978959905,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043576131086143496,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17527346163988114,
      "num_tokens": 56108102.0,
      "step": 25605
    },
    {
      "entropy": 5.562301969528198,
      "epoch": 2.5416832076220723,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043573651954664897,
      "loss": 4.9342,
      "mean_token_accuracy": 0.1906988129019737,
      "num_tokens": 56118489.0,
      "step": 25610
    },
    {
      "entropy": 5.577757120132446,
      "epoch": 2.54217943628424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043571172424594363,
      "loss": 4.9368,
      "mean_token_accuracy": 0.19454247802495955,
      "num_tokens": 56127949.0,
      "step": 25615
    },
    {
      "entropy": 5.619501829147339,
      "epoch": 2.5426756649464073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004356869249599339,
      "loss": 4.9916,
      "mean_token_accuracy": 0.1846037521958351,
      "num_tokens": 56138209.0,
      "step": 25620
    },
    {
      "entropy": 5.636900234222412,
      "epoch": 2.5431718936085748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004356621216892347,
      "loss": 5.0625,
      "mean_token_accuracy": 0.18452791422605513,
      "num_tokens": 56149005.0,
      "step": 25625
    },
    {
      "entropy": 5.630761575698853,
      "epoch": 2.5436681222707422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043563731443446095,
      "loss": 4.9848,
      "mean_token_accuracy": 0.179754638671875,
      "num_tokens": 56159406.0,
      "step": 25630
    },
    {
      "entropy": 5.633305549621582,
      "epoch": 2.5441643509329097,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043561250319622806,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18430942595005034,
      "num_tokens": 56169998.0,
      "step": 25635
    },
    {
      "entropy": 5.725721073150635,
      "epoch": 2.544660579595077,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043558768797515116,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1753906562924385,
      "num_tokens": 56181096.0,
      "step": 25640
    },
    {
      "entropy": 5.66195158958435,
      "epoch": 2.5451568082572447,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043556286877184566,
      "loss": 5.0454,
      "mean_token_accuracy": 0.17976278364658355,
      "num_tokens": 56191909.0,
      "step": 25645
    },
    {
      "entropy": 5.651175451278687,
      "epoch": 2.5456530369194126,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004355380455869269,
      "loss": 5.0649,
      "mean_token_accuracy": 0.1769494578242302,
      "num_tokens": 56202725.0,
      "step": 25650
    },
    {
      "entropy": 5.645940351486206,
      "epoch": 2.54614926558158,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004355132184210106,
      "loss": 5.0318,
      "mean_token_accuracy": 0.1846806064248085,
      "num_tokens": 56213596.0,
      "step": 25655
    },
    {
      "entropy": 5.613961219787598,
      "epoch": 2.5466454942437475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004354883872747124,
      "loss": 5.0378,
      "mean_token_accuracy": 0.18113708049058913,
      "num_tokens": 56224416.0,
      "step": 25660
    },
    {
      "entropy": 5.6164850234985355,
      "epoch": 2.547141722905915,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043546355214864796,
      "loss": 5.0242,
      "mean_token_accuracy": 0.18133242577314376,
      "num_tokens": 56234325.0,
      "step": 25665
    },
    {
      "entropy": 5.62465500831604,
      "epoch": 2.5476379515680825,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004354387130434332,
      "loss": 5.0162,
      "mean_token_accuracy": 0.18415064066648484,
      "num_tokens": 56244178.0,
      "step": 25670
    },
    {
      "entropy": 5.639575481414795,
      "epoch": 2.54813418023025,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004354138699596841,
      "loss": 5.0647,
      "mean_token_accuracy": 0.18030501902103424,
      "num_tokens": 56256036.0,
      "step": 25675
    },
    {
      "entropy": 5.635693550109863,
      "epoch": 2.548630408892418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004353890228980167,
      "loss": 5.0221,
      "mean_token_accuracy": 0.18033966422080994,
      "num_tokens": 56266203.0,
      "step": 25680
    },
    {
      "entropy": 5.639993047714233,
      "epoch": 2.5491266375545854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004353641718590471,
      "loss": 5.0441,
      "mean_token_accuracy": 0.18247295618057252,
      "num_tokens": 56276958.0,
      "step": 25685
    },
    {
      "entropy": 5.657866764068603,
      "epoch": 2.549622866216753,
      "grad_norm": 1.0,
      "learning_rate": 0.00043533931684339156,
      "loss": 5.0612,
      "mean_token_accuracy": 0.18664776682853698,
      "num_tokens": 56287941.0,
      "step": 25690
    },
    {
      "entropy": 5.609455013275147,
      "epoch": 2.5501190948789203,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004353144578516665,
      "loss": 4.9697,
      "mean_token_accuracy": 0.18172425776720047,
      "num_tokens": 56299390.0,
      "step": 25695
    },
    {
      "entropy": 5.629914855957031,
      "epoch": 2.5506153235410878,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043528959488448835,
      "loss": 5.0034,
      "mean_token_accuracy": 0.18136681616306305,
      "num_tokens": 56310762.0,
      "step": 25700
    },
    {
      "entropy": 5.666426944732666,
      "epoch": 2.5511115522032553,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043526472794247365,
      "loss": 5.0031,
      "mean_token_accuracy": 0.1859482005238533,
      "num_tokens": 56322522.0,
      "step": 25705
    },
    {
      "entropy": 5.619450569152832,
      "epoch": 2.5516077808654227,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043523985702623896,
      "loss": 5.0722,
      "mean_token_accuracy": 0.18227875381708145,
      "num_tokens": 56334124.0,
      "step": 25710
    },
    {
      "entropy": 5.600906038284302,
      "epoch": 2.55210400952759,
      "grad_norm": 1.25,
      "learning_rate": 0.00043521498213640113,
      "loss": 5.0175,
      "mean_token_accuracy": 0.1831234723329544,
      "num_tokens": 56344761.0,
      "step": 25715
    },
    {
      "entropy": 5.597951602935791,
      "epoch": 2.5526002381897577,
      "grad_norm": 1.0,
      "learning_rate": 0.00043519010327357704,
      "loss": 4.9958,
      "mean_token_accuracy": 0.182628233730793,
      "num_tokens": 56355518.0,
      "step": 25720
    },
    {
      "entropy": 5.652153730392456,
      "epoch": 2.553096466851925,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004351652204383836,
      "loss": 5.0828,
      "mean_token_accuracy": 0.17898668944835663,
      "num_tokens": 56366307.0,
      "step": 25725
    },
    {
      "entropy": 5.6664611339569095,
      "epoch": 2.5535926955140926,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004351403336314377,
      "loss": 4.984,
      "mean_token_accuracy": 0.18949312269687651,
      "num_tokens": 56377065.0,
      "step": 25730
    },
    {
      "entropy": 5.675796222686768,
      "epoch": 2.5540889241762605,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351154428533566,
      "loss": 5.0381,
      "mean_token_accuracy": 0.18632512241601945,
      "num_tokens": 56387741.0,
      "step": 25735
    },
    {
      "entropy": 5.648118162155152,
      "epoch": 2.554585152838428,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004350905481047576,
      "loss": 5.2042,
      "mean_token_accuracy": 0.17380053997039796,
      "num_tokens": 56398927.0,
      "step": 25740
    },
    {
      "entropy": 5.741234350204468,
      "epoch": 2.5550813815005955,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004350656493862579,
      "loss": 5.172,
      "mean_token_accuracy": 0.17594490796327591,
      "num_tokens": 56408791.0,
      "step": 25745
    },
    {
      "entropy": 5.69065899848938,
      "epoch": 2.555577610162763,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043504074669847517,
      "loss": 4.9766,
      "mean_token_accuracy": 0.188413405418396,
      "num_tokens": 56419206.0,
      "step": 25750
    },
    {
      "entropy": 5.733413362503052,
      "epoch": 2.5560738388249304,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004350158400420266,
      "loss": 5.1533,
      "mean_token_accuracy": 0.17249055653810502,
      "num_tokens": 56431774.0,
      "step": 25755
    },
    {
      "entropy": 5.7044844150543215,
      "epoch": 2.556570067487098,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043499092941753003,
      "loss": 5.0442,
      "mean_token_accuracy": 0.1845347046852112,
      "num_tokens": 56442048.0,
      "step": 25760
    },
    {
      "entropy": 5.640010261535645,
      "epoch": 2.557066296149266,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043496601482560327,
      "loss": 5.0274,
      "mean_token_accuracy": 0.1767486110329628,
      "num_tokens": 56453347.0,
      "step": 25765
    },
    {
      "entropy": 5.596910381317139,
      "epoch": 2.5575625248114333,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043494109626686394,
      "loss": 5.0054,
      "mean_token_accuracy": 0.19169043153524398,
      "num_tokens": 56465323.0,
      "step": 25770
    },
    {
      "entropy": 5.636179065704345,
      "epoch": 2.558058753473601,
      "grad_norm": 1.21875,
      "learning_rate": 0.00043491617374193017,
      "loss": 5.0545,
      "mean_token_accuracy": 0.1805521711707115,
      "num_tokens": 56474704.0,
      "step": 25775
    },
    {
      "entropy": 5.634581279754639,
      "epoch": 2.5585549821357683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004348912472514197,
      "loss": 5.0662,
      "mean_token_accuracy": 0.18249435871839523,
      "num_tokens": 56485028.0,
      "step": 25780
    },
    {
      "entropy": 5.643368196487427,
      "epoch": 2.5590512107979357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043486631679595097,
      "loss": 4.9663,
      "mean_token_accuracy": 0.18745721578598024,
      "num_tokens": 56496169.0,
      "step": 25785
    },
    {
      "entropy": 5.585866641998291,
      "epoch": 2.559547439460103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004348413823761421,
      "loss": 4.986,
      "mean_token_accuracy": 0.1867930293083191,
      "num_tokens": 56506324.0,
      "step": 25790
    },
    {
      "entropy": 5.570674419403076,
      "epoch": 2.5600436681222707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004348164439926112,
      "loss": 4.9338,
      "mean_token_accuracy": 0.1878335401415825,
      "num_tokens": 56516901.0,
      "step": 25795
    },
    {
      "entropy": 5.583589839935303,
      "epoch": 2.560539896784438,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043479150164597695,
      "loss": 4.9342,
      "mean_token_accuracy": 0.1870908886194229,
      "num_tokens": 56527241.0,
      "step": 25800
    },
    {
      "entropy": 5.680503416061401,
      "epoch": 2.5610361254466056,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043476655533685783,
      "loss": 5.1368,
      "mean_token_accuracy": 0.18157754242420196,
      "num_tokens": 56539078.0,
      "step": 25805
    },
    {
      "entropy": 5.636449861526489,
      "epoch": 2.561532354108773,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043474160506587234,
      "loss": 4.9883,
      "mean_token_accuracy": 0.18498046696186066,
      "num_tokens": 56551141.0,
      "step": 25810
    },
    {
      "entropy": 5.661978626251221,
      "epoch": 2.5620285827709406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004347166508336393,
      "loss": 5.0989,
      "mean_token_accuracy": 0.17896370738744735,
      "num_tokens": 56561836.0,
      "step": 25815
    },
    {
      "entropy": 5.677474689483643,
      "epoch": 2.5625248114331085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043469169264077754,
      "loss": 5.0167,
      "mean_token_accuracy": 0.184000401198864,
      "num_tokens": 56572699.0,
      "step": 25820
    },
    {
      "entropy": 5.676953554153442,
      "epoch": 2.563021040095276,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004346667304879058,
      "loss": 5.0977,
      "mean_token_accuracy": 0.1774973213672638,
      "num_tokens": 56582296.0,
      "step": 25825
    },
    {
      "entropy": 5.631569147109985,
      "epoch": 2.5635172687574435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043464176437564317,
      "loss": 5.031,
      "mean_token_accuracy": 0.18099183291196824,
      "num_tokens": 56592907.0,
      "step": 25830
    },
    {
      "entropy": 5.6172881603240965,
      "epoch": 2.564013497419611,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004346167943046089,
      "loss": 5.0063,
      "mean_token_accuracy": 0.1865838274359703,
      "num_tokens": 56604995.0,
      "step": 25835
    },
    {
      "entropy": 5.635041236877441,
      "epoch": 2.5645097260817784,
      "grad_norm": 1.03125,
      "learning_rate": 0.000434591820275422,
      "loss": 4.9862,
      "mean_token_accuracy": 0.1853833332657814,
      "num_tokens": 56614809.0,
      "step": 25840
    },
    {
      "entropy": 5.672575902938843,
      "epoch": 2.565005954743946,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004345668422887018,
      "loss": 5.0606,
      "mean_token_accuracy": 0.17765487879514694,
      "num_tokens": 56624409.0,
      "step": 25845
    },
    {
      "entropy": 5.62786717414856,
      "epoch": 2.565502183406114,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004345418603450678,
      "loss": 5.0583,
      "mean_token_accuracy": 0.17620272636413575,
      "num_tokens": 56634596.0,
      "step": 25850
    },
    {
      "entropy": 5.6381744861602785,
      "epoch": 2.5659984120682813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004345168744451395,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18624392449855803,
      "num_tokens": 56646710.0,
      "step": 25855
    },
    {
      "entropy": 5.632579040527344,
      "epoch": 2.5664946407304488,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004344918845895364,
      "loss": 5.0154,
      "mean_token_accuracy": 0.18470106720924379,
      "num_tokens": 56658427.0,
      "step": 25860
    },
    {
      "entropy": 5.664536809921264,
      "epoch": 2.5669908693926162,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004344668907788782,
      "loss": 5.093,
      "mean_token_accuracy": 0.17861920595169067,
      "num_tokens": 56669039.0,
      "step": 25865
    },
    {
      "entropy": 5.64415397644043,
      "epoch": 2.5674870980547837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004344418930137848,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18996705263853073,
      "num_tokens": 56679063.0,
      "step": 25870
    },
    {
      "entropy": 5.6814511775970455,
      "epoch": 2.567983326716951,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004344168912948759,
      "loss": 5.0308,
      "mean_token_accuracy": 0.18244960010051728,
      "num_tokens": 56690142.0,
      "step": 25875
    },
    {
      "entropy": 5.634204149246216,
      "epoch": 2.5684795553791187,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043439188562277174,
      "loss": 5.0603,
      "mean_token_accuracy": 0.1811402454972267,
      "num_tokens": 56700918.0,
      "step": 25880
    },
    {
      "entropy": 5.78638482093811,
      "epoch": 2.568975784041286,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043436687599809226,
      "loss": 5.1816,
      "mean_token_accuracy": 0.17172955423593522,
      "num_tokens": 56713520.0,
      "step": 25885
    },
    {
      "entropy": 5.6558667659759525,
      "epoch": 2.5694720127034536,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043434186242145763,
      "loss": 5.0707,
      "mean_token_accuracy": 0.17861161828041078,
      "num_tokens": 56723428.0,
      "step": 25890
    },
    {
      "entropy": 5.657552194595337,
      "epoch": 2.569968241365621,
      "grad_norm": 1.0,
      "learning_rate": 0.0004343168448934881,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18279364854097366,
      "num_tokens": 56734167.0,
      "step": 25895
    },
    {
      "entropy": 5.6661664009094235,
      "epoch": 2.5704644700277886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043429182341480423,
      "loss": 4.9688,
      "mean_token_accuracy": 0.18805532455444335,
      "num_tokens": 56745196.0,
      "step": 25900
    },
    {
      "entropy": 5.622960424423217,
      "epoch": 2.5709606986899565,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004342667979860263,
      "loss": 4.973,
      "mean_token_accuracy": 0.18052152544260025,
      "num_tokens": 56756166.0,
      "step": 25905
    },
    {
      "entropy": 5.603186082839966,
      "epoch": 2.571456927352124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043424176860777495,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18063803315162658,
      "num_tokens": 56766690.0,
      "step": 25910
    },
    {
      "entropy": 5.634601163864136,
      "epoch": 2.5719531560142914,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043421673528067095,
      "loss": 5.0574,
      "mean_token_accuracy": 0.1793111503124237,
      "num_tokens": 56776995.0,
      "step": 25915
    },
    {
      "entropy": 5.724631690979004,
      "epoch": 2.572449384676459,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043419169800533496,
      "loss": 5.1211,
      "mean_token_accuracy": 0.18054128736257552,
      "num_tokens": 56788509.0,
      "step": 25920
    },
    {
      "entropy": 5.665227794647217,
      "epoch": 2.5729456133386264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004341666567823879,
      "loss": 5.0941,
      "mean_token_accuracy": 0.170925535261631,
      "num_tokens": 56799027.0,
      "step": 25925
    },
    {
      "entropy": 5.695346164703369,
      "epoch": 2.573441842000794,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004341416116124508,
      "loss": 5.0711,
      "mean_token_accuracy": 0.18128749579191208,
      "num_tokens": 56809796.0,
      "step": 25930
    },
    {
      "entropy": 5.696785879135132,
      "epoch": 2.5739380706629618,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004341165624961445,
      "loss": 5.0759,
      "mean_token_accuracy": 0.17429264932870864,
      "num_tokens": 56820523.0,
      "step": 25935
    },
    {
      "entropy": 5.649970960617066,
      "epoch": 2.5744342993251292,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043409150943409035,
      "loss": 5.0364,
      "mean_token_accuracy": 0.1786736860871315,
      "num_tokens": 56831668.0,
      "step": 25940
    },
    {
      "entropy": 5.572697973251342,
      "epoch": 2.5749305279872967,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043406645242690967,
      "loss": 4.9384,
      "mean_token_accuracy": 0.18405273854732512,
      "num_tokens": 56841950.0,
      "step": 25945
    },
    {
      "entropy": 5.617842388153076,
      "epoch": 2.575426756649464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004340413914752237,
      "loss": 5.0647,
      "mean_token_accuracy": 0.1772662118077278,
      "num_tokens": 56852190.0,
      "step": 25950
    },
    {
      "entropy": 5.71850438117981,
      "epoch": 2.5759229853116317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004340163265796539,
      "loss": 5.0718,
      "mean_token_accuracy": 0.1801753044128418,
      "num_tokens": 56864255.0,
      "step": 25955
    },
    {
      "entropy": 5.681293725967407,
      "epoch": 2.576419213973799,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004339912577408218,
      "loss": 4.9981,
      "mean_token_accuracy": 0.18487872928380966,
      "num_tokens": 56874711.0,
      "step": 25960
    },
    {
      "entropy": 5.559211015701294,
      "epoch": 2.5769154426359666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004339661849593492,
      "loss": 4.9749,
      "mean_token_accuracy": 0.1851196989417076,
      "num_tokens": 56885790.0,
      "step": 25965
    },
    {
      "entropy": 5.639458894729614,
      "epoch": 2.577411671298134,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004339411082358577,
      "loss": 5.0426,
      "mean_token_accuracy": 0.18352247029542923,
      "num_tokens": 56896452.0,
      "step": 25970
    },
    {
      "entropy": 5.6615242004394535,
      "epoch": 2.5779078999603016,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043391602757096923,
      "loss": 5.1531,
      "mean_token_accuracy": 0.17944886088371276,
      "num_tokens": 56908216.0,
      "step": 25975
    },
    {
      "entropy": 5.692090368270874,
      "epoch": 2.578404128622469,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043389094296530565,
      "loss": 5.0405,
      "mean_token_accuracy": 0.1722586825489998,
      "num_tokens": 56918346.0,
      "step": 25980
    },
    {
      "entropy": 5.663885545730591,
      "epoch": 2.5789003572846365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004338658544194891,
      "loss": 5.1491,
      "mean_token_accuracy": 0.1706255167722702,
      "num_tokens": 56930950.0,
      "step": 25985
    },
    {
      "entropy": 5.681478643417359,
      "epoch": 2.5793965859468044,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043384076193414176,
      "loss": 5.0749,
      "mean_token_accuracy": 0.18179222345352172,
      "num_tokens": 56940622.0,
      "step": 25990
    },
    {
      "entropy": 5.645621013641358,
      "epoch": 2.579892814608972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043381566550988564,
      "loss": 5.0315,
      "mean_token_accuracy": 0.18317112922668458,
      "num_tokens": 56952463.0,
      "step": 25995
    },
    {
      "entropy": 5.6470763206481935,
      "epoch": 2.5803890432711394,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043379056514734334,
      "loss": 4.9843,
      "mean_token_accuracy": 0.18488092571496964,
      "num_tokens": 56961705.0,
      "step": 26000
    },
    {
      "entropy": 5.662886476516723,
      "epoch": 2.580885271933307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004337654608471372,
      "loss": 5.0607,
      "mean_token_accuracy": 0.17753274887800216,
      "num_tokens": 56972359.0,
      "step": 26005
    },
    {
      "entropy": 5.669077777862549,
      "epoch": 2.5813815005954743,
      "grad_norm": 1.125,
      "learning_rate": 0.00043374035260988964,
      "loss": 5.0076,
      "mean_token_accuracy": 0.18338484317064285,
      "num_tokens": 56983342.0,
      "step": 26010
    },
    {
      "entropy": 5.592480897903442,
      "epoch": 2.581877729257642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043371524043622344,
      "loss": 4.9996,
      "mean_token_accuracy": 0.1889732539653778,
      "num_tokens": 56993859.0,
      "step": 26015
    },
    {
      "entropy": 5.675951528549194,
      "epoch": 2.5823739579198093,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004336901243267614,
      "loss": 5.1135,
      "mean_token_accuracy": 0.1785951718688011,
      "num_tokens": 57006631.0,
      "step": 26020
    },
    {
      "entropy": 5.694146728515625,
      "epoch": 2.582870186581977,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000433665004282126,
      "loss": 5.0597,
      "mean_token_accuracy": 0.1774393081665039,
      "num_tokens": 57018121.0,
      "step": 26025
    },
    {
      "entropy": 5.634240293502808,
      "epoch": 2.5833664152441447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043363988030294055,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18155599236488343,
      "num_tokens": 57028127.0,
      "step": 26030
    },
    {
      "entropy": 5.635014247894287,
      "epoch": 2.583862643906312,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004336147523898279,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19395689517259598,
      "num_tokens": 57038523.0,
      "step": 26035
    },
    {
      "entropy": 5.618374490737915,
      "epoch": 2.5843588725684796,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004335896205434111,
      "loss": 4.9941,
      "mean_token_accuracy": 0.18431930094957352,
      "num_tokens": 57048641.0,
      "step": 26040
    },
    {
      "entropy": 5.582756185531617,
      "epoch": 2.584855101230647,
      "grad_norm": 1.15625,
      "learning_rate": 0.00043356448476431354,
      "loss": 4.917,
      "mean_token_accuracy": 0.18736081719398498,
      "num_tokens": 57058794.0,
      "step": 26045
    },
    {
      "entropy": 5.6538036346435545,
      "epoch": 2.5853513298928146,
      "grad_norm": 1.125,
      "learning_rate": 0.0004335393450531583,
      "loss": 5.0413,
      "mean_token_accuracy": 0.18238836526870728,
      "num_tokens": 57069814.0,
      "step": 26050
    },
    {
      "entropy": 5.755262088775635,
      "epoch": 2.585847558554982,
      "grad_norm": 1.171875,
      "learning_rate": 0.000433514201410569,
      "loss": 5.1139,
      "mean_token_accuracy": 0.18461893498897552,
      "num_tokens": 57081053.0,
      "step": 26055
    },
    {
      "entropy": 5.622925853729248,
      "epoch": 2.5863437872171495,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043348905383716905,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18461842983961105,
      "num_tokens": 57091386.0,
      "step": 26060
    },
    {
      "entropy": 5.64370174407959,
      "epoch": 2.586840015879317,
      "grad_norm": 1.234375,
      "learning_rate": 0.00043346390233358213,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18329289108514785,
      "num_tokens": 57102659.0,
      "step": 26065
    },
    {
      "entropy": 5.6963153839111325,
      "epoch": 2.5873362445414845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004334387469004319,
      "loss": 5.0628,
      "mean_token_accuracy": 0.18244686126708984,
      "num_tokens": 57114257.0,
      "step": 26070
    },
    {
      "entropy": 5.633910274505615,
      "epoch": 2.5878324732036524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004334135875383421,
      "loss": 5.0732,
      "mean_token_accuracy": 0.18235713839530945,
      "num_tokens": 57125037.0,
      "step": 26075
    },
    {
      "entropy": 5.592832374572754,
      "epoch": 2.58832870186582,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004333884242479367,
      "loss": 5.0448,
      "mean_token_accuracy": 0.17693111002445222,
      "num_tokens": 57137581.0,
      "step": 26080
    },
    {
      "entropy": 5.619929599761963,
      "epoch": 2.5888249305279873,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043336325702983966,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18810765594244003,
      "num_tokens": 57148507.0,
      "step": 26085
    },
    {
      "entropy": 5.622819471359253,
      "epoch": 2.589321159190155,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043333808588467504,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19068922847509384,
      "num_tokens": 57159518.0,
      "step": 26090
    },
    {
      "entropy": 5.61540789604187,
      "epoch": 2.5898173878523223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004333129108130671,
      "loss": 4.9955,
      "mean_token_accuracy": 0.18673209249973297,
      "num_tokens": 57170600.0,
      "step": 26095
    },
    {
      "entropy": 5.652515602111817,
      "epoch": 2.5903136165144898,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004332877318156402,
      "loss": 5.0263,
      "mean_token_accuracy": 0.18084889501333237,
      "num_tokens": 57181137.0,
      "step": 26100
    },
    {
      "entropy": 5.643892049789429,
      "epoch": 2.5908098451766572,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043326254889301856,
      "loss": 5.0298,
      "mean_token_accuracy": 0.17658105492591858,
      "num_tokens": 57191975.0,
      "step": 26105
    },
    {
      "entropy": 5.716372966766357,
      "epoch": 2.591306073838825,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004332373620458268,
      "loss": 5.0709,
      "mean_token_accuracy": 0.17599261850118636,
      "num_tokens": 57203139.0,
      "step": 26110
    },
    {
      "entropy": 5.654077196121216,
      "epoch": 2.5918023025009926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043321217127468935,
      "loss": 5.1045,
      "mean_token_accuracy": 0.1728320062160492,
      "num_tokens": 57215542.0,
      "step": 26115
    },
    {
      "entropy": 5.647805023193359,
      "epoch": 2.59229853116316,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043318697658023097,
      "loss": 5.0241,
      "mean_token_accuracy": 0.18137607872486114,
      "num_tokens": 57226525.0,
      "step": 26120
    },
    {
      "entropy": 5.646157932281494,
      "epoch": 2.5927947598253276,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004331617779630765,
      "loss": 4.9768,
      "mean_token_accuracy": 0.1927406147122383,
      "num_tokens": 57236120.0,
      "step": 26125
    },
    {
      "entropy": 5.565434980392456,
      "epoch": 2.593290988487495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331365754238507,
      "loss": 4.9499,
      "mean_token_accuracy": 0.18519456535577775,
      "num_tokens": 57246947.0,
      "step": 26130
    },
    {
      "entropy": 5.6310975551605225,
      "epoch": 2.5937872171496625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043311136896317855,
      "loss": 4.9839,
      "mean_token_accuracy": 0.18336691260337828,
      "num_tokens": 57258885.0,
      "step": 26135
    },
    {
      "entropy": 5.7012592315673825,
      "epoch": 2.59428344581183,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043308615858168517,
      "loss": 5.0637,
      "mean_token_accuracy": 0.1844007521867752,
      "num_tokens": 57270047.0,
      "step": 26140
    },
    {
      "entropy": 5.691552639007568,
      "epoch": 2.5947796744739975,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004330609442799957,
      "loss": 5.1217,
      "mean_token_accuracy": 0.18058059066534043,
      "num_tokens": 57281405.0,
      "step": 26145
    },
    {
      "entropy": 5.660006284713745,
      "epoch": 2.595275903136165,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004330357260587355,
      "loss": 5.0417,
      "mean_token_accuracy": 0.18622582107782365,
      "num_tokens": 57292880.0,
      "step": 26150
    },
    {
      "entropy": 5.637996625900269,
      "epoch": 2.5957721317983324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004330105039185297,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18538569509983063,
      "num_tokens": 57304544.0,
      "step": 26155
    },
    {
      "entropy": 5.6544300556182865,
      "epoch": 2.5962683604605004,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043298527786000393,
      "loss": 4.994,
      "mean_token_accuracy": 0.18466636836528777,
      "num_tokens": 57315869.0,
      "step": 26160
    },
    {
      "entropy": 5.692496919631958,
      "epoch": 2.596764589122668,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004329600478837837,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18311440497636794,
      "num_tokens": 57327933.0,
      "step": 26165
    },
    {
      "entropy": 5.679098224639892,
      "epoch": 2.5972608177848353,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043293481399049474,
      "loss": 5.0274,
      "mean_token_accuracy": 0.1791093870997429,
      "num_tokens": 57338941.0,
      "step": 26170
    },
    {
      "entropy": 5.63222975730896,
      "epoch": 2.597757046447003,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004329095761807626,
      "loss": 5.0375,
      "mean_token_accuracy": 0.18258484154939653,
      "num_tokens": 57350128.0,
      "step": 26175
    },
    {
      "entropy": 5.595133924484253,
      "epoch": 2.5982532751091703,
      "grad_norm": 1.0,
      "learning_rate": 0.0004328843344552133,
      "loss": 5.0121,
      "mean_token_accuracy": 0.17979067116975783,
      "num_tokens": 57361602.0,
      "step": 26180
    },
    {
      "entropy": 5.736884260177613,
      "epoch": 2.5987495037713377,
      "grad_norm": 1.25,
      "learning_rate": 0.0004328590888144726,
      "loss": 5.1049,
      "mean_token_accuracy": 0.17369866520166397,
      "num_tokens": 57373093.0,
      "step": 26185
    },
    {
      "entropy": 5.7568145275115965,
      "epoch": 2.599245732433505,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004328338392591668,
      "loss": 5.1228,
      "mean_token_accuracy": 0.17312782108783722,
      "num_tokens": 57384799.0,
      "step": 26190
    },
    {
      "entropy": 5.611915874481201,
      "epoch": 2.599741961095673,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043280858578992185,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18939533084630966,
      "num_tokens": 57395703.0,
      "step": 26195
    },
    {
      "entropy": 5.709093999862671,
      "epoch": 2.6002381897578406,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043278332840736397,
      "loss": 5.1305,
      "mean_token_accuracy": 0.17944208681583404,
      "num_tokens": 57406725.0,
      "step": 26200
    },
    {
      "entropy": 5.622258472442627,
      "epoch": 2.600734418420008,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043275806711211957,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18749370872974397,
      "num_tokens": 57417362.0,
      "step": 26205
    },
    {
      "entropy": 5.65551233291626,
      "epoch": 2.6012306470821756,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043273280190481503,
      "loss": 5.0262,
      "mean_token_accuracy": 0.18915585279464722,
      "num_tokens": 57429178.0,
      "step": 26210
    },
    {
      "entropy": 5.651227045059204,
      "epoch": 2.601726875744343,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043270753278607687,
      "loss": 5.0176,
      "mean_token_accuracy": 0.18492424935102464,
      "num_tokens": 57439741.0,
      "step": 26215
    },
    {
      "entropy": 5.607196474075318,
      "epoch": 2.6022231044065105,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043268225975653176,
      "loss": 5.0865,
      "mean_token_accuracy": 0.1813482239842415,
      "num_tokens": 57451754.0,
      "step": 26220
    },
    {
      "entropy": 5.6826300621032715,
      "epoch": 2.602719333068678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004326569828168064,
      "loss": 5.0414,
      "mean_token_accuracy": 0.18199145942926406,
      "num_tokens": 57463396.0,
      "step": 26225
    },
    {
      "entropy": 5.6392618179321286,
      "epoch": 2.6032155617308455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004326317019675275,
      "loss": 5.0187,
      "mean_token_accuracy": 0.1871027961373329,
      "num_tokens": 57474023.0,
      "step": 26230
    },
    {
      "entropy": 5.6572428226470945,
      "epoch": 2.603711790393013,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043260641720932214,
      "loss": 5.0256,
      "mean_token_accuracy": 0.18744251877069473,
      "num_tokens": 57484751.0,
      "step": 26235
    },
    {
      "entropy": 5.60026707649231,
      "epoch": 2.6042080190551804,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043258112854281726,
      "loss": 4.993,
      "mean_token_accuracy": 0.18233634084463118,
      "num_tokens": 57496502.0,
      "step": 26240
    },
    {
      "entropy": 5.649193525314331,
      "epoch": 2.6047042477173483,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004325558359686399,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18895690143108368,
      "num_tokens": 57507111.0,
      "step": 26245
    },
    {
      "entropy": 5.64849066734314,
      "epoch": 2.605200476379516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043253053948741734,
      "loss": 5.1147,
      "mean_token_accuracy": 0.17526306062936783,
      "num_tokens": 57517844.0,
      "step": 26250
    },
    {
      "entropy": 5.68176760673523,
      "epoch": 2.6056967050416833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043250523909977684,
      "loss": 5.0587,
      "mean_token_accuracy": 0.18436448276042938,
      "num_tokens": 57528047.0,
      "step": 26255
    },
    {
      "entropy": 5.691463994979858,
      "epoch": 2.6061929337038507,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004324799348063457,
      "loss": 5.039,
      "mean_token_accuracy": 0.17958420664072036,
      "num_tokens": 57538489.0,
      "step": 26260
    },
    {
      "entropy": 5.622463703155518,
      "epoch": 2.606689162366018,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004324546266077518,
      "loss": 5.0348,
      "mean_token_accuracy": 0.18707305341959,
      "num_tokens": 57549741.0,
      "step": 26265
    },
    {
      "entropy": 5.610870504379273,
      "epoch": 2.6071853910281857,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004324293145046222,
      "loss": 4.9581,
      "mean_token_accuracy": 0.18681554198265077,
      "num_tokens": 57559803.0,
      "step": 26270
    },
    {
      "entropy": 5.677463674545288,
      "epoch": 2.607681619690353,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043240399849758484,
      "loss": 5.0276,
      "mean_token_accuracy": 0.18308836668729783,
      "num_tokens": 57570609.0,
      "step": 26275
    },
    {
      "entropy": 5.748792743682861,
      "epoch": 2.608177848352521,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004323786785872675,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18459445089101792,
      "num_tokens": 57580934.0,
      "step": 26280
    },
    {
      "entropy": 5.675896739959716,
      "epoch": 2.6086740770146886,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004323533547742981,
      "loss": 5.1522,
      "mean_token_accuracy": 0.17532729357481003,
      "num_tokens": 57594671.0,
      "step": 26285
    },
    {
      "entropy": 5.687403726577759,
      "epoch": 2.609170305676856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004323280270593046,
      "loss": 5.0802,
      "mean_token_accuracy": 0.1836544007062912,
      "num_tokens": 57605582.0,
      "step": 26290
    },
    {
      "entropy": 5.664662551879883,
      "epoch": 2.6096665343390235,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043230269544291503,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18508785367012023,
      "num_tokens": 57617185.0,
      "step": 26295
    },
    {
      "entropy": 5.624679899215698,
      "epoch": 2.610162763001191,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004322773599257575,
      "loss": 4.9768,
      "mean_token_accuracy": 0.18774369060993196,
      "num_tokens": 57627205.0,
      "step": 26300
    },
    {
      "entropy": 5.57074818611145,
      "epoch": 2.6106589916633585,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043225202050846043,
      "loss": 4.9784,
      "mean_token_accuracy": 0.18569996058940888,
      "num_tokens": 57639496.0,
      "step": 26305
    },
    {
      "entropy": 5.715506076812744,
      "epoch": 2.611155220325526,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000432226677191652,
      "loss": 5.0981,
      "mean_token_accuracy": 0.1738818570971489,
      "num_tokens": 57650345.0,
      "step": 26310
    },
    {
      "entropy": 5.669410467147827,
      "epoch": 2.6116514489876934,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043220132997596084,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18494480550289155,
      "num_tokens": 57661409.0,
      "step": 26315
    },
    {
      "entropy": 5.753959989547729,
      "epoch": 2.612147677649861,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004321759788620155,
      "loss": 5.1456,
      "mean_token_accuracy": 0.17658215910196304,
      "num_tokens": 57672639.0,
      "step": 26320
    },
    {
      "entropy": 5.663569164276123,
      "epoch": 2.6126439063120284,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043215062385044446,
      "loss": 5.1297,
      "mean_token_accuracy": 0.17571263313293456,
      "num_tokens": 57682249.0,
      "step": 26325
    },
    {
      "entropy": 5.5995183944702145,
      "epoch": 2.613140134974196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043212526494187665,
      "loss": 4.9492,
      "mean_token_accuracy": 0.1931204840540886,
      "num_tokens": 57692319.0,
      "step": 26330
    },
    {
      "entropy": 5.678122568130493,
      "epoch": 2.6136363636363638,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004320999021369408,
      "loss": 5.1181,
      "mean_token_accuracy": 0.17753924578428268,
      "num_tokens": 57702334.0,
      "step": 26335
    },
    {
      "entropy": 5.670973396301269,
      "epoch": 2.6141325922985312,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043207453543626587,
      "loss": 5.0133,
      "mean_token_accuracy": 0.18140793293714524,
      "num_tokens": 57712407.0,
      "step": 26340
    },
    {
      "entropy": 5.622265815734863,
      "epoch": 2.6146288209606987,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043204916484048094,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18446770906448365,
      "num_tokens": 57724252.0,
      "step": 26345
    },
    {
      "entropy": 5.656990385055542,
      "epoch": 2.615125049622866,
      "grad_norm": 1.125,
      "learning_rate": 0.00043202379035021515,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18410003781318665,
      "num_tokens": 57735033.0,
      "step": 26350
    },
    {
      "entropy": 5.732953786849976,
      "epoch": 2.6156212782850337,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004319984119660977,
      "loss": 5.0302,
      "mean_token_accuracy": 0.18613055646419524,
      "num_tokens": 57746182.0,
      "step": 26355
    },
    {
      "entropy": 5.651144790649414,
      "epoch": 2.616117506947201,
      "grad_norm": 1.125,
      "learning_rate": 0.00043197302968875795,
      "loss": 4.9877,
      "mean_token_accuracy": 0.18508316576480865,
      "num_tokens": 57756479.0,
      "step": 26360
    },
    {
      "entropy": 5.68727068901062,
      "epoch": 2.616613735609369,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004319476435188254,
      "loss": 5.0865,
      "mean_token_accuracy": 0.17718528658151628,
      "num_tokens": 57767773.0,
      "step": 26365
    },
    {
      "entropy": 5.673148441314697,
      "epoch": 2.6171099642715365,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043192225345692933,
      "loss": 4.9846,
      "mean_token_accuracy": 0.18770713955163956,
      "num_tokens": 57777331.0,
      "step": 26370
    },
    {
      "entropy": 5.613301134109497,
      "epoch": 2.617606192933704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043189685950369963,
      "loss": 5.0106,
      "mean_token_accuracy": 0.1840173050761223,
      "num_tokens": 57788538.0,
      "step": 26375
    },
    {
      "entropy": 5.695648813247681,
      "epoch": 2.6181024215958715,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043187146165976577,
      "loss": 5.15,
      "mean_token_accuracy": 0.18019113391637803,
      "num_tokens": 57799983.0,
      "step": 26380
    },
    {
      "entropy": 5.65128755569458,
      "epoch": 2.618598650258039,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004318460599257578,
      "loss": 4.9794,
      "mean_token_accuracy": 0.17844725996255875,
      "num_tokens": 57810195.0,
      "step": 26385
    },
    {
      "entropy": 5.60472002029419,
      "epoch": 2.6190948789202064,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043182065430230545,
      "loss": 4.9422,
      "mean_token_accuracy": 0.19175882041454315,
      "num_tokens": 57821930.0,
      "step": 26390
    },
    {
      "entropy": 5.6549217224121096,
      "epoch": 2.619591107582374,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004317952447900388,
      "loss": 4.979,
      "mean_token_accuracy": 0.18312352895736694,
      "num_tokens": 57834272.0,
      "step": 26395
    },
    {
      "entropy": 5.642524147033692,
      "epoch": 2.6200873362445414,
      "grad_norm": 1.046875,
      "learning_rate": 0.000431769831389588,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18448020070791243,
      "num_tokens": 57845491.0,
      "step": 26400
    },
    {
      "entropy": 5.634185934066773,
      "epoch": 2.620583564906709,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004317444141015831,
      "loss": 5.0914,
      "mean_token_accuracy": 0.18178408592939377,
      "num_tokens": 57857816.0,
      "step": 26405
    },
    {
      "entropy": 5.650193214416504,
      "epoch": 2.6210797935688763,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00043171899292665457,
      "loss": 5.0346,
      "mean_token_accuracy": 0.17939282953739166,
      "num_tokens": 57867769.0,
      "step": 26410
    },
    {
      "entropy": 5.72166223526001,
      "epoch": 2.621576022231044,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004316935678654327,
      "loss": 5.1254,
      "mean_token_accuracy": 0.17503389865159988,
      "num_tokens": 57878027.0,
      "step": 26415
    },
    {
      "entropy": 5.652270030975342,
      "epoch": 2.6220722508932117,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043166813891854796,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18295522630214692,
      "num_tokens": 57887938.0,
      "step": 26420
    },
    {
      "entropy": 5.6744834899902346,
      "epoch": 2.622568479555379,
      "grad_norm": 1.015625,
      "learning_rate": 0.000431642706086631,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19028469324111938,
      "num_tokens": 57899375.0,
      "step": 26425
    },
    {
      "entropy": 5.674715757369995,
      "epoch": 2.6230647082175467,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043161726937031243,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17556044310331345,
      "num_tokens": 57910393.0,
      "step": 26430
    },
    {
      "entropy": 5.690600633621216,
      "epoch": 2.623560936879714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000431591828770223,
      "loss": 5.0727,
      "mean_token_accuracy": 0.18169171661138533,
      "num_tokens": 57922236.0,
      "step": 26435
    },
    {
      "entropy": 5.655585241317749,
      "epoch": 2.6240571655418816,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004315663842869938,
      "loss": 5.0403,
      "mean_token_accuracy": 0.1856545090675354,
      "num_tokens": 57931804.0,
      "step": 26440
    },
    {
      "entropy": 5.650565052032471,
      "epoch": 2.624553394204049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043154093592125554,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18041322082281114,
      "num_tokens": 57943259.0,
      "step": 26445
    },
    {
      "entropy": 5.582130861282349,
      "epoch": 2.625049622866217,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043151548367363943,
      "loss": 4.9289,
      "mean_token_accuracy": 0.1878558427095413,
      "num_tokens": 57952945.0,
      "step": 26450
    },
    {
      "entropy": 5.629783630371094,
      "epoch": 2.6255458515283845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043149002754477655,
      "loss": 5.0176,
      "mean_token_accuracy": 0.1843608871102333,
      "num_tokens": 57964030.0,
      "step": 26455
    },
    {
      "entropy": 5.635840988159179,
      "epoch": 2.626042080190552,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004314645675352982,
      "loss": 4.9899,
      "mean_token_accuracy": 0.19176736623048782,
      "num_tokens": 57974888.0,
      "step": 26460
    },
    {
      "entropy": 5.711720657348633,
      "epoch": 2.6265383088527194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004314391036458357,
      "loss": 5.0125,
      "mean_token_accuracy": 0.18389740735292434,
      "num_tokens": 57984922.0,
      "step": 26465
    },
    {
      "entropy": 5.597965240478516,
      "epoch": 2.627034537514887,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043141363587702054,
      "loss": 4.998,
      "mean_token_accuracy": 0.18639900386333466,
      "num_tokens": 57994881.0,
      "step": 26470
    },
    {
      "entropy": 5.654440021514892,
      "epoch": 2.6275307661770544,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004313881642294843,
      "loss": 5.0822,
      "mean_token_accuracy": 0.18130575567483903,
      "num_tokens": 58005335.0,
      "step": 26475
    },
    {
      "entropy": 5.664693832397461,
      "epoch": 2.628026994839222,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004313626887038585,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18886601477861403,
      "num_tokens": 58016692.0,
      "step": 26480
    },
    {
      "entropy": 5.668198490142823,
      "epoch": 2.6285232235013893,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431337209300775,
      "loss": 5.0693,
      "mean_token_accuracy": 0.18446541875600814,
      "num_tokens": 58028282.0,
      "step": 26485
    },
    {
      "entropy": 5.608150959014893,
      "epoch": 2.629019452163557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004313117260208656,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18703087121248246,
      "num_tokens": 58040108.0,
      "step": 26490
    },
    {
      "entropy": 5.585564661026001,
      "epoch": 2.6295156808257243,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004312862388647622,
      "loss": 4.9949,
      "mean_token_accuracy": 0.18712320774793625,
      "num_tokens": 58050156.0,
      "step": 26495
    },
    {
      "entropy": 5.6164322853088375,
      "epoch": 2.6300119094878918,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043126074783309687,
      "loss": 5.0007,
      "mean_token_accuracy": 0.1818867564201355,
      "num_tokens": 58060929.0,
      "step": 26500
    },
    {
      "entropy": 5.675986194610596,
      "epoch": 2.6305081381500597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004312352529265016,
      "loss": 5.0534,
      "mean_token_accuracy": 0.179039266705513,
      "num_tokens": 58071684.0,
      "step": 26505
    },
    {
      "entropy": 5.685628080368042,
      "epoch": 2.631004366812227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043120975414560886,
      "loss": 5.0694,
      "mean_token_accuracy": 0.18406491130590438,
      "num_tokens": 58082730.0,
      "step": 26510
    },
    {
      "entropy": 5.713964462280273,
      "epoch": 2.6315005954743946,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00043118425149105073,
      "loss": 5.1403,
      "mean_token_accuracy": 0.17879558950662613,
      "num_tokens": 58095278.0,
      "step": 26515
    },
    {
      "entropy": 5.785606098175049,
      "epoch": 2.631996824136562,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004311587449634597,
      "loss": 5.1801,
      "mean_token_accuracy": 0.1737062469124794,
      "num_tokens": 58107816.0,
      "step": 26520
    },
    {
      "entropy": 5.732015562057495,
      "epoch": 2.6324930527987296,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004311332345634683,
      "loss": 5.1177,
      "mean_token_accuracy": 0.17292661368846893,
      "num_tokens": 58117779.0,
      "step": 26525
    },
    {
      "entropy": 5.699132585525513,
      "epoch": 2.632989281460897,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043110772029170916,
      "loss": 5.0155,
      "mean_token_accuracy": 0.18836407214403153,
      "num_tokens": 58128665.0,
      "step": 26530
    },
    {
      "entropy": 5.696052742004395,
      "epoch": 2.633485510123065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043108220214881494,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17828207910060884,
      "num_tokens": 58139658.0,
      "step": 26535
    },
    {
      "entropy": 5.710921859741211,
      "epoch": 2.6339817387852325,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004310566801354184,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1817881777882576,
      "num_tokens": 58149980.0,
      "step": 26540
    },
    {
      "entropy": 5.653774261474609,
      "epoch": 2.6344779674474,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043103115425215246,
      "loss": 4.9485,
      "mean_token_accuracy": 0.18465085029602052,
      "num_tokens": 58159809.0,
      "step": 26545
    },
    {
      "entropy": 5.636471891403199,
      "epoch": 2.6349741961095674,
      "grad_norm": 1.125,
      "learning_rate": 0.00043100562449965016,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18160324096679686,
      "num_tokens": 58169912.0,
      "step": 26550
    },
    {
      "entropy": 5.678711700439453,
      "epoch": 2.635470424771735,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043098009087854446,
      "loss": 5.065,
      "mean_token_accuracy": 0.1832931309938431,
      "num_tokens": 58179949.0,
      "step": 26555
    },
    {
      "entropy": 5.6298912525177,
      "epoch": 2.6359666534339024,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043095455338946863,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1884173795580864,
      "num_tokens": 58189847.0,
      "step": 26560
    },
    {
      "entropy": 5.627227354049682,
      "epoch": 2.63646288209607,
      "grad_norm": 1.0625,
      "learning_rate": 0.000430929012033056,
      "loss": 4.946,
      "mean_token_accuracy": 0.189795845746994,
      "num_tokens": 58201514.0,
      "step": 26565
    },
    {
      "entropy": 5.6597517967224125,
      "epoch": 2.6369591107582373,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043090346680993983,
      "loss": 5.0383,
      "mean_token_accuracy": 0.1891452819108963,
      "num_tokens": 58211940.0,
      "step": 26570
    },
    {
      "entropy": 5.72932620048523,
      "epoch": 2.6374553394204048,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004308779177207536,
      "loss": 5.0561,
      "mean_token_accuracy": 0.1786615803837776,
      "num_tokens": 58222260.0,
      "step": 26575
    },
    {
      "entropy": 5.6436378955841064,
      "epoch": 2.6379515680825723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004308523647661309,
      "loss": 5.0052,
      "mean_token_accuracy": 0.1830901324748993,
      "num_tokens": 58232649.0,
      "step": 26580
    },
    {
      "entropy": 5.610378170013428,
      "epoch": 2.6384477967447397,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043082680794670535,
      "loss": 4.9317,
      "mean_token_accuracy": 0.19316525310277938,
      "num_tokens": 58243373.0,
      "step": 26585
    },
    {
      "entropy": 5.568605613708496,
      "epoch": 2.6389440254069076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043080124726311084,
      "loss": 5.0186,
      "mean_token_accuracy": 0.18502314686775206,
      "num_tokens": 58253863.0,
      "step": 26590
    },
    {
      "entropy": 5.6309727191925045,
      "epoch": 2.639440254069075,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004307756827159811,
      "loss": 5.1008,
      "mean_token_accuracy": 0.18810600638389588,
      "num_tokens": 58265326.0,
      "step": 26595
    },
    {
      "entropy": 5.703416538238526,
      "epoch": 2.6399364827312426,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043075011430594994,
      "loss": 5.0685,
      "mean_token_accuracy": 0.18294407725334166,
      "num_tokens": 58276006.0,
      "step": 26600
    },
    {
      "entropy": 5.640567398071289,
      "epoch": 2.64043271139341,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043072454203365175,
      "loss": 4.9324,
      "mean_token_accuracy": 0.18940143436193466,
      "num_tokens": 58285932.0,
      "step": 26605
    },
    {
      "entropy": 5.647582006454468,
      "epoch": 2.6409289400555775,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004306989658997204,
      "loss": 5.0136,
      "mean_token_accuracy": 0.18504741191864013,
      "num_tokens": 58297174.0,
      "step": 26610
    },
    {
      "entropy": 5.711141967773438,
      "epoch": 2.641425168717745,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043067338590479023,
      "loss": 5.1539,
      "mean_token_accuracy": 0.17446687072515488,
      "num_tokens": 58308509.0,
      "step": 26615
    },
    {
      "entropy": 5.641808080673218,
      "epoch": 2.641921397379913,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043064780204949543,
      "loss": 5.0574,
      "mean_token_accuracy": 0.1753106400370598,
      "num_tokens": 58318449.0,
      "step": 26620
    },
    {
      "entropy": 5.698845243453979,
      "epoch": 2.6424176260420804,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004306222143344706,
      "loss": 5.0149,
      "mean_token_accuracy": 0.1848238602280617,
      "num_tokens": 58329759.0,
      "step": 26625
    },
    {
      "entropy": 5.660292243957519,
      "epoch": 2.642913854704248,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004305966227603503,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18415968865156174,
      "num_tokens": 58339363.0,
      "step": 26630
    },
    {
      "entropy": 5.606242561340332,
      "epoch": 2.6434100833664154,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305710273277689,
      "loss": 5.0298,
      "mean_token_accuracy": 0.17971438318490982,
      "num_tokens": 58349482.0,
      "step": 26635
    },
    {
      "entropy": 5.684519720077515,
      "epoch": 2.643906312028583,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043054542803736126,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17900317758321763,
      "num_tokens": 58359879.0,
      "step": 26640
    },
    {
      "entropy": 5.736082553863525,
      "epoch": 2.6444025406907503,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043051982488976216,
      "loss": 5.1011,
      "mean_token_accuracy": 0.17222164571285248,
      "num_tokens": 58370475.0,
      "step": 26645
    },
    {
      "entropy": 5.6942095279693605,
      "epoch": 2.644898769352918,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004304942178856066,
      "loss": 5.0352,
      "mean_token_accuracy": 0.18235959559679032,
      "num_tokens": 58381992.0,
      "step": 26650
    },
    {
      "entropy": 5.6517603397369385,
      "epoch": 2.6453949980150853,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043046860702552944,
      "loss": 4.9952,
      "mean_token_accuracy": 0.18960391730070114,
      "num_tokens": 58391616.0,
      "step": 26655
    },
    {
      "entropy": 5.70772795677185,
      "epoch": 2.6458912266772527,
      "grad_norm": 1.25,
      "learning_rate": 0.0004304429923101658,
      "loss": 5.0092,
      "mean_token_accuracy": 0.18468829542398452,
      "num_tokens": 58401800.0,
      "step": 26660
    },
    {
      "entropy": 5.631481313705445,
      "epoch": 2.64638745533942,
      "grad_norm": 1.03125,
      "learning_rate": 0.000430417373740151,
      "loss": 4.9918,
      "mean_token_accuracy": 0.1864980489015579,
      "num_tokens": 58413376.0,
      "step": 26665
    },
    {
      "entropy": 5.543612432479859,
      "epoch": 2.6468836840015877,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004303917513161201,
      "loss": 4.9636,
      "mean_token_accuracy": 0.18857150971889497,
      "num_tokens": 58423772.0,
      "step": 26670
    },
    {
      "entropy": 5.6667979717254635,
      "epoch": 2.6473799126637556,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004303661250387088,
      "loss": 5.0712,
      "mean_token_accuracy": 0.17800713181495667,
      "num_tokens": 58434838.0,
      "step": 26675
    },
    {
      "entropy": 5.6880792617797855,
      "epoch": 2.647876141325923,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004303404949085522,
      "loss": 5.0894,
      "mean_token_accuracy": 0.17923627197742462,
      "num_tokens": 58445036.0,
      "step": 26680
    },
    {
      "entropy": 5.649956798553466,
      "epoch": 2.6483723699880906,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004303148609262861,
      "loss": 5.0123,
      "mean_token_accuracy": 0.1846483066678047,
      "num_tokens": 58455558.0,
      "step": 26685
    },
    {
      "entropy": 5.632132196426392,
      "epoch": 2.648868598650258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004302892230925462,
      "loss": 5.0576,
      "mean_token_accuracy": 0.17279154062271118,
      "num_tokens": 58467716.0,
      "step": 26690
    },
    {
      "entropy": 5.620901870727539,
      "epoch": 2.6493648273124255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043026358140796806,
      "loss": 5.0355,
      "mean_token_accuracy": 0.18112232834100722,
      "num_tokens": 58479421.0,
      "step": 26695
    },
    {
      "entropy": 5.6048980236053465,
      "epoch": 2.649861055974593,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004302379358731878,
      "loss": 5.0275,
      "mean_token_accuracy": 0.18656516671180726,
      "num_tokens": 58490115.0,
      "step": 26700
    },
    {
      "entropy": 5.613460493087769,
      "epoch": 2.6503572846367605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004302122864888411,
      "loss": 4.9732,
      "mean_token_accuracy": 0.18542134910821914,
      "num_tokens": 58501711.0,
      "step": 26705
    },
    {
      "entropy": 5.670157241821289,
      "epoch": 2.6508535132989284,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004301866332555643,
      "loss": 5.005,
      "mean_token_accuracy": 0.19033184349536897,
      "num_tokens": 58513441.0,
      "step": 26710
    },
    {
      "entropy": 5.634325647354126,
      "epoch": 2.651349741961096,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043016097617399327,
      "loss": 4.9982,
      "mean_token_accuracy": 0.1853046879172325,
      "num_tokens": 58523940.0,
      "step": 26715
    },
    {
      "entropy": 5.631616163253784,
      "epoch": 2.6518459706232633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004301353152447644,
      "loss": 5.0538,
      "mean_token_accuracy": 0.1849694013595581,
      "num_tokens": 58536354.0,
      "step": 26720
    },
    {
      "entropy": 5.70519437789917,
      "epoch": 2.652342199285431,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043010965046851404,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18081564754247664,
      "num_tokens": 58547096.0,
      "step": 26725
    },
    {
      "entropy": 5.662531614303589,
      "epoch": 2.6528384279475983,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043008398184587847,
      "loss": 5.0714,
      "mean_token_accuracy": 0.17900378704071046,
      "num_tokens": 58557951.0,
      "step": 26730
    },
    {
      "entropy": 5.611026811599731,
      "epoch": 2.6533346566097658,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043005830937749447,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18583527207374573,
      "num_tokens": 58569384.0,
      "step": 26735
    },
    {
      "entropy": 5.631719350814819,
      "epoch": 2.6538308852719332,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004300326330639984,
      "loss": 5.0313,
      "mean_token_accuracy": 0.18020059913396835,
      "num_tokens": 58580308.0,
      "step": 26740
    },
    {
      "entropy": 5.622963905334473,
      "epoch": 2.6543271139341007,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004300069529060272,
      "loss": 5.0442,
      "mean_token_accuracy": 0.17844083309173583,
      "num_tokens": 58591780.0,
      "step": 26745
    },
    {
      "entropy": 5.654569435119629,
      "epoch": 2.654823342596268,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042998126890421754,
      "loss": 5.0513,
      "mean_token_accuracy": 0.18161566257476808,
      "num_tokens": 58601477.0,
      "step": 26750
    },
    {
      "entropy": 5.661498308181763,
      "epoch": 2.6553195712584357,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042995558105920627,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1889594629406929,
      "num_tokens": 58612531.0,
      "step": 26755
    },
    {
      "entropy": 5.645244693756103,
      "epoch": 2.6558157999206036,
      "grad_norm": 1.0,
      "learning_rate": 0.00042992988937163066,
      "loss": 5.0852,
      "mean_token_accuracy": 0.17822430431842803,
      "num_tokens": 58624951.0,
      "step": 26760
    },
    {
      "entropy": 5.64632511138916,
      "epoch": 2.656312028582771,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004299041938421275,
      "loss": 5.1254,
      "mean_token_accuracy": 0.1798737168312073,
      "num_tokens": 58636484.0,
      "step": 26765
    },
    {
      "entropy": 5.692964744567871,
      "epoch": 2.6568082572449385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042987849447133423,
      "loss": 5.0326,
      "mean_token_accuracy": 0.17759032398462296,
      "num_tokens": 58647386.0,
      "step": 26770
    },
    {
      "entropy": 5.713610458374023,
      "epoch": 2.657304485907106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004298527912598879,
      "loss": 5.1039,
      "mean_token_accuracy": 0.18283108174800872,
      "num_tokens": 58658656.0,
      "step": 26775
    },
    {
      "entropy": 5.670279121398925,
      "epoch": 2.6578007145692735,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004298270842084261,
      "loss": 4.987,
      "mean_token_accuracy": 0.18524612933397294,
      "num_tokens": 58669901.0,
      "step": 26780
    },
    {
      "entropy": 5.621233701705933,
      "epoch": 2.658296943231441,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042980137331758626,
      "loss": 5.0689,
      "mean_token_accuracy": 0.18220627903938294,
      "num_tokens": 58680849.0,
      "step": 26785
    },
    {
      "entropy": 5.623178577423095,
      "epoch": 2.6587931718936084,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042977565858800593,
      "loss": 5.0015,
      "mean_token_accuracy": 0.17933868616819382,
      "num_tokens": 58690642.0,
      "step": 26790
    },
    {
      "entropy": 5.601643085479736,
      "epoch": 2.6592894005557763,
      "grad_norm": 1.125,
      "learning_rate": 0.00042974994002032275,
      "loss": 4.9534,
      "mean_token_accuracy": 0.18884531110525132,
      "num_tokens": 58700480.0,
      "step": 26795
    },
    {
      "entropy": 5.660499238967896,
      "epoch": 2.659785629217944,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042972421761517456,
      "loss": 5.0858,
      "mean_token_accuracy": 0.1746712639927864,
      "num_tokens": 58712443.0,
      "step": 26800
    },
    {
      "entropy": 5.62268099784851,
      "epoch": 2.6602818578801113,
      "grad_norm": 1.125,
      "learning_rate": 0.0004296984913731991,
      "loss": 4.9833,
      "mean_token_accuracy": 0.1812335029244423,
      "num_tokens": 58721172.0,
      "step": 26805
    },
    {
      "entropy": 5.659602451324463,
      "epoch": 2.6607780865422788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004296727612950344,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18722205609083176,
      "num_tokens": 58731060.0,
      "step": 26810
    },
    {
      "entropy": 5.643565940856933,
      "epoch": 2.6612743152044462,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004296470273813186,
      "loss": 5.0295,
      "mean_token_accuracy": 0.1866572842001915,
      "num_tokens": 58742189.0,
      "step": 26815
    },
    {
      "entropy": 5.7019964218139645,
      "epoch": 2.6617705438666137,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004296212896326896,
      "loss": 5.0469,
      "mean_token_accuracy": 0.18758414536714554,
      "num_tokens": 58753617.0,
      "step": 26820
    },
    {
      "entropy": 5.63217134475708,
      "epoch": 2.662266772528781,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042959554804978596,
      "loss": 5.055,
      "mean_token_accuracy": 0.18427814841270446,
      "num_tokens": 58765299.0,
      "step": 26825
    },
    {
      "entropy": 5.697397136688233,
      "epoch": 2.6627630011909487,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042956980263324576,
      "loss": 4.9832,
      "mean_token_accuracy": 0.18759999871253968,
      "num_tokens": 58775280.0,
      "step": 26830
    },
    {
      "entropy": 5.708413314819336,
      "epoch": 2.663259229853116,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004295440533837075,
      "loss": 4.9781,
      "mean_token_accuracy": 0.18907142281532288,
      "num_tokens": 58786078.0,
      "step": 26835
    },
    {
      "entropy": 5.682507514953613,
      "epoch": 2.6637554585152836,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042951830030180983,
      "loss": 5.0637,
      "mean_token_accuracy": 0.1811142757534981,
      "num_tokens": 58796665.0,
      "step": 26840
    },
    {
      "entropy": 5.6898761749267575,
      "epoch": 2.6642516871774515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042949254338819123,
      "loss": 5.0706,
      "mean_token_accuracy": 0.1848100543022156,
      "num_tokens": 58808150.0,
      "step": 26845
    },
    {
      "entropy": 5.705758666992187,
      "epoch": 2.664747915839619,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042946678264349036,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18400920927524567,
      "num_tokens": 58818278.0,
      "step": 26850
    },
    {
      "entropy": 5.654260063171387,
      "epoch": 2.6652441445017865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042944101806834627,
      "loss": 5.0485,
      "mean_token_accuracy": 0.18083391934633256,
      "num_tokens": 58828847.0,
      "step": 26855
    },
    {
      "entropy": 5.636255979537964,
      "epoch": 2.665740373163954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004294152496633977,
      "loss": 5.0193,
      "mean_token_accuracy": 0.18276432752609253,
      "num_tokens": 58839698.0,
      "step": 26860
    },
    {
      "entropy": 5.595601558685303,
      "epoch": 2.6662366018261214,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042938947742928357,
      "loss": 4.9883,
      "mean_token_accuracy": 0.18337797671556472,
      "num_tokens": 58850794.0,
      "step": 26865
    },
    {
      "entropy": 5.663086557388306,
      "epoch": 2.666732830488289,
      "grad_norm": 1.0,
      "learning_rate": 0.00042936370136664314,
      "loss": 5.0822,
      "mean_token_accuracy": 0.17753103971481324,
      "num_tokens": 58861590.0,
      "step": 26870
    },
    {
      "entropy": 5.693813896179199,
      "epoch": 2.6672290591504564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004293379214761156,
      "loss": 5.1051,
      "mean_token_accuracy": 0.17758218795061112,
      "num_tokens": 58873796.0,
      "step": 26875
    },
    {
      "entropy": 5.6803466796875,
      "epoch": 2.6677252878126243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004293121377583401,
      "loss": 5.0624,
      "mean_token_accuracy": 0.18288323581218718,
      "num_tokens": 58885296.0,
      "step": 26880
    },
    {
      "entropy": 5.6461669921875,
      "epoch": 2.6682215164747918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004292863502139561,
      "loss": 5.0955,
      "mean_token_accuracy": 0.18052715808153152,
      "num_tokens": 58897764.0,
      "step": 26885
    },
    {
      "entropy": 5.684815359115601,
      "epoch": 2.6687177451369593,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004292605588436031,
      "loss": 5.0559,
      "mean_token_accuracy": 0.17900653630495073,
      "num_tokens": 58908964.0,
      "step": 26890
    },
    {
      "entropy": 5.5613922595977785,
      "epoch": 2.6692139737991267,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042923476364792063,
      "loss": 4.9187,
      "mean_token_accuracy": 0.18763911873102188,
      "num_tokens": 58919152.0,
      "step": 26895
    },
    {
      "entropy": 5.5607317924499515,
      "epoch": 2.669710202461294,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004292089646275484,
      "loss": 5.0596,
      "mean_token_accuracy": 0.18434956669807434,
      "num_tokens": 58928849.0,
      "step": 26900
    },
    {
      "entropy": 5.721317625045776,
      "epoch": 2.6702064311234617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004291831617831261,
      "loss": 5.1699,
      "mean_token_accuracy": 0.17781679481267929,
      "num_tokens": 58941560.0,
      "step": 26905
    },
    {
      "entropy": 5.730229139328003,
      "epoch": 2.670702659785629,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042915735511529356,
      "loss": 5.1193,
      "mean_token_accuracy": 0.17243372946977614,
      "num_tokens": 58953365.0,
      "step": 26910
    },
    {
      "entropy": 5.675168228149414,
      "epoch": 2.6711988884477966,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004291315446246909,
      "loss": 5.023,
      "mean_token_accuracy": 0.18534417301416398,
      "num_tokens": 58964602.0,
      "step": 26915
    },
    {
      "entropy": 5.656284284591675,
      "epoch": 2.671695117109964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042910573031195804,
      "loss": 5.0292,
      "mean_token_accuracy": 0.18517274856567384,
      "num_tokens": 58976154.0,
      "step": 26920
    },
    {
      "entropy": 5.650727272033691,
      "epoch": 2.6721913457721316,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004290799121777351,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18566546440124512,
      "num_tokens": 58985989.0,
      "step": 26925
    },
    {
      "entropy": 5.728303098678589,
      "epoch": 2.672687574434299,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004290540902226624,
      "loss": 5.1248,
      "mean_token_accuracy": 0.17260902523994445,
      "num_tokens": 58997080.0,
      "step": 26930
    },
    {
      "entropy": 5.653652811050415,
      "epoch": 2.673183803096467,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004290282644473802,
      "loss": 5.0459,
      "mean_token_accuracy": 0.18177807480096816,
      "num_tokens": 59008683.0,
      "step": 26935
    },
    {
      "entropy": 5.648606681823731,
      "epoch": 2.6736800317586344,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004290024348525289,
      "loss": 4.9649,
      "mean_token_accuracy": 0.18906575739383696,
      "num_tokens": 59019019.0,
      "step": 26940
    },
    {
      "entropy": 5.615896081924438,
      "epoch": 2.674176260420802,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004289766014387492,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18541242331266403,
      "num_tokens": 59029706.0,
      "step": 26945
    },
    {
      "entropy": 5.628919076919556,
      "epoch": 2.6746724890829694,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004289507642066813,
      "loss": 4.9649,
      "mean_token_accuracy": 0.1864032819867134,
      "num_tokens": 59040208.0,
      "step": 26950
    },
    {
      "entropy": 5.611232471466065,
      "epoch": 2.675168717745137,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004289249231569664,
      "loss": 5.0015,
      "mean_token_accuracy": 0.19420036375522615,
      "num_tokens": 59050909.0,
      "step": 26955
    },
    {
      "entropy": 5.579736185073853,
      "epoch": 2.6756649464073043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042889907829024506,
      "loss": 4.991,
      "mean_token_accuracy": 0.1873885542154312,
      "num_tokens": 59059834.0,
      "step": 26960
    },
    {
      "entropy": 5.662536001205444,
      "epoch": 2.6761611750694723,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042887322960715813,
      "loss": 5.0602,
      "mean_token_accuracy": 0.17748401015996934,
      "num_tokens": 59071348.0,
      "step": 26965
    },
    {
      "entropy": 5.648468732833862,
      "epoch": 2.6766574037316397,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004288473771083468,
      "loss": 5.0105,
      "mean_token_accuracy": 0.180013570189476,
      "num_tokens": 59083053.0,
      "step": 26970
    },
    {
      "entropy": 5.682869577407837,
      "epoch": 2.677153632393807,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004288215207944519,
      "loss": 5.0947,
      "mean_token_accuracy": 0.17682799994945525,
      "num_tokens": 59093400.0,
      "step": 26975
    },
    {
      "entropy": 5.641188144683838,
      "epoch": 2.6776498610559747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042879566066611474,
      "loss": 5.0679,
      "mean_token_accuracy": 0.17711251080036164,
      "num_tokens": 59103653.0,
      "step": 26980
    },
    {
      "entropy": 5.674331283569336,
      "epoch": 2.678146089718142,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004287697967239767,
      "loss": 5.0643,
      "mean_token_accuracy": 0.1819843217730522,
      "num_tokens": 59116660.0,
      "step": 26985
    },
    {
      "entropy": 5.606666517257691,
      "epoch": 2.6786423183803096,
      "grad_norm": 1.171875,
      "learning_rate": 0.000428743928968679,
      "loss": 4.9464,
      "mean_token_accuracy": 0.18893352001905442,
      "num_tokens": 59127424.0,
      "step": 26990
    },
    {
      "entropy": 5.671489524841308,
      "epoch": 2.679138547042477,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042871805740086317,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1804300293326378,
      "num_tokens": 59137566.0,
      "step": 26995
    },
    {
      "entropy": 5.584001970291138,
      "epoch": 2.6796347757046446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004286921820211707,
      "loss": 4.9701,
      "mean_token_accuracy": 0.18876097947359086,
      "num_tokens": 59148404.0,
      "step": 27000
    },
    {
      "epoch": 2.6796347757046446,
      "eval_entropy": 5.401918924797966,
      "eval_loss": 5.15195369720459,
      "eval_mean_token_accuracy": 0.18225784066079434,
      "eval_num_tokens": 59148404.0,
      "eval_runtime": 20.6496,
      "eval_samples_per_second": 1630.495,
      "eval_steps_per_second": 203.83,
      "step": 27000
    },
    {
      "entropy": 5.584519958496093,
      "epoch": 2.680131004366812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042866630283024327,
      "loss": 5.0266,
      "mean_token_accuracy": 0.17981680184602739,
      "num_tokens": 59158097.0,
      "step": 27005
    },
    {
      "entropy": 5.554842805862426,
      "epoch": 2.6806272330289795,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004286404198287227,
      "loss": 4.9036,
      "mean_token_accuracy": 0.19506526440382005,
      "num_tokens": 59168940.0,
      "step": 27010
    },
    {
      "entropy": 5.643365812301636,
      "epoch": 2.681123461691147,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004286145330172508,
      "loss": 5.0228,
      "mean_token_accuracy": 0.18273395895957947,
      "num_tokens": 59180231.0,
      "step": 27015
    },
    {
      "entropy": 5.702926254272461,
      "epoch": 2.681619690353315,
      "grad_norm": 1.125,
      "learning_rate": 0.0004285886423964695,
      "loss": 5.1215,
      "mean_token_accuracy": 0.1724468007683754,
      "num_tokens": 59191565.0,
      "step": 27020
    },
    {
      "entropy": 5.642581415176392,
      "epoch": 2.6821159190154824,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004285627479670208,
      "loss": 5.1048,
      "mean_token_accuracy": 0.1781492203474045,
      "num_tokens": 59203394.0,
      "step": 27025
    },
    {
      "entropy": 5.576269483566284,
      "epoch": 2.68261214767765,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004285368497295468,
      "loss": 4.9655,
      "mean_token_accuracy": 0.18239316642284392,
      "num_tokens": 59213987.0,
      "step": 27030
    },
    {
      "entropy": 5.6325243473052975,
      "epoch": 2.6831083763398174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004285109476846898,
      "loss": 5.0026,
      "mean_token_accuracy": 0.18734180331230163,
      "num_tokens": 59225825.0,
      "step": 27035
    },
    {
      "entropy": 5.577541208267212,
      "epoch": 2.683604605001985,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004284850418330921,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18771986067295074,
      "num_tokens": 59235978.0,
      "step": 27040
    },
    {
      "entropy": 5.718644285202027,
      "epoch": 2.6841008336641523,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042845913217539613,
      "loss": 5.1523,
      "mean_token_accuracy": 0.17963545322418212,
      "num_tokens": 59247505.0,
      "step": 27045
    },
    {
      "entropy": 5.640657806396485,
      "epoch": 2.6845970623263202,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042843321871224426,
      "loss": 4.9481,
      "mean_token_accuracy": 0.18675633519887924,
      "num_tokens": 59258401.0,
      "step": 27050
    },
    {
      "entropy": 5.665589141845703,
      "epoch": 2.6850932909884877,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042840730144427927,
      "loss": 5.0323,
      "mean_token_accuracy": 0.1867706909775734,
      "num_tokens": 59268068.0,
      "step": 27055
    },
    {
      "entropy": 5.633105707168579,
      "epoch": 2.685589519650655,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042838138037214375,
      "loss": 5.0229,
      "mean_token_accuracy": 0.18506719321012496,
      "num_tokens": 59279567.0,
      "step": 27060
    },
    {
      "entropy": 5.644492673873901,
      "epoch": 2.6860857483128227,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042835545549648054,
      "loss": 4.9639,
      "mean_token_accuracy": 0.18795686960220337,
      "num_tokens": 59289862.0,
      "step": 27065
    },
    {
      "entropy": 5.655943393707275,
      "epoch": 2.68658197697499,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042832952681793235,
      "loss": 5.005,
      "mean_token_accuracy": 0.19091366231441498,
      "num_tokens": 59301304.0,
      "step": 27070
    },
    {
      "entropy": 5.646803569793701,
      "epoch": 2.6870782056371576,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042830359433714243,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1813417688012123,
      "num_tokens": 59312184.0,
      "step": 27075
    },
    {
      "entropy": 5.609905862808228,
      "epoch": 2.687574434299325,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042827765805475367,
      "loss": 4.9112,
      "mean_token_accuracy": 0.18997137546539306,
      "num_tokens": 59322688.0,
      "step": 27080
    },
    {
      "entropy": 5.610944652557373,
      "epoch": 2.6880706629614926,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042825171797140926,
      "loss": 4.978,
      "mean_token_accuracy": 0.1893126428127289,
      "num_tokens": 59333349.0,
      "step": 27085
    },
    {
      "entropy": 5.599378967285157,
      "epoch": 2.68856689162366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004282257740877525,
      "loss": 4.9739,
      "mean_token_accuracy": 0.19330060482025146,
      "num_tokens": 59344473.0,
      "step": 27090
    },
    {
      "entropy": 5.640298795700073,
      "epoch": 2.6890631202858275,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004281998264044267,
      "loss": 5.049,
      "mean_token_accuracy": 0.18440739810466766,
      "num_tokens": 59355993.0,
      "step": 27095
    },
    {
      "entropy": 5.723005294799805,
      "epoch": 2.689559348947995,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004281738749220753,
      "loss": 5.1091,
      "mean_token_accuracy": 0.18090903460979463,
      "num_tokens": 59367182.0,
      "step": 27100
    },
    {
      "entropy": 5.729381418228149,
      "epoch": 2.690055577610163,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042814791964134184,
      "loss": 5.077,
      "mean_token_accuracy": 0.18066031038761138,
      "num_tokens": 59379235.0,
      "step": 27105
    },
    {
      "entropy": 5.6548220157623295,
      "epoch": 2.6905518062723304,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042812196056287,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18800578862428666,
      "num_tokens": 59388514.0,
      "step": 27110
    },
    {
      "entropy": 5.635708332061768,
      "epoch": 2.691048034934498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042809599768730357,
      "loss": 5.036,
      "mean_token_accuracy": 0.18138365298509598,
      "num_tokens": 59399496.0,
      "step": 27115
    },
    {
      "entropy": 5.607694435119629,
      "epoch": 2.6915442635966653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004280700310152861,
      "loss": 5.0248,
      "mean_token_accuracy": 0.18721487671136855,
      "num_tokens": 59411986.0,
      "step": 27120
    },
    {
      "entropy": 5.7541149139404295,
      "epoch": 2.692040492258833,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004280440605474619,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1695254474878311,
      "num_tokens": 59424232.0,
      "step": 27125
    },
    {
      "entropy": 5.720369052886963,
      "epoch": 2.6925367209210003,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004280180862844747,
      "loss": 5.0614,
      "mean_token_accuracy": 0.17734398543834687,
      "num_tokens": 59435201.0,
      "step": 27130
    },
    {
      "entropy": 5.696267795562744,
      "epoch": 2.693032949583168,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042799210822696865,
      "loss": 5.097,
      "mean_token_accuracy": 0.17638006955385208,
      "num_tokens": 59445552.0,
      "step": 27135
    },
    {
      "entropy": 5.741204261779785,
      "epoch": 2.6935291782453357,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000427966126375588,
      "loss": 5.0913,
      "mean_token_accuracy": 0.173358453810215,
      "num_tokens": 59455634.0,
      "step": 27140
    },
    {
      "entropy": 5.685871171951294,
      "epoch": 2.694025406907503,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004279401407309771,
      "loss": 5.0601,
      "mean_token_accuracy": 0.17948268055915834,
      "num_tokens": 59467955.0,
      "step": 27145
    },
    {
      "entropy": 5.668799924850464,
      "epoch": 2.6945216355696706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004279141512937802,
      "loss": 5.0616,
      "mean_token_accuracy": 0.18047458976507186,
      "num_tokens": 59479566.0,
      "step": 27150
    },
    {
      "entropy": 5.574527311325073,
      "epoch": 2.695017864231838,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004278881580646418,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18450209349393845,
      "num_tokens": 59489967.0,
      "step": 27155
    },
    {
      "entropy": 5.623745584487915,
      "epoch": 2.6955140928940056,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004278621610442066,
      "loss": 5.0486,
      "mean_token_accuracy": 0.17964051067829132,
      "num_tokens": 59501640.0,
      "step": 27160
    },
    {
      "entropy": 5.654453325271606,
      "epoch": 2.696010321556173,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004278361602331192,
      "loss": 4.9775,
      "mean_token_accuracy": 0.1919047251343727,
      "num_tokens": 59512767.0,
      "step": 27165
    },
    {
      "entropy": 5.6644550323486325,
      "epoch": 2.6965065502183405,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042781015563202434,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18782842308282852,
      "num_tokens": 59522188.0,
      "step": 27170
    },
    {
      "entropy": 5.645979404449463,
      "epoch": 2.697002778880508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004277841472415669,
      "loss": 4.9968,
      "mean_token_accuracy": 0.188828244805336,
      "num_tokens": 59534871.0,
      "step": 27175
    },
    {
      "entropy": 5.757172870635986,
      "epoch": 2.6974990075426755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004277581350623918,
      "loss": 5.1293,
      "mean_token_accuracy": 0.17202058285474778,
      "num_tokens": 59546858.0,
      "step": 27180
    },
    {
      "entropy": 5.628689384460449,
      "epoch": 2.697995236204843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004277321190951441,
      "loss": 4.9756,
      "mean_token_accuracy": 0.18336831480264665,
      "num_tokens": 59557022.0,
      "step": 27185
    },
    {
      "entropy": 5.600583267211914,
      "epoch": 2.698491464867011,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042770609934046904,
      "loss": 4.9946,
      "mean_token_accuracy": 0.19027036130428315,
      "num_tokens": 59567564.0,
      "step": 27190
    },
    {
      "entropy": 5.748032283782959,
      "epoch": 2.6989876935291783,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042768007579901176,
      "loss": 5.1335,
      "mean_token_accuracy": 0.1742416188120842,
      "num_tokens": 59579122.0,
      "step": 27195
    },
    {
      "entropy": 5.741862154006958,
      "epoch": 2.699483922191346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042765404847141755,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1850947007536888,
      "num_tokens": 59590477.0,
      "step": 27200
    },
    {
      "entropy": 5.646586084365845,
      "epoch": 2.6999801508535133,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004276280173583319,
      "loss": 4.9888,
      "mean_token_accuracy": 0.18468654602766038,
      "num_tokens": 59601217.0,
      "step": 27205
    },
    {
      "entropy": 5.585481977462768,
      "epoch": 2.7004763795156808,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042760198246040025,
      "loss": 5.0096,
      "mean_token_accuracy": 0.19008333534002303,
      "num_tokens": 59612271.0,
      "step": 27210
    },
    {
      "entropy": 5.690351057052612,
      "epoch": 2.7009726081778482,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004275759437782683,
      "loss": 5.0871,
      "mean_token_accuracy": 0.1796427011489868,
      "num_tokens": 59623948.0,
      "step": 27215
    },
    {
      "entropy": 5.631183624267578,
      "epoch": 2.701468836840016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042754990131258175,
      "loss": 4.9579,
      "mean_token_accuracy": 0.1871839925646782,
      "num_tokens": 59635720.0,
      "step": 27220
    },
    {
      "entropy": 5.692176151275635,
      "epoch": 2.7019650655021836,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004275238550639863,
      "loss": 5.0662,
      "mean_token_accuracy": 0.18303292542695998,
      "num_tokens": 59645171.0,
      "step": 27225
    },
    {
      "entropy": 5.734927606582642,
      "epoch": 2.702461294164351,
      "grad_norm": 1.140625,
      "learning_rate": 0.000427497805033128,
      "loss": 5.1406,
      "mean_token_accuracy": 0.17657694220542908,
      "num_tokens": 59655266.0,
      "step": 27230
    },
    {
      "entropy": 5.646162748336792,
      "epoch": 2.7029575228265186,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004274717512206527,
      "loss": 5.0436,
      "mean_token_accuracy": 0.18472324460744857,
      "num_tokens": 59667055.0,
      "step": 27235
    },
    {
      "entropy": 5.644345474243164,
      "epoch": 2.703453751488686,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004274456936272065,
      "loss": 4.9975,
      "mean_token_accuracy": 0.18412665724754335,
      "num_tokens": 59678621.0,
      "step": 27240
    },
    {
      "entropy": 5.609760856628418,
      "epoch": 2.7039499801508535,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042741963225343554,
      "loss": 4.9901,
      "mean_token_accuracy": 0.1878741890192032,
      "num_tokens": 59690514.0,
      "step": 27245
    },
    {
      "entropy": 5.695537281036377,
      "epoch": 2.704446208813021,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004273935670999861,
      "loss": 5.0628,
      "mean_token_accuracy": 0.18152913600206375,
      "num_tokens": 59701271.0,
      "step": 27250
    },
    {
      "entropy": 5.646013164520264,
      "epoch": 2.7049424374751885,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042736749816750476,
      "loss": 5.1182,
      "mean_token_accuracy": 0.1744782030582428,
      "num_tokens": 59712787.0,
      "step": 27255
    },
    {
      "entropy": 5.6027826309204105,
      "epoch": 2.705438666137356,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004273414254566376,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18779198080301285,
      "num_tokens": 59722787.0,
      "step": 27260
    },
    {
      "entropy": 5.652699708938599,
      "epoch": 2.7059348947995234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004273153489680315,
      "loss": 5.0427,
      "mean_token_accuracy": 0.1827496126294136,
      "num_tokens": 59733302.0,
      "step": 27265
    },
    {
      "entropy": 5.614335775375366,
      "epoch": 2.706431123461691,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042728926870233287,
      "loss": 5.036,
      "mean_token_accuracy": 0.18168635070323944,
      "num_tokens": 59743290.0,
      "step": 27270
    },
    {
      "entropy": 5.686984586715698,
      "epoch": 2.706927352123859,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042726318466018855,
      "loss": 5.1034,
      "mean_token_accuracy": 0.17971232831478118,
      "num_tokens": 59754145.0,
      "step": 27275
    },
    {
      "entropy": 5.685484457015991,
      "epoch": 2.7074235807860263,
      "grad_norm": 1.125,
      "learning_rate": 0.0004272370968422454,
      "loss": 5.0537,
      "mean_token_accuracy": 0.18346502631902695,
      "num_tokens": 59764075.0,
      "step": 27280
    },
    {
      "entropy": 5.641423130035401,
      "epoch": 2.7079198094481938,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042721100524915017,
      "loss": 5.0767,
      "mean_token_accuracy": 0.17873225808143617,
      "num_tokens": 59775495.0,
      "step": 27285
    },
    {
      "entropy": 5.577423334121704,
      "epoch": 2.7084160381103612,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042718490988155014,
      "loss": 4.9284,
      "mean_token_accuracy": 0.20294324308633804,
      "num_tokens": 59787006.0,
      "step": 27290
    },
    {
      "entropy": 5.671135520935058,
      "epoch": 2.7089122667725287,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042715881074009216,
      "loss": 5.0576,
      "mean_token_accuracy": 0.17521505504846574,
      "num_tokens": 59797680.0,
      "step": 27295
    },
    {
      "entropy": 5.630173397064209,
      "epoch": 2.709408495434696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042713270782542364,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18554079085588454,
      "num_tokens": 59809666.0,
      "step": 27300
    },
    {
      "entropy": 5.601640892028809,
      "epoch": 2.7099047240968637,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004271066011381917,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18665453791618347,
      "num_tokens": 59819889.0,
      "step": 27305
    },
    {
      "entropy": 5.597252702713012,
      "epoch": 2.7104009527590316,
      "grad_norm": 1.125,
      "learning_rate": 0.0004270804906790438,
      "loss": 4.9508,
      "mean_token_accuracy": 0.19314661026000976,
      "num_tokens": 59830344.0,
      "step": 27310
    },
    {
      "entropy": 5.617011880874633,
      "epoch": 2.710897181421199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004270543764486275,
      "loss": 4.9981,
      "mean_token_accuracy": 0.19212463051080703,
      "num_tokens": 59843498.0,
      "step": 27315
    },
    {
      "entropy": 5.717956399917602,
      "epoch": 2.7113934100833665,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004270282584475903,
      "loss": 5.0525,
      "mean_token_accuracy": 0.18314781188964843,
      "num_tokens": 59854633.0,
      "step": 27320
    },
    {
      "entropy": 5.6491364479064945,
      "epoch": 2.711889638745534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042700213667657983,
      "loss": 5.1093,
      "mean_token_accuracy": 0.18094449788331984,
      "num_tokens": 59865925.0,
      "step": 27325
    },
    {
      "entropy": 5.621878385543823,
      "epoch": 2.7123858674077015,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000426976011136244,
      "loss": 4.9996,
      "mean_token_accuracy": 0.18517425805330276,
      "num_tokens": 59877196.0,
      "step": 27330
    },
    {
      "entropy": 5.697592878341675,
      "epoch": 2.712882096069869,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042694988182723047,
      "loss": 5.0788,
      "mean_token_accuracy": 0.1753239795565605,
      "num_tokens": 59887411.0,
      "step": 27335
    },
    {
      "entropy": 5.629831790924072,
      "epoch": 2.7133783247320364,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042692374875018735,
      "loss": 5.0365,
      "mean_token_accuracy": 0.1911686033010483,
      "num_tokens": 59899123.0,
      "step": 27340
    },
    {
      "entropy": 5.589451885223388,
      "epoch": 2.713874553394204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042689761190576263,
      "loss": 4.9827,
      "mean_token_accuracy": 0.19020217955112456,
      "num_tokens": 59910326.0,
      "step": 27345
    },
    {
      "entropy": 5.685828590393067,
      "epoch": 2.7143707820563714,
      "grad_norm": 1.125,
      "learning_rate": 0.00042687147129460447,
      "loss": 5.0748,
      "mean_token_accuracy": 0.1786923199892044,
      "num_tokens": 59920722.0,
      "step": 27350
    },
    {
      "entropy": 5.716787719726563,
      "epoch": 2.714867010718539,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042684532691736104,
      "loss": 5.0494,
      "mean_token_accuracy": 0.18094804435968398,
      "num_tokens": 59932568.0,
      "step": 27355
    },
    {
      "entropy": 5.6648579120635985,
      "epoch": 2.715363239380707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042681917877468076,
      "loss": 5.0976,
      "mean_token_accuracy": 0.17399242371320725,
      "num_tokens": 59943805.0,
      "step": 27360
    },
    {
      "entropy": 5.641012382507324,
      "epoch": 2.7158594680428743,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042679302686721187,
      "loss": 4.9845,
      "mean_token_accuracy": 0.1907743588089943,
      "num_tokens": 59954455.0,
      "step": 27365
    },
    {
      "entropy": 5.650668382644653,
      "epoch": 2.7163556967050417,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004267668711956031,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18545941412448883,
      "num_tokens": 59966722.0,
      "step": 27370
    },
    {
      "entropy": 5.639348745346069,
      "epoch": 2.716851925367209,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042674071176050296,
      "loss": 5.0182,
      "mean_token_accuracy": 0.18588501811027527,
      "num_tokens": 59976702.0,
      "step": 27375
    },
    {
      "entropy": 5.62450270652771,
      "epoch": 2.7173481540293767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042671454856256,
      "loss": 4.9874,
      "mean_token_accuracy": 0.1855453446507454,
      "num_tokens": 59987038.0,
      "step": 27380
    },
    {
      "entropy": 5.66150016784668,
      "epoch": 2.717844382691544,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004266883816024233,
      "loss": 5.0487,
      "mean_token_accuracy": 0.18267505317926408,
      "num_tokens": 59996755.0,
      "step": 27385
    },
    {
      "entropy": 5.650114917755127,
      "epoch": 2.7183406113537116,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004266622108807415,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18280827701091767,
      "num_tokens": 60006419.0,
      "step": 27390
    },
    {
      "entropy": 5.6799764156341555,
      "epoch": 2.7188368400158796,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004266360363981637,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18523354530334474,
      "num_tokens": 60017793.0,
      "step": 27395
    },
    {
      "entropy": 5.628332185745239,
      "epoch": 2.719333068678047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042660985815533905,
      "loss": 4.97,
      "mean_token_accuracy": 0.18482168912887573,
      "num_tokens": 60029392.0,
      "step": 27400
    },
    {
      "entropy": 5.667222881317139,
      "epoch": 2.7198292973402145,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042658367615291644,
      "loss": 5.0965,
      "mean_token_accuracy": 0.18446321785449982,
      "num_tokens": 60040467.0,
      "step": 27405
    },
    {
      "entropy": 5.629010438919067,
      "epoch": 2.720325526002382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004265574903915454,
      "loss": 5.0301,
      "mean_token_accuracy": 0.17990383207798005,
      "num_tokens": 60051583.0,
      "step": 27410
    },
    {
      "entropy": 5.648243808746338,
      "epoch": 2.7208217546645495,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004265313008718751,
      "loss": 5.0399,
      "mean_token_accuracy": 0.17777639329433442,
      "num_tokens": 60062218.0,
      "step": 27415
    },
    {
      "entropy": 5.625511837005615,
      "epoch": 2.721317983326717,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042650510759455516,
      "loss": 5.0209,
      "mean_token_accuracy": 0.1897749647498131,
      "num_tokens": 60074455.0,
      "step": 27420
    },
    {
      "entropy": 5.672637319564819,
      "epoch": 2.7218142119888844,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042647891056023495,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18233124166727066,
      "num_tokens": 60085442.0,
      "step": 27425
    },
    {
      "entropy": 5.6928369998931885,
      "epoch": 2.722310440651052,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042645270976956426,
      "loss": 5.0844,
      "mean_token_accuracy": 0.18397756069898605,
      "num_tokens": 60096441.0,
      "step": 27430
    },
    {
      "entropy": 5.71308650970459,
      "epoch": 2.7228066693132194,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042642650522319253,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18740772902965547,
      "num_tokens": 60107399.0,
      "step": 27435
    },
    {
      "entropy": 5.675365304946899,
      "epoch": 2.723302897975387,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004264002969217699,
      "loss": 5.0187,
      "mean_token_accuracy": 0.18672911822795868,
      "num_tokens": 60117568.0,
      "step": 27440
    },
    {
      "entropy": 5.68816385269165,
      "epoch": 2.7237991266375547,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004263740848659461,
      "loss": 5.0593,
      "mean_token_accuracy": 0.18431686013936996,
      "num_tokens": 60127214.0,
      "step": 27445
    },
    {
      "entropy": 5.636187982559204,
      "epoch": 2.724295355299722,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042634786905637123,
      "loss": 4.9957,
      "mean_token_accuracy": 0.18509362041950225,
      "num_tokens": 60138167.0,
      "step": 27450
    },
    {
      "entropy": 5.6106343269348145,
      "epoch": 2.7247915839618897,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004263216494936953,
      "loss": 4.9919,
      "mean_token_accuracy": 0.18398575633764266,
      "num_tokens": 60148841.0,
      "step": 27455
    },
    {
      "entropy": 5.6058063983917235,
      "epoch": 2.725287812624057,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042629542617856855,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18217118829488754,
      "num_tokens": 60159512.0,
      "step": 27460
    },
    {
      "entropy": 5.631308317184448,
      "epoch": 2.7257840412862246,
      "grad_norm": 1.125,
      "learning_rate": 0.0004262691991116412,
      "loss": 5.009,
      "mean_token_accuracy": 0.18100976347923278,
      "num_tokens": 60170463.0,
      "step": 27465
    },
    {
      "entropy": 5.57224702835083,
      "epoch": 2.726280269948392,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004262429682935637,
      "loss": 4.8818,
      "mean_token_accuracy": 0.19267417043447493,
      "num_tokens": 60180431.0,
      "step": 27470
    },
    {
      "entropy": 5.63006591796875,
      "epoch": 2.7267764986105596,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004262167337249865,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18035856485366822,
      "num_tokens": 60190410.0,
      "step": 27475
    },
    {
      "entropy": 5.6305821418762205,
      "epoch": 2.7272727272727275,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004261904954065601,
      "loss": 4.9853,
      "mean_token_accuracy": 0.182123064994812,
      "num_tokens": 60201909.0,
      "step": 27480
    },
    {
      "entropy": 5.707838439941407,
      "epoch": 2.727768955934895,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004261642533389352,
      "loss": 4.9876,
      "mean_token_accuracy": 0.18645013570785524,
      "num_tokens": 60212577.0,
      "step": 27485
    },
    {
      "entropy": 5.641513776779175,
      "epoch": 2.7282651845970625,
      "grad_norm": 1.125,
      "learning_rate": 0.00042613800752276264,
      "loss": 5.0808,
      "mean_token_accuracy": 0.18145448118448257,
      "num_tokens": 60222615.0,
      "step": 27490
    },
    {
      "entropy": 5.635886001586914,
      "epoch": 2.72876141325923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042611175795869306,
      "loss": 5.1053,
      "mean_token_accuracy": 0.17835942655801773,
      "num_tokens": 60234338.0,
      "step": 27495
    },
    {
      "entropy": 5.6891851902008055,
      "epoch": 2.7292576419213974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004260855046473775,
      "loss": 4.9809,
      "mean_token_accuracy": 0.1877891391515732,
      "num_tokens": 60244781.0,
      "step": 27500
    },
    {
      "entropy": 5.695287895202637,
      "epoch": 2.729753870583565,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042605924758946706,
      "loss": 5.0593,
      "mean_token_accuracy": 0.1819610044360161,
      "num_tokens": 60255150.0,
      "step": 27505
    },
    {
      "entropy": 5.626006460189819,
      "epoch": 2.7302500992457324,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042603298678561277,
      "loss": 5.0203,
      "mean_token_accuracy": 0.1839043378829956,
      "num_tokens": 60266073.0,
      "step": 27510
    },
    {
      "entropy": 5.657427835464477,
      "epoch": 2.7307463279079,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004260067222364659,
      "loss": 5.0656,
      "mean_token_accuracy": 0.1792394071817398,
      "num_tokens": 60277520.0,
      "step": 27515
    },
    {
      "entropy": 5.679428005218506,
      "epoch": 2.7312425565700673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042598045394267757,
      "loss": 5.044,
      "mean_token_accuracy": 0.18106625974178314,
      "num_tokens": 60288530.0,
      "step": 27520
    },
    {
      "entropy": 5.660749435424805,
      "epoch": 2.731738785232235,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042595418190489944,
      "loss": 4.977,
      "mean_token_accuracy": 0.18829621076583863,
      "num_tokens": 60299926.0,
      "step": 27525
    },
    {
      "entropy": 5.719200277328492,
      "epoch": 2.7322350138944023,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042592790612378277,
      "loss": 5.151,
      "mean_token_accuracy": 0.17660026699304582,
      "num_tokens": 60312129.0,
      "step": 27530
    },
    {
      "entropy": 5.668578815460205,
      "epoch": 2.73273124255657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042590162659997933,
      "loss": 5.0546,
      "mean_token_accuracy": 0.18027278482913972,
      "num_tokens": 60322405.0,
      "step": 27535
    },
    {
      "entropy": 5.717739248275757,
      "epoch": 2.7332274712187377,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004258753433341407,
      "loss": 5.0464,
      "mean_token_accuracy": 0.1772638350725174,
      "num_tokens": 60333112.0,
      "step": 27540
    },
    {
      "entropy": 5.680772590637207,
      "epoch": 2.733723699880905,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042584905632691867,
      "loss": 5.0155,
      "mean_token_accuracy": 0.1847972720861435,
      "num_tokens": 60343500.0,
      "step": 27545
    },
    {
      "entropy": 5.572921323776245,
      "epoch": 2.7342199285430726,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004258227655789651,
      "loss": 5.01,
      "mean_token_accuracy": 0.1890001878142357,
      "num_tokens": 60354711.0,
      "step": 27550
    },
    {
      "entropy": 5.610660362243652,
      "epoch": 2.73471615720524,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004257964710909318,
      "loss": 5.0336,
      "mean_token_accuracy": 0.17590565830469132,
      "num_tokens": 60365064.0,
      "step": 27555
    },
    {
      "entropy": 5.660860681533814,
      "epoch": 2.7352123858674076,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042577017286347116,
      "loss": 5.0511,
      "mean_token_accuracy": 0.18102008402347564,
      "num_tokens": 60376425.0,
      "step": 27560
    },
    {
      "entropy": 5.6578747749328615,
      "epoch": 2.7357086145295755,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042574387089723504,
      "loss": 4.95,
      "mean_token_accuracy": 0.18873940110206605,
      "num_tokens": 60386745.0,
      "step": 27565
    },
    {
      "entropy": 5.628534412384033,
      "epoch": 2.736204843191743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042571756519287567,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18237075060606003,
      "num_tokens": 60397038.0,
      "step": 27570
    },
    {
      "entropy": 5.620749473571777,
      "epoch": 2.7367010718539104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042569125575104557,
      "loss": 5.0168,
      "mean_token_accuracy": 0.18325221836566924,
      "num_tokens": 60406955.0,
      "step": 27575
    },
    {
      "entropy": 5.654900169372558,
      "epoch": 2.737197300516078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042566494257239695,
      "loss": 5.0266,
      "mean_token_accuracy": 0.18520811051130295,
      "num_tokens": 60418245.0,
      "step": 27580
    },
    {
      "entropy": 5.6869340419769285,
      "epoch": 2.7376935291782454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004256386256575824,
      "loss": 5.0317,
      "mean_token_accuracy": 0.1777902737259865,
      "num_tokens": 60429310.0,
      "step": 27585
    },
    {
      "entropy": 5.676470518112183,
      "epoch": 2.738189757840413,
      "grad_norm": 1.125,
      "learning_rate": 0.0004256123050072546,
      "loss": 5.0778,
      "mean_token_accuracy": 0.18465646058321,
      "num_tokens": 60441355.0,
      "step": 27590
    },
    {
      "entropy": 5.739621353149414,
      "epoch": 2.7386859865025803,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042558598062206616,
      "loss": 5.2051,
      "mean_token_accuracy": 0.1754053235054016,
      "num_tokens": 60452853.0,
      "step": 27595
    },
    {
      "entropy": 5.659017848968506,
      "epoch": 2.739182215164748,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042555965250266984,
      "loss": 5.0367,
      "mean_token_accuracy": 0.18338130414485931,
      "num_tokens": 60462934.0,
      "step": 27600
    },
    {
      "entropy": 5.714774560928345,
      "epoch": 2.7396784438269153,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042553332064971866,
      "loss": 5.1122,
      "mean_token_accuracy": 0.17438598871231079,
      "num_tokens": 60475575.0,
      "step": 27605
    },
    {
      "entropy": 5.5667520523071286,
      "epoch": 2.7401746724890828,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042550698506386536,
      "loss": 4.9004,
      "mean_token_accuracy": 0.20030098408460617,
      "num_tokens": 60486495.0,
      "step": 27610
    },
    {
      "entropy": 5.614031839370727,
      "epoch": 2.7406709011512502,
      "grad_norm": 1.25,
      "learning_rate": 0.00042548064574576323,
      "loss": 5.0166,
      "mean_token_accuracy": 0.1899981677532196,
      "num_tokens": 60496806.0,
      "step": 27615
    },
    {
      "entropy": 5.661921167373658,
      "epoch": 2.741167129813418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004254543026960654,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18640949726104736,
      "num_tokens": 60508905.0,
      "step": 27620
    },
    {
      "entropy": 5.605615282058716,
      "epoch": 2.7416633584755856,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004254279559154249,
      "loss": 4.9786,
      "mean_token_accuracy": 0.18297578692436217,
      "num_tokens": 60519603.0,
      "step": 27625
    },
    {
      "entropy": 5.639845752716065,
      "epoch": 2.742159587137753,
      "grad_norm": 1.0,
      "learning_rate": 0.00042540160540449536,
      "loss": 5.0602,
      "mean_token_accuracy": 0.18271466344594955,
      "num_tokens": 60530977.0,
      "step": 27630
    },
    {
      "entropy": 5.636430263519287,
      "epoch": 2.7426558157999206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004253752511639301,
      "loss": 5.0615,
      "mean_token_accuracy": 0.17958287000656128,
      "num_tokens": 60541354.0,
      "step": 27635
    },
    {
      "entropy": 5.641478538513184,
      "epoch": 2.743152044462088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004253488931943826,
      "loss": 5.0358,
      "mean_token_accuracy": 0.18718815892934798,
      "num_tokens": 60552531.0,
      "step": 27640
    },
    {
      "entropy": 5.59478874206543,
      "epoch": 2.7436482731242555,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042532253149650653,
      "loss": 5.0087,
      "mean_token_accuracy": 0.18893203437328338,
      "num_tokens": 60562552.0,
      "step": 27645
    },
    {
      "entropy": 5.649020957946777,
      "epoch": 2.7441445017864234,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004252961660709556,
      "loss": 5.0264,
      "mean_token_accuracy": 0.17636396884918212,
      "num_tokens": 60572849.0,
      "step": 27650
    },
    {
      "entropy": 5.646630716323853,
      "epoch": 2.744640730448591,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042526979691838353,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18879534751176835,
      "num_tokens": 60584095.0,
      "step": 27655
    },
    {
      "entropy": 5.679950428009033,
      "epoch": 2.7451369591107584,
      "grad_norm": 1.125,
      "learning_rate": 0.00042524342403944446,
      "loss": 5.0707,
      "mean_token_accuracy": 0.18624247163534163,
      "num_tokens": 60594353.0,
      "step": 27660
    },
    {
      "entropy": 5.702851009368897,
      "epoch": 2.745633187772926,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042521704743479213,
      "loss": 5.0599,
      "mean_token_accuracy": 0.18375449180603026,
      "num_tokens": 60604558.0,
      "step": 27665
    },
    {
      "entropy": 5.6237585067749025,
      "epoch": 2.7461294164350933,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042519066710508064,
      "loss": 5.0129,
      "mean_token_accuracy": 0.17831090539693834,
      "num_tokens": 60615670.0,
      "step": 27670
    },
    {
      "entropy": 5.569155216217041,
      "epoch": 2.746625645097261,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042516428305096435,
      "loss": 4.9521,
      "mean_token_accuracy": 0.18535391241312027,
      "num_tokens": 60626452.0,
      "step": 27675
    },
    {
      "entropy": 5.6337682723999025,
      "epoch": 2.7471218737594283,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004251378952730973,
      "loss": 5.0298,
      "mean_token_accuracy": 0.1872106373310089,
      "num_tokens": 60637122.0,
      "step": 27680
    },
    {
      "entropy": 5.683755683898926,
      "epoch": 2.7476181024215958,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004251115037721341,
      "loss": 5.0121,
      "mean_token_accuracy": 0.18926577866077424,
      "num_tokens": 60648835.0,
      "step": 27685
    },
    {
      "entropy": 5.6811925888061525,
      "epoch": 2.7481143310837632,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042508510854872905,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1784917965531349,
      "num_tokens": 60658537.0,
      "step": 27690
    },
    {
      "entropy": 5.666929292678833,
      "epoch": 2.7486105597459307,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00042505870960353665,
      "loss": 5.0369,
      "mean_token_accuracy": 0.18836702555418014,
      "num_tokens": 60670450.0,
      "step": 27695
    },
    {
      "entropy": 5.7231298923492435,
      "epoch": 2.749106788408098,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004250323069372116,
      "loss": 5.0486,
      "mean_token_accuracy": 0.18878818601369857,
      "num_tokens": 60680539.0,
      "step": 27700
    },
    {
      "entropy": 5.6939146518707275,
      "epoch": 2.749603017070266,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042500590055040866,
      "loss": 5.1233,
      "mean_token_accuracy": 0.18045013695955275,
      "num_tokens": 60693106.0,
      "step": 27705
    },
    {
      "entropy": 5.726310062408447,
      "epoch": 2.7500992457324336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042497949044378267,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17934447526931763,
      "num_tokens": 60703509.0,
      "step": 27710
    },
    {
      "entropy": 5.688462781906128,
      "epoch": 2.750595474394601,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042495307661798837,
      "loss": 5.0842,
      "mean_token_accuracy": 0.17858813256025313,
      "num_tokens": 60714113.0,
      "step": 27715
    },
    {
      "entropy": 5.750570964813233,
      "epoch": 2.7510917030567685,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042492665907368094,
      "loss": 5.1288,
      "mean_token_accuracy": 0.17728557288646699,
      "num_tokens": 60726826.0,
      "step": 27720
    },
    {
      "entropy": 5.659419059753418,
      "epoch": 2.751587931718936,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042490023781151546,
      "loss": 5.0453,
      "mean_token_accuracy": 0.18173159062862396,
      "num_tokens": 60738443.0,
      "step": 27725
    },
    {
      "entropy": 5.6280069828033445,
      "epoch": 2.7520841603811035,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042487381283214706,
      "loss": 5.0051,
      "mean_token_accuracy": 0.18093528747558593,
      "num_tokens": 60748865.0,
      "step": 27730
    },
    {
      "entropy": 5.662254333496094,
      "epoch": 2.7525803890432714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004248473841362311,
      "loss": 5.0353,
      "mean_token_accuracy": 0.18826731145381928,
      "num_tokens": 60760001.0,
      "step": 27735
    },
    {
      "entropy": 5.6983846664428714,
      "epoch": 2.753076617705439,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042482095172442286,
      "loss": 5.0779,
      "mean_token_accuracy": 0.18748919665813446,
      "num_tokens": 60772003.0,
      "step": 27740
    },
    {
      "entropy": 5.618401765823364,
      "epoch": 2.7535728463676064,
      "grad_norm": 1.125,
      "learning_rate": 0.0004247945155973778,
      "loss": 4.9599,
      "mean_token_accuracy": 0.19284712970256807,
      "num_tokens": 60783167.0,
      "step": 27745
    },
    {
      "entropy": 5.650554037094116,
      "epoch": 2.754069075029774,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004247680757557516,
      "loss": 5.058,
      "mean_token_accuracy": 0.18472394198179246,
      "num_tokens": 60793917.0,
      "step": 27750
    },
    {
      "entropy": 5.737556600570679,
      "epoch": 2.7545653036919413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042474163220019987,
      "loss": 5.1534,
      "mean_token_accuracy": 0.17327350229024888,
      "num_tokens": 60805166.0,
      "step": 27755
    },
    {
      "entropy": 5.7650199890136715,
      "epoch": 2.7550615323541088,
      "grad_norm": 1.125,
      "learning_rate": 0.00042471518493137836,
      "loss": 5.0758,
      "mean_token_accuracy": 0.1865039959549904,
      "num_tokens": 60816984.0,
      "step": 27760
    },
    {
      "entropy": 5.619897127151489,
      "epoch": 2.7555577610162763,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004246887339499427,
      "loss": 4.9999,
      "mean_token_accuracy": 0.1847267121076584,
      "num_tokens": 60829481.0,
      "step": 27765
    },
    {
      "entropy": 5.6593985080719,
      "epoch": 2.7560539896784437,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042466227925654917,
      "loss": 5.0001,
      "mean_token_accuracy": 0.1872779607772827,
      "num_tokens": 60839376.0,
      "step": 27770
    },
    {
      "entropy": 5.606841039657593,
      "epoch": 2.756550218340611,
      "grad_norm": 1.125,
      "learning_rate": 0.0004246358208518535,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18153008073568344,
      "num_tokens": 60850383.0,
      "step": 27775
    },
    {
      "entropy": 5.634746980667114,
      "epoch": 2.7570464470027787,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042460935873651196,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18367213755846024,
      "num_tokens": 60861249.0,
      "step": 27780
    },
    {
      "entropy": 5.6717511177062985,
      "epoch": 2.757542675664946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042458289291118066,
      "loss": 5.0345,
      "mean_token_accuracy": 0.1825373187661171,
      "num_tokens": 60872220.0,
      "step": 27785
    },
    {
      "entropy": 5.621320819854736,
      "epoch": 2.758038904327114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042455642337651595,
      "loss": 5.0249,
      "mean_token_accuracy": 0.1811293661594391,
      "num_tokens": 60882006.0,
      "step": 27790
    },
    {
      "entropy": 5.629112005233765,
      "epoch": 2.7585351329892815,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004245299501331743,
      "loss": 4.9537,
      "mean_token_accuracy": 0.19372481405735015,
      "num_tokens": 60891732.0,
      "step": 27795
    },
    {
      "entropy": 5.699230051040649,
      "epoch": 2.759031361651449,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042450347318181194,
      "loss": 5.0797,
      "mean_token_accuracy": 0.18079608380794526,
      "num_tokens": 60902990.0,
      "step": 27800
    },
    {
      "entropy": 5.6542352676391605,
      "epoch": 2.7595275903136165,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042447699252308566,
      "loss": 5.0257,
      "mean_token_accuracy": 0.17648455500602722,
      "num_tokens": 60914673.0,
      "step": 27805
    },
    {
      "entropy": 5.617784833908081,
      "epoch": 2.760023818975784,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004244505081576521,
      "loss": 4.9757,
      "mean_token_accuracy": 0.18967270702123643,
      "num_tokens": 60925872.0,
      "step": 27810
    },
    {
      "entropy": 5.638317346572876,
      "epoch": 2.7605200476379514,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042442402008616793,
      "loss": 5.083,
      "mean_token_accuracy": 0.18043978959321977,
      "num_tokens": 60936589.0,
      "step": 27815
    },
    {
      "entropy": 5.642530727386474,
      "epoch": 2.7610162763001194,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042439752830929,
      "loss": 5.0292,
      "mean_token_accuracy": 0.1806761920452118,
      "num_tokens": 60947213.0,
      "step": 27820
    },
    {
      "entropy": 5.703270721435547,
      "epoch": 2.761512504962287,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004243710328276754,
      "loss": 5.0338,
      "mean_token_accuracy": 0.18663505464792252,
      "num_tokens": 60958615.0,
      "step": 27825
    },
    {
      "entropy": 5.6436793327331545,
      "epoch": 2.7620087336244543,
      "grad_norm": 1.09375,
      "learning_rate": 0.000424344533641981,
      "loss": 4.9897,
      "mean_token_accuracy": 0.18960287123918534,
      "num_tokens": 60968830.0,
      "step": 27830
    },
    {
      "entropy": 5.617725658416748,
      "epoch": 2.762504962286622,
      "grad_norm": 1.09375,
      "learning_rate": 0.000424318030752864,
      "loss": 4.9574,
      "mean_token_accuracy": 0.1845041885972023,
      "num_tokens": 60979744.0,
      "step": 27835
    },
    {
      "entropy": 5.6814470291137695,
      "epoch": 2.7630011909487893,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004242915241609815,
      "loss": 5.0562,
      "mean_token_accuracy": 0.17685190886259078,
      "num_tokens": 60990291.0,
      "step": 27840
    },
    {
      "entropy": 5.670430469512939,
      "epoch": 2.7634974196109567,
      "grad_norm": 1.125,
      "learning_rate": 0.000424265013866991,
      "loss": 5.0216,
      "mean_token_accuracy": 0.1833343520760536,
      "num_tokens": 61002170.0,
      "step": 27845
    },
    {
      "entropy": 5.629031944274902,
      "epoch": 2.763993648273124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004242384998715497,
      "loss": 5.0601,
      "mean_token_accuracy": 0.1759025290608406,
      "num_tokens": 61012639.0,
      "step": 27850
    },
    {
      "entropy": 5.701525020599365,
      "epoch": 2.7644898769352917,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004242119821753153,
      "loss": 5.025,
      "mean_token_accuracy": 0.1804855927824974,
      "num_tokens": 61022657.0,
      "step": 27855
    },
    {
      "entropy": 5.640690851211548,
      "epoch": 2.764986105597459,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004241854607789452,
      "loss": 4.9696,
      "mean_token_accuracy": 0.18924693167209625,
      "num_tokens": 61033108.0,
      "step": 27860
    },
    {
      "entropy": 5.599376821517945,
      "epoch": 2.7654823342596266,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042415893568309714,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18349212855100633,
      "num_tokens": 61043651.0,
      "step": 27865
    },
    {
      "entropy": 5.563090753555298,
      "epoch": 2.765978562921794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00042413240688842884,
      "loss": 4.9518,
      "mean_token_accuracy": 0.18834821730852128,
      "num_tokens": 61053234.0,
      "step": 27870
    },
    {
      "entropy": 5.6230003356933596,
      "epoch": 2.766474791583962,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004241058743955982,
      "loss": 5.0061,
      "mean_token_accuracy": 0.18268449902534484,
      "num_tokens": 61065178.0,
      "step": 27875
    },
    {
      "entropy": 5.629685306549073,
      "epoch": 2.7669710202461295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004240793382052633,
      "loss": 4.9646,
      "mean_token_accuracy": 0.1971358075737953,
      "num_tokens": 61075709.0,
      "step": 27880
    },
    {
      "entropy": 5.667858457565307,
      "epoch": 2.767467248908297,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000424052798318082,
      "loss": 5.0803,
      "mean_token_accuracy": 0.17230141013860703,
      "num_tokens": 61086880.0,
      "step": 27885
    },
    {
      "entropy": 5.620935583114624,
      "epoch": 2.7679634775704645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004240262547347124,
      "loss": 5.033,
      "mean_token_accuracy": 0.18749147951602935,
      "num_tokens": 61097008.0,
      "step": 27890
    },
    {
      "entropy": 5.649658298492431,
      "epoch": 2.768459706232632,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042399970745581283,
      "loss": 5.0023,
      "mean_token_accuracy": 0.18510909527540206,
      "num_tokens": 61107844.0,
      "step": 27895
    },
    {
      "entropy": 5.615605306625366,
      "epoch": 2.7689559348947994,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042397315648204164,
      "loss": 4.9723,
      "mean_token_accuracy": 0.19360601156949997,
      "num_tokens": 61118652.0,
      "step": 27900
    },
    {
      "entropy": 5.700351285934448,
      "epoch": 2.7694521635569673,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042394660181405716,
      "loss": 5.0951,
      "mean_token_accuracy": 0.18013325780630113,
      "num_tokens": 61129143.0,
      "step": 27905
    },
    {
      "entropy": 5.663667678833008,
      "epoch": 2.769948392219135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004239200434525179,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18484596610069276,
      "num_tokens": 61140623.0,
      "step": 27910
    },
    {
      "entropy": 5.6142536163330075,
      "epoch": 2.7704446208813023,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004238934813980824,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18010045886039733,
      "num_tokens": 61150998.0,
      "step": 27915
    },
    {
      "entropy": 5.649796009063721,
      "epoch": 2.7709408495434698,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042386691565140947,
      "loss": 5.025,
      "mean_token_accuracy": 0.1880584880709648,
      "num_tokens": 61162086.0,
      "step": 27920
    },
    {
      "entropy": 5.674909067153931,
      "epoch": 2.7714370782056372,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042384034621315773,
      "loss": 5.0114,
      "mean_token_accuracy": 0.18336241841316223,
      "num_tokens": 61171806.0,
      "step": 27925
    },
    {
      "entropy": 5.658399486541748,
      "epoch": 2.7719333068678047,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004238137730839862,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18632082641124725,
      "num_tokens": 61182056.0,
      "step": 27930
    },
    {
      "entropy": 5.678248167037964,
      "epoch": 2.772429535529972,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004237871962645538,
      "loss": 5.0207,
      "mean_token_accuracy": 0.18955037146806716,
      "num_tokens": 61193231.0,
      "step": 27935
    },
    {
      "entropy": 5.610466003417969,
      "epoch": 2.7729257641921397,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042376061575551946,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18500090539455413,
      "num_tokens": 61203785.0,
      "step": 27940
    },
    {
      "entropy": 5.614484977722168,
      "epoch": 2.773421992854307,
      "grad_norm": 1.0,
      "learning_rate": 0.0004237340315575425,
      "loss": 5.0119,
      "mean_token_accuracy": 0.18843696117401124,
      "num_tokens": 61215244.0,
      "step": 27945
    },
    {
      "entropy": 5.687121629714966,
      "epoch": 2.7739182215164746,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004237074436712819,
      "loss": 4.9914,
      "mean_token_accuracy": 0.18104442059993744,
      "num_tokens": 61227147.0,
      "step": 27950
    },
    {
      "entropy": 5.702968788146973,
      "epoch": 2.774414450178642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042368085209739715,
      "loss": 5.0528,
      "mean_token_accuracy": 0.18729435205459594,
      "num_tokens": 61237687.0,
      "step": 27955
    },
    {
      "entropy": 5.561707592010498,
      "epoch": 2.77491067884081,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004236542568365478,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18754327297210693,
      "num_tokens": 61247700.0,
      "step": 27960
    },
    {
      "entropy": 5.662714004516602,
      "epoch": 2.7754069075029775,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00042362765788939307,
      "loss": 4.9831,
      "mean_token_accuracy": 0.181393164396286,
      "num_tokens": 61258407.0,
      "step": 27965
    },
    {
      "entropy": 5.727023887634277,
      "epoch": 2.775903136165145,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004236010552565927,
      "loss": 5.1337,
      "mean_token_accuracy": 0.1763608142733574,
      "num_tokens": 61269065.0,
      "step": 27970
    },
    {
      "entropy": 5.694483280181885,
      "epoch": 2.7763993648273124,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042357444893880645,
      "loss": 5.0257,
      "mean_token_accuracy": 0.18069480061531068,
      "num_tokens": 61280767.0,
      "step": 27975
    },
    {
      "entropy": 5.597509384155273,
      "epoch": 2.77689559348948,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042354783893669385,
      "loss": 4.9727,
      "mean_token_accuracy": 0.18225322663784027,
      "num_tokens": 61291967.0,
      "step": 27980
    },
    {
      "entropy": 5.68665714263916,
      "epoch": 2.7773918221516474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004235212252509151,
      "loss": 5.144,
      "mean_token_accuracy": 0.17410053163766862,
      "num_tokens": 61303572.0,
      "step": 27985
    },
    {
      "entropy": 5.703855323791504,
      "epoch": 2.777888050813815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042349460788212995,
      "loss": 5.0475,
      "mean_token_accuracy": 0.18361503034830093,
      "num_tokens": 61315674.0,
      "step": 27990
    },
    {
      "entropy": 5.688141107559204,
      "epoch": 2.7783842794759828,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042346798683099844,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18339960277080536,
      "num_tokens": 61326334.0,
      "step": 27995
    },
    {
      "entropy": 5.653124570846558,
      "epoch": 2.7788805081381502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004234413620981809,
      "loss": 5.1171,
      "mean_token_accuracy": 0.18028786033391953,
      "num_tokens": 61337505.0,
      "step": 28000
    },
    {
      "entropy": 5.655402040481567,
      "epoch": 2.7793767368003177,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00042341473368433735,
      "loss": 5.0637,
      "mean_token_accuracy": 0.18797844648361206,
      "num_tokens": 61348451.0,
      "step": 28005
    },
    {
      "entropy": 5.742032194137574,
      "epoch": 2.779872965462485,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004233881015901282,
      "loss": 5.2016,
      "mean_token_accuracy": 0.1719684988260269,
      "num_tokens": 61358356.0,
      "step": 28010
    },
    {
      "entropy": 5.782065153121948,
      "epoch": 2.7803691941246527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004233614658162139,
      "loss": 5.0489,
      "mean_token_accuracy": 0.18857255280017854,
      "num_tokens": 61370996.0,
      "step": 28015
    },
    {
      "entropy": 5.738521862030029,
      "epoch": 2.78086542278682,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042333482636325504,
      "loss": 5.0738,
      "mean_token_accuracy": 0.1798695906996727,
      "num_tokens": 61383928.0,
      "step": 28020
    },
    {
      "entropy": 5.682621335983276,
      "epoch": 2.7813616514489876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000423308183231912,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1832788273692131,
      "num_tokens": 61394277.0,
      "step": 28025
    },
    {
      "entropy": 5.602226686477661,
      "epoch": 2.781857880111155,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042328153642284567,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18922929018735885,
      "num_tokens": 61405928.0,
      "step": 28030
    },
    {
      "entropy": 5.705266952514648,
      "epoch": 2.7823541087733226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004232548859367167,
      "loss": 5.0922,
      "mean_token_accuracy": 0.17926039248704911,
      "num_tokens": 61417893.0,
      "step": 28035
    },
    {
      "entropy": 5.686022520065308,
      "epoch": 2.78285033743549,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000423228231774186,
      "loss": 5.0554,
      "mean_token_accuracy": 0.1812805712223053,
      "num_tokens": 61428099.0,
      "step": 28040
    },
    {
      "entropy": 5.6475804328918455,
      "epoch": 2.783346566097658,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004232015739359146,
      "loss": 4.9739,
      "mean_token_accuracy": 0.19056009501218796,
      "num_tokens": 61437636.0,
      "step": 28045
    },
    {
      "entropy": 5.601137113571167,
      "epoch": 2.7838427947598254,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042317491242256346,
      "loss": 4.9603,
      "mean_token_accuracy": 0.18171384185552597,
      "num_tokens": 61448038.0,
      "step": 28050
    },
    {
      "entropy": 5.632596778869629,
      "epoch": 2.784339023421993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004231482472347938,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18116715550422668,
      "num_tokens": 61460258.0,
      "step": 28055
    },
    {
      "entropy": 5.634225797653198,
      "epoch": 2.7848352520841604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004231215783732668,
      "loss": 5.0354,
      "mean_token_accuracy": 0.18490842282772063,
      "num_tokens": 61472274.0,
      "step": 28060
    },
    {
      "entropy": 5.658372640609741,
      "epoch": 2.785331480746328,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042309490583864395,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18912916332483293,
      "num_tokens": 61482975.0,
      "step": 28065
    },
    {
      "entropy": 5.695724964141846,
      "epoch": 2.7858277094084953,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004230682296315865,
      "loss": 5.0404,
      "mean_token_accuracy": 0.17615651786327363,
      "num_tokens": 61494223.0,
      "step": 28070
    },
    {
      "entropy": 5.7482240200042725,
      "epoch": 2.786323938070663,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000423041549752756,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1801214650273323,
      "num_tokens": 61505981.0,
      "step": 28075
    },
    {
      "entropy": 5.592228698730469,
      "epoch": 2.7868201667328307,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004230148662028141,
      "loss": 4.9795,
      "mean_token_accuracy": 0.19099920094013215,
      "num_tokens": 61516410.0,
      "step": 28080
    },
    {
      "entropy": 5.614271354675293,
      "epoch": 2.787316395394998,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042298817898242237,
      "loss": 4.9881,
      "mean_token_accuracy": 0.1818405419588089,
      "num_tokens": 61526551.0,
      "step": 28085
    },
    {
      "entropy": 5.709369564056397,
      "epoch": 2.7878126240571657,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042296148809224275,
      "loss": 5.1274,
      "mean_token_accuracy": 0.17715153247117996,
      "num_tokens": 61538214.0,
      "step": 28090
    },
    {
      "entropy": 5.6734415054321286,
      "epoch": 2.788308852719333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004229347935329371,
      "loss": 5.0783,
      "mean_token_accuracy": 0.17701806873083115,
      "num_tokens": 61549726.0,
      "step": 28095
    },
    {
      "entropy": 5.729780769348144,
      "epoch": 2.7888050813815006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042290809530516724,
      "loss": 5.1583,
      "mean_token_accuracy": 0.17137616276741027,
      "num_tokens": 61560070.0,
      "step": 28100
    },
    {
      "entropy": 5.741299772262574,
      "epoch": 2.789301310043668,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004228813934095954,
      "loss": 5.0907,
      "mean_token_accuracy": 0.17774919122457505,
      "num_tokens": 61571348.0,
      "step": 28105
    },
    {
      "entropy": 5.631013011932373,
      "epoch": 2.7897975387058356,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042285468784688363,
      "loss": 4.9616,
      "mean_token_accuracy": 0.18988422602415084,
      "num_tokens": 61581981.0,
      "step": 28110
    },
    {
      "entropy": 5.654374074935913,
      "epoch": 2.790293767368003,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042282797861769423,
      "loss": 5.0616,
      "mean_token_accuracy": 0.1832409456372261,
      "num_tokens": 61591248.0,
      "step": 28115
    },
    {
      "entropy": 5.659319257736206,
      "epoch": 2.7907899960301705,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004228012657226895,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1834576666355133,
      "num_tokens": 61602432.0,
      "step": 28120
    },
    {
      "entropy": 5.590486907958985,
      "epoch": 2.791286224692338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042277454916253186,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18399685025215148,
      "num_tokens": 61613418.0,
      "step": 28125
    },
    {
      "entropy": 5.627384185791016,
      "epoch": 2.791782453354506,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004227478289378838,
      "loss": 5.0242,
      "mean_token_accuracy": 0.1783231645822525,
      "num_tokens": 61623783.0,
      "step": 28130
    },
    {
      "entropy": 5.733232259750366,
      "epoch": 2.7922786820166734,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00042272110504940795,
      "loss": 5.1022,
      "mean_token_accuracy": 0.17916015386581421,
      "num_tokens": 61633838.0,
      "step": 28135
    },
    {
      "entropy": 5.663054323196411,
      "epoch": 2.792774910678841,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042269437749776703,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18778206408023834,
      "num_tokens": 61645982.0,
      "step": 28140
    },
    {
      "entropy": 5.649596261978149,
      "epoch": 2.7932711393410083,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004226676462836238,
      "loss": 4.967,
      "mean_token_accuracy": 0.18431106507778167,
      "num_tokens": 61656030.0,
      "step": 28145
    },
    {
      "entropy": 5.598747587203979,
      "epoch": 2.793767368003176,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004226409114076411,
      "loss": 4.9954,
      "mean_token_accuracy": 0.1823603630065918,
      "num_tokens": 61666746.0,
      "step": 28150
    },
    {
      "entropy": 5.699208450317383,
      "epoch": 2.7942635966653433,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042261417287048196,
      "loss": 5.154,
      "mean_token_accuracy": 0.1795966684818268,
      "num_tokens": 61677681.0,
      "step": 28155
    },
    {
      "entropy": 5.745540475845337,
      "epoch": 2.7947598253275108,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004225874306728094,
      "loss": 5.1068,
      "mean_token_accuracy": 0.17722053080797195,
      "num_tokens": 61689564.0,
      "step": 28160
    },
    {
      "entropy": 5.666355848312378,
      "epoch": 2.7952560539896787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042256068481528666,
      "loss": 5.0097,
      "mean_token_accuracy": 0.18427261859178543,
      "num_tokens": 61701074.0,
      "step": 28165
    },
    {
      "entropy": 5.5946142196655275,
      "epoch": 2.795752282651846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004225339352985768,
      "loss": 4.9984,
      "mean_token_accuracy": 0.187604159116745,
      "num_tokens": 61712203.0,
      "step": 28170
    },
    {
      "entropy": 5.65030460357666,
      "epoch": 2.7962485113140136,
      "grad_norm": 1.125,
      "learning_rate": 0.00042250718212334334,
      "loss": 4.9652,
      "mean_token_accuracy": 0.1840137153863907,
      "num_tokens": 61723362.0,
      "step": 28175
    },
    {
      "entropy": 5.633122253417969,
      "epoch": 2.796744739976181,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042248042529024956,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18947168588638305,
      "num_tokens": 61733454.0,
      "step": 28180
    },
    {
      "entropy": 5.686996698379517,
      "epoch": 2.7972409686383486,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042245366479995894,
      "loss": 5.1008,
      "mean_token_accuracy": 0.17714795619249343,
      "num_tokens": 61745014.0,
      "step": 28185
    },
    {
      "entropy": 5.64125690460205,
      "epoch": 2.797737197300516,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042242690065313535,
      "loss": 5.0448,
      "mean_token_accuracy": 0.18193560242652893,
      "num_tokens": 61755334.0,
      "step": 28190
    },
    {
      "entropy": 5.583362007141114,
      "epoch": 2.7982334259626835,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042240013285044215,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19004108607769013,
      "num_tokens": 61766024.0,
      "step": 28195
    },
    {
      "entropy": 5.6807159900665285,
      "epoch": 2.798729654624851,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042237336139254336,
      "loss": 5.0325,
      "mean_token_accuracy": 0.18865990191698073,
      "num_tokens": 61778228.0,
      "step": 28200
    },
    {
      "entropy": 5.689754819869995,
      "epoch": 2.7992258832870185,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042234658628010274,
      "loss": 5.0669,
      "mean_token_accuracy": 0.1767919421195984,
      "num_tokens": 61789072.0,
      "step": 28205
    },
    {
      "entropy": 5.693597078323364,
      "epoch": 2.799722111949186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042231980751378435,
      "loss": 5.0271,
      "mean_token_accuracy": 0.1874741718173027,
      "num_tokens": 61799503.0,
      "step": 28210
    },
    {
      "entropy": 5.642817974090576,
      "epoch": 2.8002183406113534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042229302509425205,
      "loss": 4.9777,
      "mean_token_accuracy": 0.18740370869636536,
      "num_tokens": 61811063.0,
      "step": 28215
    },
    {
      "entropy": 5.6783417701721195,
      "epoch": 2.8007145692735214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004222662390221702,
      "loss": 5.09,
      "mean_token_accuracy": 0.179064017534256,
      "num_tokens": 61822832.0,
      "step": 28220
    },
    {
      "entropy": 5.660078859329223,
      "epoch": 2.801210797935689,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004222394492982029,
      "loss": 5.0622,
      "mean_token_accuracy": 0.17942069470882416,
      "num_tokens": 61832456.0,
      "step": 28225
    },
    {
      "entropy": 5.647758913040161,
      "epoch": 2.8017070265978563,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004222126559230146,
      "loss": 4.9981,
      "mean_token_accuracy": 0.18654464483261107,
      "num_tokens": 61843043.0,
      "step": 28230
    },
    {
      "entropy": 5.6170923709869385,
      "epoch": 2.802203255260024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004221858588972696,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17734377831220627,
      "num_tokens": 61853249.0,
      "step": 28235
    },
    {
      "entropy": 5.5791278839111325,
      "epoch": 2.8026994839221913,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042215905822163245,
      "loss": 4.9666,
      "mean_token_accuracy": 0.1896164208650589,
      "num_tokens": 61863842.0,
      "step": 28240
    },
    {
      "entropy": 5.66473126411438,
      "epoch": 2.8031957125843587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004221322538967678,
      "loss": 5.057,
      "mean_token_accuracy": 0.18433575034141542,
      "num_tokens": 61876924.0,
      "step": 28245
    },
    {
      "entropy": 5.742309474945069,
      "epoch": 2.8036919412465267,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004221054459233403,
      "loss": 5.0936,
      "mean_token_accuracy": 0.1753251388669014,
      "num_tokens": 61886766.0,
      "step": 28250
    },
    {
      "entropy": 5.740855693817139,
      "epoch": 2.804188169908694,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042207863430201466,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1852702558040619,
      "num_tokens": 61896538.0,
      "step": 28255
    },
    {
      "entropy": 5.619879150390625,
      "epoch": 2.8046843985708616,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042205181903345586,
      "loss": 5.0267,
      "mean_token_accuracy": 0.17767500579357148,
      "num_tokens": 61907328.0,
      "step": 28260
    },
    {
      "entropy": 5.637238168716431,
      "epoch": 2.805180627233029,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004220250001183288,
      "loss": 5.0558,
      "mean_token_accuracy": 0.1801632434129715,
      "num_tokens": 61918797.0,
      "step": 28265
    },
    {
      "entropy": 5.678374195098877,
      "epoch": 2.8056768558951966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004219981775572986,
      "loss": 5.079,
      "mean_token_accuracy": 0.18435271829366684,
      "num_tokens": 61928755.0,
      "step": 28270
    },
    {
      "entropy": 5.661672830581665,
      "epoch": 2.806173084557364,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004219713513510303,
      "loss": 5.05,
      "mean_token_accuracy": 0.18221821635961533,
      "num_tokens": 61939993.0,
      "step": 28275
    },
    {
      "entropy": 5.59008092880249,
      "epoch": 2.8066693132195315,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042194452150018917,
      "loss": 4.9107,
      "mean_token_accuracy": 0.1963660791516304,
      "num_tokens": 61950982.0,
      "step": 28280
    },
    {
      "entropy": 5.651943063735962,
      "epoch": 2.807165541881699,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042191768800544055,
      "loss": 5.0147,
      "mean_token_accuracy": 0.18054611682891847,
      "num_tokens": 61961786.0,
      "step": 28285
    },
    {
      "entropy": 5.648893547058106,
      "epoch": 2.8076617705438665,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042189085086744986,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19300242960453035,
      "num_tokens": 61972548.0,
      "step": 28290
    },
    {
      "entropy": 5.627050447463989,
      "epoch": 2.808157999206034,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004218640100868826,
      "loss": 4.9966,
      "mean_token_accuracy": 0.1867861971259117,
      "num_tokens": 61981822.0,
      "step": 28295
    },
    {
      "entropy": 5.6797767162323,
      "epoch": 2.8086542278682014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042183716566440433,
      "loss": 5.1234,
      "mean_token_accuracy": 0.17217372506856918,
      "num_tokens": 61992997.0,
      "step": 28300
    },
    {
      "entropy": 5.680510520935059,
      "epoch": 2.8091504565303693,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004218103176006808,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17969800382852555,
      "num_tokens": 62004391.0,
      "step": 28305
    },
    {
      "entropy": 5.662801027297974,
      "epoch": 2.809646685192537,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004217834658963777,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18307759165763854,
      "num_tokens": 62014099.0,
      "step": 28310
    },
    {
      "entropy": 5.6721107959747314,
      "epoch": 2.8101429138547043,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042175661055216093,
      "loss": 4.9973,
      "mean_token_accuracy": 0.18318615555763246,
      "num_tokens": 62023872.0,
      "step": 28315
    },
    {
      "entropy": 5.680202198028565,
      "epoch": 2.8106391425168717,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004217297515686965,
      "loss": 5.0077,
      "mean_token_accuracy": 0.1816487580537796,
      "num_tokens": 62035566.0,
      "step": 28320
    },
    {
      "entropy": 5.637287759780884,
      "epoch": 2.811135371179039,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004217028889466503,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1788201466202736,
      "num_tokens": 62046742.0,
      "step": 28325
    },
    {
      "entropy": 5.643389368057251,
      "epoch": 2.8116315998412067,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004216760226866887,
      "loss": 4.9832,
      "mean_token_accuracy": 0.18372279405593872,
      "num_tokens": 62056545.0,
      "step": 28330
    },
    {
      "entropy": 5.682279109954834,
      "epoch": 2.8121278285033746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042164915278947775,
      "loss": 5.0897,
      "mean_token_accuracy": 0.18192748725414276,
      "num_tokens": 62066414.0,
      "step": 28335
    },
    {
      "entropy": 5.603960227966309,
      "epoch": 2.812624057165542,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004216222792556838,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18415651470422745,
      "num_tokens": 62078406.0,
      "step": 28340
    },
    {
      "entropy": 5.639530420303345,
      "epoch": 2.8131202858277096,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004215954020859732,
      "loss": 5.0476,
      "mean_token_accuracy": 0.17765790224075317,
      "num_tokens": 62089516.0,
      "step": 28345
    },
    {
      "entropy": 5.653100776672363,
      "epoch": 2.813616514489877,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042156852128101264,
      "loss": 5.0207,
      "mean_token_accuracy": 0.18657675683498381,
      "num_tokens": 62099287.0,
      "step": 28350
    },
    {
      "entropy": 5.631223154067993,
      "epoch": 2.8141127431520445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042154163684146855,
      "loss": 4.976,
      "mean_token_accuracy": 0.18441150039434434,
      "num_tokens": 62111110.0,
      "step": 28355
    },
    {
      "entropy": 5.760529613494873,
      "epoch": 2.814608971814212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004215147487680076,
      "loss": 5.1491,
      "mean_token_accuracy": 0.1744990900158882,
      "num_tokens": 62123614.0,
      "step": 28360
    },
    {
      "entropy": 5.650915479660034,
      "epoch": 2.8151052004763795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042148785706129657,
      "loss": 5.0917,
      "mean_token_accuracy": 0.1820780873298645,
      "num_tokens": 62133682.0,
      "step": 28365
    },
    {
      "entropy": 5.626917982101441,
      "epoch": 2.815601429138547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004214609617220024,
      "loss": 5.011,
      "mean_token_accuracy": 0.18609764575958251,
      "num_tokens": 62144802.0,
      "step": 28370
    },
    {
      "entropy": 5.668693113327026,
      "epoch": 2.8160976578007144,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042143406275079186,
      "loss": 4.9918,
      "mean_token_accuracy": 0.1788462609052658,
      "num_tokens": 62154205.0,
      "step": 28375
    },
    {
      "entropy": 5.6216278076171875,
      "epoch": 2.816593886462882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004214071601483322,
      "loss": 4.9222,
      "mean_token_accuracy": 0.19649873673915863,
      "num_tokens": 62164648.0,
      "step": 28380
    },
    {
      "entropy": 5.6184814929962155,
      "epoch": 2.8170901151250494,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004213802539152905,
      "loss": 5.0084,
      "mean_token_accuracy": 0.18635935932397843,
      "num_tokens": 62176073.0,
      "step": 28385
    },
    {
      "entropy": 5.671701288223266,
      "epoch": 2.8175863437872173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042135334405233384,
      "loss": 5.046,
      "mean_token_accuracy": 0.18120885044336318,
      "num_tokens": 62187350.0,
      "step": 28390
    },
    {
      "entropy": 5.613169002532959,
      "epoch": 2.8180825724493848,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004213264305601296,
      "loss": 4.9687,
      "mean_token_accuracy": 0.1799824595451355,
      "num_tokens": 62197729.0,
      "step": 28395
    },
    {
      "entropy": 5.611161756515503,
      "epoch": 2.8185788011115522,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042129951343934527,
      "loss": 4.9618,
      "mean_token_accuracy": 0.19256726652383804,
      "num_tokens": 62209047.0,
      "step": 28400
    },
    {
      "entropy": 5.6480268955230715,
      "epoch": 2.8190750297737197,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004212725926906482,
      "loss": 5.0113,
      "mean_token_accuracy": 0.18157424181699752,
      "num_tokens": 62218859.0,
      "step": 28405
    },
    {
      "entropy": 5.627857971191406,
      "epoch": 2.819571258435887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042124566831470604,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18510057032108307,
      "num_tokens": 62230396.0,
      "step": 28410
    },
    {
      "entropy": 5.6242560863494875,
      "epoch": 2.8200674870980547,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004212187403121864,
      "loss": 4.9784,
      "mean_token_accuracy": 0.1889033645391464,
      "num_tokens": 62241790.0,
      "step": 28415
    },
    {
      "entropy": 5.616982936859131,
      "epoch": 2.8205637157602226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042119180868375707,
      "loss": 4.9549,
      "mean_token_accuracy": 0.1871782049536705,
      "num_tokens": 62253820.0,
      "step": 28420
    },
    {
      "entropy": 5.613757085800171,
      "epoch": 2.82105994442239,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004211648734300859,
      "loss": 5.0217,
      "mean_token_accuracy": 0.18587653934955597,
      "num_tokens": 62263875.0,
      "step": 28425
    },
    {
      "entropy": 5.646945953369141,
      "epoch": 2.8215561730845575,
      "grad_norm": 1.125,
      "learning_rate": 0.00042113793455184084,
      "loss": 5.043,
      "mean_token_accuracy": 0.1790230840444565,
      "num_tokens": 62275390.0,
      "step": 28430
    },
    {
      "entropy": 5.653855085372925,
      "epoch": 2.822052401746725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004211109920496899,
      "loss": 5.0744,
      "mean_token_accuracy": 0.17600519359111785,
      "num_tokens": 62286969.0,
      "step": 28435
    },
    {
      "entropy": 5.655211496353149,
      "epoch": 2.8225486304088925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00042108404592430126,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18186463713645934,
      "num_tokens": 62297810.0,
      "step": 28440
    },
    {
      "entropy": 5.638306570053101,
      "epoch": 2.82304485907106,
      "grad_norm": 1.046875,
      "learning_rate": 0.000421057096176343,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18408167511224746,
      "num_tokens": 62309474.0,
      "step": 28445
    },
    {
      "entropy": 5.613359212875366,
      "epoch": 2.8235410877332274,
      "grad_norm": 1.0,
      "learning_rate": 0.0004210301428064835,
      "loss": 5.0472,
      "mean_token_accuracy": 0.18427337110042571,
      "num_tokens": 62320022.0,
      "step": 28450
    },
    {
      "entropy": 5.696625137329102,
      "epoch": 2.824037316395395,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004210031858153911,
      "loss": 5.0357,
      "mean_token_accuracy": 0.19177151024341582,
      "num_tokens": 62330450.0,
      "step": 28455
    },
    {
      "entropy": 5.6645361423492435,
      "epoch": 2.8245335450575624,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042097622520373434,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18730973750352858,
      "num_tokens": 62341419.0,
      "step": 28460
    },
    {
      "entropy": 5.673176097869873,
      "epoch": 2.82502977371973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004209492609721817,
      "loss": 5.0859,
      "mean_token_accuracy": 0.18045921325683595,
      "num_tokens": 62353312.0,
      "step": 28465
    },
    {
      "entropy": 5.675391340255738,
      "epoch": 2.8255260023818973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004209222931214019,
      "loss": 5.035,
      "mean_token_accuracy": 0.1836095556616783,
      "num_tokens": 62364838.0,
      "step": 28470
    },
    {
      "entropy": 5.655869865417481,
      "epoch": 2.8260222310440652,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042089532165206367,
      "loss": 4.9499,
      "mean_token_accuracy": 0.19687241762876512,
      "num_tokens": 62376443.0,
      "step": 28475
    },
    {
      "entropy": 5.704240942001343,
      "epoch": 2.8265184597062327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004208683465648358,
      "loss": 5.1408,
      "mean_token_accuracy": 0.17663374692201614,
      "num_tokens": 62387263.0,
      "step": 28480
    },
    {
      "entropy": 5.661428594589234,
      "epoch": 2.8270146883684,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042084136786038737,
      "loss": 4.9654,
      "mean_token_accuracy": 0.18883695155382157,
      "num_tokens": 62398929.0,
      "step": 28485
    },
    {
      "entropy": 5.669475984573364,
      "epoch": 2.8275109170305677,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004208143855393872,
      "loss": 5.0807,
      "mean_token_accuracy": 0.17477056086063386,
      "num_tokens": 62409516.0,
      "step": 28490
    },
    {
      "entropy": 5.70139594078064,
      "epoch": 2.828007145692735,
      "grad_norm": 1.0,
      "learning_rate": 0.00042078739960250446,
      "loss": 5.0363,
      "mean_token_accuracy": 0.18845886290073394,
      "num_tokens": 62421856.0,
      "step": 28495
    },
    {
      "entropy": 5.678028583526611,
      "epoch": 2.8285033743549026,
      "grad_norm": 1.125,
      "learning_rate": 0.00042076041005040835,
      "loss": 4.9728,
      "mean_token_accuracy": 0.18302861750125884,
      "num_tokens": 62432340.0,
      "step": 28500
    },
    {
      "entropy": 5.633771896362305,
      "epoch": 2.8289996030170705,
      "grad_norm": 1.125,
      "learning_rate": 0.0004207334168837682,
      "loss": 5.0564,
      "mean_token_accuracy": 0.17725768238306044,
      "num_tokens": 62442652.0,
      "step": 28505
    },
    {
      "entropy": 5.671031713485718,
      "epoch": 2.829495831679238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004207064201032533,
      "loss": 4.9862,
      "mean_token_accuracy": 0.1914087191224098,
      "num_tokens": 62454503.0,
      "step": 28510
    },
    {
      "entropy": 5.596483278274536,
      "epoch": 2.8299920603414055,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042067941970953316,
      "loss": 4.9641,
      "mean_token_accuracy": 0.18966117650270461,
      "num_tokens": 62464858.0,
      "step": 28515
    },
    {
      "entropy": 5.672163772583008,
      "epoch": 2.830488289003573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042065241570327734,
      "loss": 5.0428,
      "mean_token_accuracy": 0.1785295605659485,
      "num_tokens": 62475153.0,
      "step": 28520
    },
    {
      "entropy": 5.667665910720825,
      "epoch": 2.8309845176657404,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004206254080851555,
      "loss": 5.0866,
      "mean_token_accuracy": 0.17936773598194122,
      "num_tokens": 62487883.0,
      "step": 28525
    },
    {
      "entropy": 5.643094110488891,
      "epoch": 2.831480746327908,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042059839685583726,
      "loss": 4.9905,
      "mean_token_accuracy": 0.18520406484603882,
      "num_tokens": 62499905.0,
      "step": 28530
    },
    {
      "entropy": 5.675932550430298,
      "epoch": 2.8319769749900754,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042057138201599247,
      "loss": 5.0625,
      "mean_token_accuracy": 0.17918011993169786,
      "num_tokens": 62511403.0,
      "step": 28535
    },
    {
      "entropy": 5.644376373291015,
      "epoch": 2.832473203652243,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042054436356629113,
      "loss": 5.0093,
      "mean_token_accuracy": 0.19169364124536514,
      "num_tokens": 62522543.0,
      "step": 28540
    },
    {
      "entropy": 5.617155647277832,
      "epoch": 2.8329694323144103,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004205173415074032,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18510881960391998,
      "num_tokens": 62533053.0,
      "step": 28545
    },
    {
      "entropy": 5.670138359069824,
      "epoch": 2.833465660976578,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004204903158399989,
      "loss": 5.1038,
      "mean_token_accuracy": 0.17678156942129136,
      "num_tokens": 62543699.0,
      "step": 28550
    },
    {
      "entropy": 5.627392244338989,
      "epoch": 2.8339618896387453,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004204632865647481,
      "loss": 5.0445,
      "mean_token_accuracy": 0.18649811148643494,
      "num_tokens": 62554121.0,
      "step": 28555
    },
    {
      "entropy": 5.667775630950928,
      "epoch": 2.834458118300913,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004204362536823213,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18429603427648544,
      "num_tokens": 62564010.0,
      "step": 28560
    },
    {
      "entropy": 5.624179363250732,
      "epoch": 2.8349543469630807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042040921719338877,
      "loss": 4.9514,
      "mean_token_accuracy": 0.18798715621232986,
      "num_tokens": 62574451.0,
      "step": 28565
    },
    {
      "entropy": 5.644273281097412,
      "epoch": 2.835450575625248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000420382177098621,
      "loss": 5.0344,
      "mean_token_accuracy": 0.18180313110351562,
      "num_tokens": 62585298.0,
      "step": 28570
    },
    {
      "entropy": 5.667098188400269,
      "epoch": 2.8359468042874156,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042035513339868855,
      "loss": 5.0704,
      "mean_token_accuracy": 0.1854921504855156,
      "num_tokens": 62596248.0,
      "step": 28575
    },
    {
      "entropy": 5.686450242996216,
      "epoch": 2.836443032949583,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004203280860942619,
      "loss": 5.0472,
      "mean_token_accuracy": 0.17974888831377028,
      "num_tokens": 62607233.0,
      "step": 28580
    },
    {
      "entropy": 5.656635379791259,
      "epoch": 2.8369392616117506,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004203010351860119,
      "loss": 4.9676,
      "mean_token_accuracy": 0.191437728703022,
      "num_tokens": 62618564.0,
      "step": 28585
    },
    {
      "entropy": 5.621137809753418,
      "epoch": 2.837435490273918,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004202739806746094,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18165084719657898,
      "num_tokens": 62630049.0,
      "step": 28590
    },
    {
      "entropy": 5.674604272842407,
      "epoch": 2.837931718936086,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004202469225607252,
      "loss": 5.0998,
      "mean_token_accuracy": 0.17453185915946962,
      "num_tokens": 62640887.0,
      "step": 28595
    },
    {
      "entropy": 5.671249961853027,
      "epoch": 2.8384279475982535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004202198608450302,
      "loss": 5.056,
      "mean_token_accuracy": 0.17916857004165648,
      "num_tokens": 62651816.0,
      "step": 28600
    },
    {
      "entropy": 5.635812282562256,
      "epoch": 2.838924176260421,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004201927955281957,
      "loss": 5.0066,
      "mean_token_accuracy": 0.18236065655946732,
      "num_tokens": 62662601.0,
      "step": 28605
    },
    {
      "entropy": 5.592185974121094,
      "epoch": 2.8394204049225884,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004201657266108926,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18789913207292558,
      "num_tokens": 62673661.0,
      "step": 28610
    },
    {
      "entropy": 5.599120044708252,
      "epoch": 2.839916633584756,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042013865409379244,
      "loss": 4.984,
      "mean_token_accuracy": 0.1906283974647522,
      "num_tokens": 62683984.0,
      "step": 28615
    },
    {
      "entropy": 5.681260967254639,
      "epoch": 2.8404128622469234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201115779775663,
      "loss": 5.0387,
      "mean_token_accuracy": 0.18047577291727065,
      "num_tokens": 62695338.0,
      "step": 28620
    },
    {
      "entropy": 5.6913776874542235,
      "epoch": 2.840909090909091,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004200844982628857,
      "loss": 5.0851,
      "mean_token_accuracy": 0.18112850785255433,
      "num_tokens": 62706341.0,
      "step": 28625
    },
    {
      "entropy": 5.664693212509155,
      "epoch": 2.8414053195712583,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004200574149504222,
      "loss": 5.0708,
      "mean_token_accuracy": 0.18741879910230635,
      "num_tokens": 62717895.0,
      "step": 28630
    },
    {
      "entropy": 5.6611621379852295,
      "epoch": 2.841901548233426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004200303280408474,
      "loss": 5.0471,
      "mean_token_accuracy": 0.17916405946016312,
      "num_tokens": 62729745.0,
      "step": 28635
    },
    {
      "entropy": 5.651779842376709,
      "epoch": 2.8423977768955933,
      "grad_norm": 1.171875,
      "learning_rate": 0.00042000323753483296,
      "loss": 5.0252,
      "mean_token_accuracy": 0.18828121125698088,
      "num_tokens": 62740298.0,
      "step": 28640
    },
    {
      "entropy": 5.613928508758545,
      "epoch": 2.842894005557761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041997614343305067,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1911056011915207,
      "num_tokens": 62751292.0,
      "step": 28645
    },
    {
      "entropy": 5.66034517288208,
      "epoch": 2.8433902342199286,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041994904573617247,
      "loss": 5.1189,
      "mean_token_accuracy": 0.1769956797361374,
      "num_tokens": 62761789.0,
      "step": 28650
    },
    {
      "entropy": 5.620557880401611,
      "epoch": 2.843886462882096,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041992194444487026,
      "loss": 5.0103,
      "mean_token_accuracy": 0.18628036081790925,
      "num_tokens": 62773317.0,
      "step": 28655
    },
    {
      "entropy": 5.622793865203858,
      "epoch": 2.8443826915442636,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041989483955981607,
      "loss": 4.9829,
      "mean_token_accuracy": 0.1871124565601349,
      "num_tokens": 62783780.0,
      "step": 28660
    },
    {
      "entropy": 5.6573038578033445,
      "epoch": 2.844878920206431,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041986773108168216,
      "loss": 5.011,
      "mean_token_accuracy": 0.18539528250694276,
      "num_tokens": 62794535.0,
      "step": 28665
    },
    {
      "entropy": 5.623552846908569,
      "epoch": 2.8453751488685985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041984061901114057,
      "loss": 5.0125,
      "mean_token_accuracy": 0.178994083404541,
      "num_tokens": 62804410.0,
      "step": 28670
    },
    {
      "entropy": 5.679698753356933,
      "epoch": 2.845871377530766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004198135033488638,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19271342903375627,
      "num_tokens": 62814693.0,
      "step": 28675
    },
    {
      "entropy": 5.670564794540406,
      "epoch": 2.846367606192934,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041978638409552417,
      "loss": 4.9963,
      "mean_token_accuracy": 0.18166814148426055,
      "num_tokens": 62826114.0,
      "step": 28680
    },
    {
      "entropy": 5.618617868423462,
      "epoch": 2.8468638348551014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004197592612517942,
      "loss": 5.0398,
      "mean_token_accuracy": 0.18268163800239562,
      "num_tokens": 62838130.0,
      "step": 28685
    },
    {
      "entropy": 5.7492475509643555,
      "epoch": 2.847360063517269,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004197321348183465,
      "loss": 5.0981,
      "mean_token_accuracy": 0.1826516345143318,
      "num_tokens": 62849290.0,
      "step": 28690
    },
    {
      "entropy": 5.718512725830078,
      "epoch": 2.8478562921794364,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041970500479585364,
      "loss": 5.1506,
      "mean_token_accuracy": 0.17850203812122345,
      "num_tokens": 62860888.0,
      "step": 28695
    },
    {
      "entropy": 5.685663223266602,
      "epoch": 2.848352520841604,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041967787118498854,
      "loss": 5.062,
      "mean_token_accuracy": 0.1905467301607132,
      "num_tokens": 62870231.0,
      "step": 28700
    },
    {
      "entropy": 5.6666899681091305,
      "epoch": 2.8488487495037713,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004196507339864241,
      "loss": 4.9808,
      "mean_token_accuracy": 0.18726201206445695,
      "num_tokens": 62881147.0,
      "step": 28705
    },
    {
      "entropy": 5.6736362934112545,
      "epoch": 2.849344978165939,
      "grad_norm": 1.15625,
      "learning_rate": 0.000419623593200833,
      "loss": 5.0024,
      "mean_token_accuracy": 0.18285225480794906,
      "num_tokens": 62891272.0,
      "step": 28710
    },
    {
      "entropy": 5.581662225723266,
      "epoch": 2.8498412068281063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041959644882888844,
      "loss": 4.9883,
      "mean_token_accuracy": 0.18214948177337648,
      "num_tokens": 62902376.0,
      "step": 28715
    },
    {
      "entropy": 5.591905641555786,
      "epoch": 2.8503374354902737,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041956930087126357,
      "loss": 4.9893,
      "mean_token_accuracy": 0.1907816633582115,
      "num_tokens": 62913811.0,
      "step": 28720
    },
    {
      "entropy": 5.642810440063476,
      "epoch": 2.850833664152441,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041954214932863156,
      "loss": 5.0545,
      "mean_token_accuracy": 0.17752996683120728,
      "num_tokens": 62926001.0,
      "step": 28725
    },
    {
      "entropy": 5.646038436889649,
      "epoch": 2.851329892814609,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004195149942016657,
      "loss": 5.0063,
      "mean_token_accuracy": 0.18225767463445663,
      "num_tokens": 62936183.0,
      "step": 28730
    },
    {
      "entropy": 5.634921073913574,
      "epoch": 2.8518261214767766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004194878354910394,
      "loss": 4.9769,
      "mean_token_accuracy": 0.1912127986550331,
      "num_tokens": 62947427.0,
      "step": 28735
    },
    {
      "entropy": 5.656744527816772,
      "epoch": 2.852322350138944,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041946067319742605,
      "loss": 5.0653,
      "mean_token_accuracy": 0.18482588678598405,
      "num_tokens": 62958630.0,
      "step": 28740
    },
    {
      "entropy": 5.635790205001831,
      "epoch": 2.8528185788011116,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041943350732149935,
      "loss": 4.9908,
      "mean_token_accuracy": 0.185550357401371,
      "num_tokens": 62968856.0,
      "step": 28745
    },
    {
      "entropy": 5.622393798828125,
      "epoch": 2.853314807463279,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041940633786393295,
      "loss": 4.9198,
      "mean_token_accuracy": 0.19125569015741348,
      "num_tokens": 62978393.0,
      "step": 28750
    },
    {
      "entropy": 5.611555337905884,
      "epoch": 2.8538110361254465,
      "grad_norm": 1.328125,
      "learning_rate": 0.00041937916482540045,
      "loss": 5.0206,
      "mean_token_accuracy": 0.18724679052829743,
      "num_tokens": 62989786.0,
      "step": 28755
    },
    {
      "entropy": 5.6548606872558596,
      "epoch": 2.854307264787614,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00041935198820657575,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18986773043870925,
      "num_tokens": 63000021.0,
      "step": 28760
    },
    {
      "entropy": 5.6440122604370115,
      "epoch": 2.854803493449782,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004193248080081329,
      "loss": 5.0285,
      "mean_token_accuracy": 0.1773261234164238,
      "num_tokens": 63011398.0,
      "step": 28765
    },
    {
      "entropy": 5.683233547210693,
      "epoch": 2.8552997221119494,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004192976242307457,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19006709158420562,
      "num_tokens": 63021155.0,
      "step": 28770
    },
    {
      "entropy": 5.6759308815002445,
      "epoch": 2.855795950774117,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004192704368750884,
      "loss": 5.0558,
      "mean_token_accuracy": 0.1797106981277466,
      "num_tokens": 63031510.0,
      "step": 28775
    },
    {
      "entropy": 5.696593189239502,
      "epoch": 2.8562921794362843,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004192432459418351,
      "loss": 5.051,
      "mean_token_accuracy": 0.18285465091466904,
      "num_tokens": 63041978.0,
      "step": 28780
    },
    {
      "entropy": 5.68225793838501,
      "epoch": 2.856788408098452,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004192160514316602,
      "loss": 5.0949,
      "mean_token_accuracy": 0.1758950471878052,
      "num_tokens": 63052194.0,
      "step": 28785
    },
    {
      "entropy": 5.586061477661133,
      "epoch": 2.8572846367606193,
      "grad_norm": 1.125,
      "learning_rate": 0.0004191888533452379,
      "loss": 4.949,
      "mean_token_accuracy": 0.19139712303876877,
      "num_tokens": 63062144.0,
      "step": 28790
    },
    {
      "entropy": 5.614221477508545,
      "epoch": 2.8577808654227868,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004191616516832428,
      "loss": 5.052,
      "mean_token_accuracy": 0.1782183438539505,
      "num_tokens": 63072403.0,
      "step": 28795
    },
    {
      "entropy": 5.628628301620483,
      "epoch": 2.8582770940849542,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004191344464463494,
      "loss": 4.9831,
      "mean_token_accuracy": 0.18737420290708542,
      "num_tokens": 63084799.0,
      "step": 28800
    },
    {
      "entropy": 5.6527900218963625,
      "epoch": 2.8587733227471217,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041910723763523224,
      "loss": 5.0634,
      "mean_token_accuracy": 0.18839122354984283,
      "num_tokens": 63095437.0,
      "step": 28805
    },
    {
      "entropy": 5.634169054031372,
      "epoch": 2.859269551409289,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004190800252505661,
      "loss": 5.0854,
      "mean_token_accuracy": 0.17917048633098603,
      "num_tokens": 63106743.0,
      "step": 28810
    },
    {
      "entropy": 5.6029235363006595,
      "epoch": 2.8597657800714567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004190528092930259,
      "loss": 5.0094,
      "mean_token_accuracy": 0.18349028378725052,
      "num_tokens": 63120525.0,
      "step": 28815
    },
    {
      "entropy": 5.680508852005005,
      "epoch": 2.8602620087336246,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004190255897632863,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18339774757623672,
      "num_tokens": 63131510.0,
      "step": 28820
    },
    {
      "entropy": 5.7220982074737545,
      "epoch": 2.860758237395792,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041899836666202254,
      "loss": 5.0689,
      "mean_token_accuracy": 0.1889885902404785,
      "num_tokens": 63143094.0,
      "step": 28825
    },
    {
      "entropy": 5.602231407165528,
      "epoch": 2.8612544660579595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004189711399899095,
      "loss": 4.9787,
      "mean_token_accuracy": 0.1911844328045845,
      "num_tokens": 63155159.0,
      "step": 28830
    },
    {
      "entropy": 5.596693801879883,
      "epoch": 2.861750694720127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004189439097476224,
      "loss": 5.0215,
      "mean_token_accuracy": 0.19252488762140274,
      "num_tokens": 63167132.0,
      "step": 28835
    },
    {
      "entropy": 5.647830390930176,
      "epoch": 2.8622469233822945,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004189166759358366,
      "loss": 5.0228,
      "mean_token_accuracy": 0.18790313005447387,
      "num_tokens": 63176864.0,
      "step": 28840
    },
    {
      "entropy": 5.671120071411133,
      "epoch": 2.862743152044462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041888943855522727,
      "loss": 5.0312,
      "mean_token_accuracy": 0.1810614511370659,
      "num_tokens": 63188003.0,
      "step": 28845
    },
    {
      "entropy": 5.613817024230957,
      "epoch": 2.86323938070663,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041886219760646997,
      "loss": 4.9671,
      "mean_token_accuracy": 0.18503632098436357,
      "num_tokens": 63197634.0,
      "step": 28850
    },
    {
      "entropy": 5.590407943725586,
      "epoch": 2.8637356093687973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004188349530902402,
      "loss": 4.9393,
      "mean_token_accuracy": 0.18779747039079667,
      "num_tokens": 63207441.0,
      "step": 28855
    },
    {
      "entropy": 5.597579002380371,
      "epoch": 2.864231838030965,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041880770500721346,
      "loss": 4.9864,
      "mean_token_accuracy": 0.1870836943387985,
      "num_tokens": 63217541.0,
      "step": 28860
    },
    {
      "entropy": 5.60846529006958,
      "epoch": 2.8647280666931323,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004187804533580655,
      "loss": 4.9623,
      "mean_token_accuracy": 0.1901414006948471,
      "num_tokens": 63228391.0,
      "step": 28865
    },
    {
      "entropy": 5.642832279205322,
      "epoch": 2.8652242953552998,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004187531981434721,
      "loss": 4.9882,
      "mean_token_accuracy": 0.1914323464035988,
      "num_tokens": 63238462.0,
      "step": 28870
    },
    {
      "entropy": 5.612261581420898,
      "epoch": 2.8657205240174672,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041872593936410923,
      "loss": 5.0532,
      "mean_token_accuracy": 0.18401677906513214,
      "num_tokens": 63249305.0,
      "step": 28875
    },
    {
      "entropy": 5.6917589664459225,
      "epoch": 2.8662167526796347,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041869867702065266,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18486017882823944,
      "num_tokens": 63260832.0,
      "step": 28880
    },
    {
      "entropy": 5.622360801696777,
      "epoch": 2.866712981341802,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004186714111137787,
      "loss": 5.0204,
      "mean_token_accuracy": 0.18287938684225083,
      "num_tokens": 63273689.0,
      "step": 28885
    },
    {
      "entropy": 5.704988193511963,
      "epoch": 2.8672092100039697,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004186441416441632,
      "loss": 5.1717,
      "mean_token_accuracy": 0.18085987120866776,
      "num_tokens": 63284912.0,
      "step": 28890
    },
    {
      "entropy": 5.702650499343872,
      "epoch": 2.867705438666137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004186168686124825,
      "loss": 5.0666,
      "mean_token_accuracy": 0.17877527177333832,
      "num_tokens": 63296428.0,
      "step": 28895
    },
    {
      "entropy": 5.645402717590332,
      "epoch": 2.8682016673283046,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041858959201941307,
      "loss": 4.9676,
      "mean_token_accuracy": 0.1875320389866829,
      "num_tokens": 63307509.0,
      "step": 28900
    },
    {
      "entropy": 5.683566761016846,
      "epoch": 2.8686978959904725,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041856231186563107,
      "loss": 5.1073,
      "mean_token_accuracy": 0.17424225807189941,
      "num_tokens": 63319194.0,
      "step": 28905
    },
    {
      "entropy": 5.598350763320923,
      "epoch": 2.86919412465264,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041853502815181306,
      "loss": 4.954,
      "mean_token_accuracy": 0.18748095780611038,
      "num_tokens": 63329122.0,
      "step": 28910
    },
    {
      "entropy": 5.638651943206787,
      "epoch": 2.8696903533148075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004185077408786357,
      "loss": 5.0259,
      "mean_token_accuracy": 0.18310668170452118,
      "num_tokens": 63340304.0,
      "step": 28915
    },
    {
      "entropy": 5.682317066192627,
      "epoch": 2.870186581976975,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004184804500467756,
      "loss": 5.033,
      "mean_token_accuracy": 0.1803188666701317,
      "num_tokens": 63351029.0,
      "step": 28920
    },
    {
      "entropy": 5.601900196075439,
      "epoch": 2.8706828106391424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041845315565690945,
      "loss": 4.9638,
      "mean_token_accuracy": 0.18738200515508652,
      "num_tokens": 63361339.0,
      "step": 28925
    },
    {
      "entropy": 5.676943254470825,
      "epoch": 2.87117903930131,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041842585770971417,
      "loss": 5.0232,
      "mean_token_accuracy": 0.1904005452990532,
      "num_tokens": 63373431.0,
      "step": 28930
    },
    {
      "entropy": 5.624940490722656,
      "epoch": 2.871675267963478,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004183985562058667,
      "loss": 4.9907,
      "mean_token_accuracy": 0.18484213799238206,
      "num_tokens": 63384096.0,
      "step": 28935
    },
    {
      "entropy": 5.722109699249268,
      "epoch": 2.8721714966256453,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000418371251146044,
      "loss": 5.0606,
      "mean_token_accuracy": 0.1799267575144768,
      "num_tokens": 63394852.0,
      "step": 28940
    },
    {
      "entropy": 5.672243261337281,
      "epoch": 2.872667725287813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004183439425309232,
      "loss": 5.0265,
      "mean_token_accuracy": 0.18857532739639282,
      "num_tokens": 63405575.0,
      "step": 28945
    },
    {
      "entropy": 5.610588693618775,
      "epoch": 2.8731639539499803,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004183166303611816,
      "loss": 5.0546,
      "mean_token_accuracy": 0.18649081885814667,
      "num_tokens": 63416677.0,
      "step": 28950
    },
    {
      "entropy": 5.668358564376831,
      "epoch": 2.8736601826121477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004182893146374963,
      "loss": 5.1079,
      "mean_token_accuracy": 0.18196916133165358,
      "num_tokens": 63428613.0,
      "step": 28955
    },
    {
      "entropy": 5.680631017684936,
      "epoch": 2.874156411274315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041826199536054476,
      "loss": 5.0539,
      "mean_token_accuracy": 0.18084003478288652,
      "num_tokens": 63439321.0,
      "step": 28960
    },
    {
      "entropy": 5.6682148456573485,
      "epoch": 2.8746526399364827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004182346725310045,
      "loss": 5.0817,
      "mean_token_accuracy": 0.1817978024482727,
      "num_tokens": 63451279.0,
      "step": 28965
    },
    {
      "entropy": 5.681896781921386,
      "epoch": 2.87514886859865,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041820734614955295,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18840765506029128,
      "num_tokens": 63460801.0,
      "step": 28970
    },
    {
      "entropy": 5.685109615325928,
      "epoch": 2.8756450972608176,
      "grad_norm": 1.125,
      "learning_rate": 0.0004181800162168678,
      "loss": 5.0117,
      "mean_token_accuracy": 0.1898718625307083,
      "num_tokens": 63470232.0,
      "step": 28975
    },
    {
      "entropy": 5.604404926300049,
      "epoch": 2.876141325922985,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004181526827336268,
      "loss": 5.0257,
      "mean_token_accuracy": 0.18537116497755052,
      "num_tokens": 63480247.0,
      "step": 28980
    },
    {
      "entropy": 5.65711669921875,
      "epoch": 2.8766375545851526,
      "grad_norm": 1.0,
      "learning_rate": 0.0004181253457005077,
      "loss": 5.0201,
      "mean_token_accuracy": 0.18050152510404588,
      "num_tokens": 63492104.0,
      "step": 28985
    },
    {
      "entropy": 5.677733564376831,
      "epoch": 2.8771337832473205,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041809800511818853,
      "loss": 5.0332,
      "mean_token_accuracy": 0.18577434122562408,
      "num_tokens": 63502784.0,
      "step": 28990
    },
    {
      "entropy": 5.693964958190918,
      "epoch": 2.877630011909488,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041807066098734704,
      "loss": 5.1164,
      "mean_token_accuracy": 0.18546211570501328,
      "num_tokens": 63514150.0,
      "step": 28995
    },
    {
      "entropy": 5.665675210952759,
      "epoch": 2.8781262405716554,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004180433133086615,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18634116053581237,
      "num_tokens": 63526682.0,
      "step": 29000
    },
    {
      "entropy": 5.647340631484985,
      "epoch": 2.878622469233823,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004180159620828101,
      "loss": 5.0703,
      "mean_token_accuracy": 0.1795322597026825,
      "num_tokens": 63537448.0,
      "step": 29005
    },
    {
      "entropy": 5.637281799316407,
      "epoch": 2.8791186978959904,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004179886073104709,
      "loss": 4.9932,
      "mean_token_accuracy": 0.18109891265630723,
      "num_tokens": 63547489.0,
      "step": 29010
    },
    {
      "entropy": 5.628810691833496,
      "epoch": 2.879614926558158,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004179612489923224,
      "loss": 4.9819,
      "mean_token_accuracy": 0.19188993871212007,
      "num_tokens": 63558324.0,
      "step": 29015
    },
    {
      "entropy": 5.617344903945923,
      "epoch": 2.880111155220326,
      "grad_norm": 1.09375,
      "learning_rate": 0.000417933887129043,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18333812355995177,
      "num_tokens": 63568630.0,
      "step": 29020
    },
    {
      "entropy": 5.611018800735474,
      "epoch": 2.8806073838824933,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004179065217213111,
      "loss": 4.9488,
      "mean_token_accuracy": 0.18644191473722457,
      "num_tokens": 63578068.0,
      "step": 29025
    },
    {
      "entropy": 5.687215757369995,
      "epoch": 2.8811036125446607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041787915276980556,
      "loss": 5.069,
      "mean_token_accuracy": 0.17616955935955048,
      "num_tokens": 63589103.0,
      "step": 29030
    },
    {
      "entropy": 5.646406555175782,
      "epoch": 2.881599841206828,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041785178027520486,
      "loss": 5.0109,
      "mean_token_accuracy": 0.1908353388309479,
      "num_tokens": 63600305.0,
      "step": 29035
    },
    {
      "entropy": 5.711560010910034,
      "epoch": 2.8820960698689957,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041782440423818775,
      "loss": 5.0749,
      "mean_token_accuracy": 0.18139332830905913,
      "num_tokens": 63611910.0,
      "step": 29040
    },
    {
      "entropy": 5.6655846118927,
      "epoch": 2.882592298531163,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004177970246594332,
      "loss": 4.9921,
      "mean_token_accuracy": 0.1841554284095764,
      "num_tokens": 63622674.0,
      "step": 29045
    },
    {
      "entropy": 5.6580626487731935,
      "epoch": 2.8830885271933306,
      "grad_norm": 1.0,
      "learning_rate": 0.00041776964153962017,
      "loss": 5.0748,
      "mean_token_accuracy": 0.17746581137180328,
      "num_tokens": 63633274.0,
      "step": 29050
    },
    {
      "entropy": 5.654553842544556,
      "epoch": 2.883584755855498,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041774225487942765,
      "loss": 5.0238,
      "mean_token_accuracy": 0.19017728716135024,
      "num_tokens": 63644738.0,
      "step": 29055
    },
    {
      "entropy": 5.674219989776612,
      "epoch": 2.8840809845176656,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004177148646795348,
      "loss": 5.0922,
      "mean_token_accuracy": 0.18255554735660554,
      "num_tokens": 63655468.0,
      "step": 29060
    },
    {
      "entropy": 5.611102294921875,
      "epoch": 2.884577213179833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041768747094062084,
      "loss": 4.9466,
      "mean_token_accuracy": 0.18602794408798218,
      "num_tokens": 63667675.0,
      "step": 29065
    },
    {
      "entropy": 5.64500937461853,
      "epoch": 2.8850734418420005,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041766007366336494,
      "loss": 5.0594,
      "mean_token_accuracy": 0.17948453426361083,
      "num_tokens": 63679251.0,
      "step": 29070
    },
    {
      "entropy": 5.614674997329712,
      "epoch": 2.8855696705041685,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041763267284844675,
      "loss": 4.985,
      "mean_token_accuracy": 0.1908552974462509,
      "num_tokens": 63690809.0,
      "step": 29075
    },
    {
      "entropy": 5.644348907470703,
      "epoch": 2.886065899166336,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004176052684965456,
      "loss": 4.9875,
      "mean_token_accuracy": 0.1826278641819954,
      "num_tokens": 63702434.0,
      "step": 29080
    },
    {
      "entropy": 5.714014101028442,
      "epoch": 2.8865621278285034,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004175778606083411,
      "loss": 5.0993,
      "mean_token_accuracy": 0.18656526356935502,
      "num_tokens": 63713173.0,
      "step": 29085
    },
    {
      "entropy": 5.695063734054566,
      "epoch": 2.887058356490671,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004175504491845128,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1807670533657074,
      "num_tokens": 63723956.0,
      "step": 29090
    },
    {
      "entropy": 5.640679264068604,
      "epoch": 2.8875545851528384,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041752303422574067,
      "loss": 5.0957,
      "mean_token_accuracy": 0.17492161691188812,
      "num_tokens": 63735280.0,
      "step": 29095
    },
    {
      "entropy": 5.686241674423218,
      "epoch": 2.888050813815006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004174956157327043,
      "loss": 5.0618,
      "mean_token_accuracy": 0.17653166651725768,
      "num_tokens": 63746075.0,
      "step": 29100
    },
    {
      "entropy": 5.665151643753052,
      "epoch": 2.8885470424771738,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004174681937060837,
      "loss": 5.025,
      "mean_token_accuracy": 0.18678808212280273,
      "num_tokens": 63758152.0,
      "step": 29105
    },
    {
      "entropy": 5.628698825836182,
      "epoch": 2.8890432711393412,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041744076814655893,
      "loss": 5.0516,
      "mean_token_accuracy": 0.1878634959459305,
      "num_tokens": 63768855.0,
      "step": 29110
    },
    {
      "entropy": 5.664821815490723,
      "epoch": 2.8895394998015087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041741333905481,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1780032128095627,
      "num_tokens": 63780258.0,
      "step": 29115
    },
    {
      "entropy": 5.708815336227417,
      "epoch": 2.890035728463676,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004173859064315172,
      "loss": 5.0909,
      "mean_token_accuracy": 0.18359940350055695,
      "num_tokens": 63791959.0,
      "step": 29120
    },
    {
      "entropy": 5.673575448989868,
      "epoch": 2.8905319571258437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041735847027736067,
      "loss": 5.0619,
      "mean_token_accuracy": 0.18149505108594893,
      "num_tokens": 63802859.0,
      "step": 29125
    },
    {
      "entropy": 5.626009464263916,
      "epoch": 2.891028185788011,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004173310305930209,
      "loss": 4.9988,
      "mean_token_accuracy": 0.17767027765512466,
      "num_tokens": 63814275.0,
      "step": 29130
    },
    {
      "entropy": 5.715307521820068,
      "epoch": 2.8915244144501786,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00041730358737917823,
      "loss": 5.038,
      "mean_token_accuracy": 0.1834655523300171,
      "num_tokens": 63824633.0,
      "step": 29135
    },
    {
      "entropy": 5.6862977027893065,
      "epoch": 2.892020643112346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041727614063651313,
      "loss": 5.0403,
      "mean_token_accuracy": 0.186515174806118,
      "num_tokens": 63836478.0,
      "step": 29140
    },
    {
      "entropy": 5.714416933059693,
      "epoch": 2.8925168717745136,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004172486903657065,
      "loss": 5.0387,
      "mean_token_accuracy": 0.18227681815624236,
      "num_tokens": 63847917.0,
      "step": 29145
    },
    {
      "entropy": 5.658496236801147,
      "epoch": 2.893013100436681,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004172212365674388,
      "loss": 5.0549,
      "mean_token_accuracy": 0.18800588995218276,
      "num_tokens": 63857258.0,
      "step": 29150
    },
    {
      "entropy": 5.636815071105957,
      "epoch": 2.8935093290988485,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00041719377924239083,
      "loss": 5.104,
      "mean_token_accuracy": 0.180440554022789,
      "num_tokens": 63868627.0,
      "step": 29155
    },
    {
      "entropy": 5.689479970932007,
      "epoch": 2.8940055577610164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041716631839124356,
      "loss": 5.0802,
      "mean_token_accuracy": 0.18569062799215316,
      "num_tokens": 63879954.0,
      "step": 29160
    },
    {
      "entropy": 5.78488335609436,
      "epoch": 2.894501786423184,
      "grad_norm": 1.15625,
      "learning_rate": 0.000417138854014678,
      "loss": 5.1338,
      "mean_token_accuracy": 0.18359114229679108,
      "num_tokens": 63891147.0,
      "step": 29165
    },
    {
      "entropy": 5.64517183303833,
      "epoch": 2.8949980150853514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004171113861133751,
      "loss": 4.9995,
      "mean_token_accuracy": 0.18055322170257568,
      "num_tokens": 63902380.0,
      "step": 29170
    },
    {
      "entropy": 5.647266578674317,
      "epoch": 2.895494243747519,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004170839146880161,
      "loss": 5.0679,
      "mean_token_accuracy": 0.18839465230703353,
      "num_tokens": 63912557.0,
      "step": 29175
    },
    {
      "entropy": 5.707425975799561,
      "epoch": 2.8959904724096863,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004170564397392821,
      "loss": 5.0911,
      "mean_token_accuracy": 0.178343702852726,
      "num_tokens": 63923959.0,
      "step": 29180
    },
    {
      "entropy": 5.705269479751587,
      "epoch": 2.896486701071854,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004170289612678545,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18825327455997468,
      "num_tokens": 63935600.0,
      "step": 29185
    },
    {
      "entropy": 5.7046366214752195,
      "epoch": 2.8969829297340213,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041700147927441477,
      "loss": 5.1636,
      "mean_token_accuracy": 0.17160772085189818,
      "num_tokens": 63946371.0,
      "step": 29190
    },
    {
      "entropy": 5.677399826049805,
      "epoch": 2.897479158396189,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041697399375964435,
      "loss": 5.1024,
      "mean_token_accuracy": 0.17970819324254989,
      "num_tokens": 63956019.0,
      "step": 29195
    },
    {
      "entropy": 5.682320404052734,
      "epoch": 2.8979753870583567,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041694650472422467,
      "loss": 5.0409,
      "mean_token_accuracy": 0.18075598925352096,
      "num_tokens": 63967330.0,
      "step": 29200
    },
    {
      "entropy": 5.655933046340943,
      "epoch": 2.898471615720524,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041691901216883763,
      "loss": 5.0397,
      "mean_token_accuracy": 0.19010513871908188,
      "num_tokens": 63978952.0,
      "step": 29205
    },
    {
      "entropy": 5.639018821716308,
      "epoch": 2.8989678443826916,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000416891516094165,
      "loss": 5.0589,
      "mean_token_accuracy": 0.1768221825361252,
      "num_tokens": 63989913.0,
      "step": 29210
    },
    {
      "entropy": 5.5954913139343265,
      "epoch": 2.899464073044859,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004168640165008884,
      "loss": 4.9578,
      "mean_token_accuracy": 0.1930540382862091,
      "num_tokens": 64000033.0,
      "step": 29215
    },
    {
      "entropy": 5.642793607711792,
      "epoch": 2.8999603017070266,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041683651338968994,
      "loss": 5.0516,
      "mean_token_accuracy": 0.1778510645031929,
      "num_tokens": 64010746.0,
      "step": 29220
    },
    {
      "entropy": 5.651125764846801,
      "epoch": 2.900456530369194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041680900676125147,
      "loss": 5.009,
      "mean_token_accuracy": 0.1832447648048401,
      "num_tokens": 64021179.0,
      "step": 29225
    },
    {
      "entropy": 5.708410120010376,
      "epoch": 2.9009527590313615,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041678149661625525,
      "loss": 5.0719,
      "mean_token_accuracy": 0.179323311150074,
      "num_tokens": 64032953.0,
      "step": 29230
    },
    {
      "entropy": 5.635750389099121,
      "epoch": 2.901448987693529,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004167539829553835,
      "loss": 4.9122,
      "mean_token_accuracy": 0.2010921910405159,
      "num_tokens": 64042548.0,
      "step": 29235
    },
    {
      "entropy": 5.6367676734924315,
      "epoch": 2.9019452163556965,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041672646577931835,
      "loss": 4.9654,
      "mean_token_accuracy": 0.18218363225460052,
      "num_tokens": 64053254.0,
      "step": 29240
    },
    {
      "entropy": 5.61321530342102,
      "epoch": 2.9024414450178644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041669894508874225,
      "loss": 5.0349,
      "mean_token_accuracy": 0.17452775537967682,
      "num_tokens": 64063307.0,
      "step": 29245
    },
    {
      "entropy": 5.631636667251587,
      "epoch": 2.902937673680032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041667142088433766,
      "loss": 5.0246,
      "mean_token_accuracy": 0.19053286761045457,
      "num_tokens": 64074274.0,
      "step": 29250
    },
    {
      "entropy": 5.517972755432129,
      "epoch": 2.9034339023421993,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041664389316678713,
      "loss": 4.8951,
      "mean_token_accuracy": 0.20139629095792771,
      "num_tokens": 64084243.0,
      "step": 29255
    },
    {
      "entropy": 5.637271356582642,
      "epoch": 2.903930131004367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004166163619367733,
      "loss": 4.9876,
      "mean_token_accuracy": 0.19009556770324706,
      "num_tokens": 64095899.0,
      "step": 29260
    },
    {
      "entropy": 5.614692449569702,
      "epoch": 2.9044263596665343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041658882719497876,
      "loss": 4.9202,
      "mean_token_accuracy": 0.1861192539334297,
      "num_tokens": 64107408.0,
      "step": 29265
    },
    {
      "entropy": 5.701054096221924,
      "epoch": 2.9049225883287018,
      "grad_norm": 1.125,
      "learning_rate": 0.0004165612889420864,
      "loss": 5.0799,
      "mean_token_accuracy": 0.18005398511886597,
      "num_tokens": 64118935.0,
      "step": 29270
    },
    {
      "entropy": 5.679817390441895,
      "epoch": 2.9054188169908692,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004165337471787792,
      "loss": 5.0397,
      "mean_token_accuracy": 0.18110102117061616,
      "num_tokens": 64129085.0,
      "step": 29275
    },
    {
      "entropy": 5.608068943023682,
      "epoch": 2.905915045653037,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041650620190574004,
      "loss": 4.9411,
      "mean_token_accuracy": 0.19275281727313995,
      "num_tokens": 64139605.0,
      "step": 29280
    },
    {
      "entropy": 5.688938951492309,
      "epoch": 2.9064112743152046,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041647865312365186,
      "loss": 5.0808,
      "mean_token_accuracy": 0.18053552657365798,
      "num_tokens": 64151171.0,
      "step": 29285
    },
    {
      "entropy": 5.678523969650269,
      "epoch": 2.906907502977372,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041645110083319805,
      "loss": 5.048,
      "mean_token_accuracy": 0.18008213937282563,
      "num_tokens": 64162417.0,
      "step": 29290
    },
    {
      "entropy": 5.682714033126831,
      "epoch": 2.9074037316395396,
      "grad_norm": 1.25,
      "learning_rate": 0.00041642354503506175,
      "loss": 5.0409,
      "mean_token_accuracy": 0.182305771112442,
      "num_tokens": 64172755.0,
      "step": 29295
    },
    {
      "entropy": 5.671325588226319,
      "epoch": 2.907899960301707,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041639598572992625,
      "loss": 5.016,
      "mean_token_accuracy": 0.18196941465139388,
      "num_tokens": 64184270.0,
      "step": 29300
    },
    {
      "entropy": 5.663750457763672,
      "epoch": 2.9083961889638745,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041636842291847495,
      "loss": 5.0787,
      "mean_token_accuracy": 0.1778179883956909,
      "num_tokens": 64195400.0,
      "step": 29305
    },
    {
      "entropy": 5.698828125,
      "epoch": 2.908892417626042,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041634085660139145,
      "loss": 5.0614,
      "mean_token_accuracy": 0.18245899975299834,
      "num_tokens": 64205101.0,
      "step": 29310
    },
    {
      "entropy": 5.621560764312744,
      "epoch": 2.9093886462882095,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004163132867793592,
      "loss": 4.9336,
      "mean_token_accuracy": 0.18829631805419922,
      "num_tokens": 64216047.0,
      "step": 29315
    },
    {
      "entropy": 5.6016068935394285,
      "epoch": 2.909884874950377,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000416285713453062,
      "loss": 5.0433,
      "mean_token_accuracy": 0.18667684346437455,
      "num_tokens": 64226195.0,
      "step": 29320
    },
    {
      "entropy": 5.665503883361817,
      "epoch": 2.9103811036125444,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004162581366231835,
      "loss": 5.0132,
      "mean_token_accuracy": 0.19367197901010513,
      "num_tokens": 64236405.0,
      "step": 29325
    },
    {
      "entropy": 5.634744453430176,
      "epoch": 2.9108773322747123,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041623055629040764,
      "loss": 4.967,
      "mean_token_accuracy": 0.18541204929351807,
      "num_tokens": 64246989.0,
      "step": 29330
    },
    {
      "entropy": 5.62405219078064,
      "epoch": 2.91137356093688,
      "grad_norm": 1.0,
      "learning_rate": 0.0004162029724554183,
      "loss": 5.0016,
      "mean_token_accuracy": 0.18247146308422088,
      "num_tokens": 64257928.0,
      "step": 29335
    },
    {
      "entropy": 5.687788438796997,
      "epoch": 2.9118697895990473,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041617538511889946,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18333820551633834,
      "num_tokens": 64268985.0,
      "step": 29340
    },
    {
      "entropy": 5.762788248062134,
      "epoch": 2.9123660182612148,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004161477942815353,
      "loss": 5.1663,
      "mean_token_accuracy": 0.17916905432939528,
      "num_tokens": 64279907.0,
      "step": 29345
    },
    {
      "entropy": 5.718074798583984,
      "epoch": 2.9128622469233822,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041612019994401,
      "loss": 5.0064,
      "mean_token_accuracy": 0.18756390661001204,
      "num_tokens": 64291496.0,
      "step": 29350
    },
    {
      "entropy": 5.696304082870483,
      "epoch": 2.9133584755855497,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004160926021070078,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18967108726501464,
      "num_tokens": 64302160.0,
      "step": 29355
    },
    {
      "entropy": 5.646494626998901,
      "epoch": 2.913854704247717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004160650007712132,
      "loss": 5.077,
      "mean_token_accuracy": 0.1797836869955063,
      "num_tokens": 64313671.0,
      "step": 29360
    },
    {
      "entropy": 5.642529058456421,
      "epoch": 2.914350932909885,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041603739593731046,
      "loss": 5.0655,
      "mean_token_accuracy": 0.1782272055745125,
      "num_tokens": 64325376.0,
      "step": 29365
    },
    {
      "entropy": 5.711380529403686,
      "epoch": 2.9148471615720526,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004160097876059842,
      "loss": 5.0727,
      "mean_token_accuracy": 0.18247185051441192,
      "num_tokens": 64336048.0,
      "step": 29370
    },
    {
      "entropy": 5.760016441345215,
      "epoch": 2.91534339023422,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004159821757779191,
      "loss": 5.1003,
      "mean_token_accuracy": 0.18083226084709167,
      "num_tokens": 64346833.0,
      "step": 29375
    },
    {
      "entropy": 5.637574768066406,
      "epoch": 2.9158396188963875,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041595456045379975,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18257290571928025,
      "num_tokens": 64359656.0,
      "step": 29380
    },
    {
      "entropy": 5.604636335372925,
      "epoch": 2.916335847558555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004159269416343111,
      "loss": 4.9652,
      "mean_token_accuracy": 0.18397252410650253,
      "num_tokens": 64371130.0,
      "step": 29385
    },
    {
      "entropy": 5.624562358856201,
      "epoch": 2.9168320762207225,
      "grad_norm": 1.15625,
      "learning_rate": 0.000415899319320138,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18797527402639388,
      "num_tokens": 64381826.0,
      "step": 29390
    },
    {
      "entropy": 5.733919715881347,
      "epoch": 2.91732830488289,
      "grad_norm": 1.125,
      "learning_rate": 0.00041587169351196536,
      "loss": 5.1302,
      "mean_token_accuracy": 0.17327823340892792,
      "num_tokens": 64392725.0,
      "step": 29395
    },
    {
      "entropy": 5.723057460784912,
      "epoch": 2.9178245335450574,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041584406421047825,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18325423896312715,
      "num_tokens": 64402587.0,
      "step": 29400
    },
    {
      "entropy": 5.674096298217774,
      "epoch": 2.918320762207225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041581643141636184,
      "loss": 5.0662,
      "mean_token_accuracy": 0.18714593052864076,
      "num_tokens": 64414692.0,
      "step": 29405
    },
    {
      "entropy": 5.649024915695191,
      "epoch": 2.9188169908693924,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004157887951303014,
      "loss": 5.0832,
      "mean_token_accuracy": 0.18105038106441498,
      "num_tokens": 64427278.0,
      "step": 29410
    },
    {
      "entropy": 5.641086673736572,
      "epoch": 2.91931321953156,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004157611553529822,
      "loss": 4.9725,
      "mean_token_accuracy": 0.18614497631788254,
      "num_tokens": 64438221.0,
      "step": 29415
    },
    {
      "entropy": 5.6682600498199465,
      "epoch": 2.919809448193728,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041573351208508966,
      "loss": 5.12,
      "mean_token_accuracy": 0.1799928814172745,
      "num_tokens": 64450139.0,
      "step": 29420
    },
    {
      "entropy": 5.656809282302857,
      "epoch": 2.9203056768558953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041570586532730926,
      "loss": 5.0477,
      "mean_token_accuracy": 0.1744733139872551,
      "num_tokens": 64461523.0,
      "step": 29425
    },
    {
      "entropy": 5.683545637130737,
      "epoch": 2.9208019055180627,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004156782150803266,
      "loss": 5.1036,
      "mean_token_accuracy": 0.17522041201591493,
      "num_tokens": 64474206.0,
      "step": 29430
    },
    {
      "entropy": 5.685872888565063,
      "epoch": 2.92129813418023,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041565056134482733,
      "loss": 5.0496,
      "mean_token_accuracy": 0.18029904067516328,
      "num_tokens": 64485585.0,
      "step": 29435
    },
    {
      "entropy": 5.716071844100952,
      "epoch": 2.9217943628423977,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004156229041214972,
      "loss": 5.0881,
      "mean_token_accuracy": 0.18330438286066056,
      "num_tokens": 64497237.0,
      "step": 29440
    },
    {
      "entropy": 5.6431097984313965,
      "epoch": 2.922290591504565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004155952434110221,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18537342250347139,
      "num_tokens": 64508235.0,
      "step": 29445
    },
    {
      "entropy": 5.710089254379272,
      "epoch": 2.922786820166733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004155675792140879,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18636616915464402,
      "num_tokens": 64520345.0,
      "step": 29450
    },
    {
      "entropy": 5.699161767959595,
      "epoch": 2.9232830488289006,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004155399115313806,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1793277993798256,
      "num_tokens": 64531428.0,
      "step": 29455
    },
    {
      "entropy": 5.722307109832764,
      "epoch": 2.923779277491068,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041551224036358644,
      "loss": 5.0828,
      "mean_token_accuracy": 0.17317097634077072,
      "num_tokens": 64541489.0,
      "step": 29460
    },
    {
      "entropy": 5.625987243652344,
      "epoch": 2.9242755061532355,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041548456571139145,
      "loss": 4.9684,
      "mean_token_accuracy": 0.18959822207689286,
      "num_tokens": 64553026.0,
      "step": 29465
    },
    {
      "entropy": 5.705714178085327,
      "epoch": 2.924771734815403,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041545688757548184,
      "loss": 5.1388,
      "mean_token_accuracy": 0.17593677192926407,
      "num_tokens": 64565988.0,
      "step": 29470
    },
    {
      "entropy": 5.686604261398315,
      "epoch": 2.9252679634775705,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041542920595654423,
      "loss": 5.016,
      "mean_token_accuracy": 0.18469203561544417,
      "num_tokens": 64578098.0,
      "step": 29475
    },
    {
      "entropy": 5.610626697540283,
      "epoch": 2.925764192139738,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041540152085526476,
      "loss": 4.9317,
      "mean_token_accuracy": 0.1813234269618988,
      "num_tokens": 64588931.0,
      "step": 29480
    },
    {
      "entropy": 5.5904130935668945,
      "epoch": 2.9262604208019054,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041537383227233023,
      "loss": 4.9501,
      "mean_token_accuracy": 0.1931142270565033,
      "num_tokens": 64599636.0,
      "step": 29485
    },
    {
      "entropy": 5.631572961807251,
      "epoch": 2.926756649464073,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004153461402084271,
      "loss": 5.058,
      "mean_token_accuracy": 0.18327219784259796,
      "num_tokens": 64609355.0,
      "step": 29490
    },
    {
      "entropy": 5.625305843353272,
      "epoch": 2.9272528781262404,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004153184446642421,
      "loss": 5.0147,
      "mean_token_accuracy": 0.19080494195222855,
      "num_tokens": 64619969.0,
      "step": 29495
    },
    {
      "entropy": 5.663408517837524,
      "epoch": 2.927749106788408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041529074564046204,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18987379521131514,
      "num_tokens": 64629538.0,
      "step": 29500
    },
    {
      "entropy": 5.689943504333496,
      "epoch": 2.9282453354505757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004152630431377738,
      "loss": 5.035,
      "mean_token_accuracy": 0.18784327507019044,
      "num_tokens": 64640255.0,
      "step": 29505
    },
    {
      "entropy": 5.616406059265136,
      "epoch": 2.9287415641127432,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004152353371568643,
      "loss": 4.9454,
      "mean_token_accuracy": 0.1878476545214653,
      "num_tokens": 64650051.0,
      "step": 29510
    },
    {
      "entropy": 5.592715167999268,
      "epoch": 2.9292377927749107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041520762769842065,
      "loss": 4.9981,
      "mean_token_accuracy": 0.19343506395816804,
      "num_tokens": 64660575.0,
      "step": 29515
    },
    {
      "entropy": 5.681214618682861,
      "epoch": 2.929734021437078,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041517991476312994,
      "loss": 5.0587,
      "mean_token_accuracy": 0.1849284812808037,
      "num_tokens": 64671486.0,
      "step": 29520
    },
    {
      "entropy": 5.652558040618897,
      "epoch": 2.9302302500992456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004151521983516794,
      "loss": 4.9792,
      "mean_token_accuracy": 0.1815241754055023,
      "num_tokens": 64682315.0,
      "step": 29525
    },
    {
      "entropy": 5.696198654174805,
      "epoch": 2.930726478761413,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004151244784647563,
      "loss": 5.1006,
      "mean_token_accuracy": 0.18339456170797347,
      "num_tokens": 64693940.0,
      "step": 29530
    },
    {
      "entropy": 5.657015228271485,
      "epoch": 2.931222707423581,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004150967551030482,
      "loss": 4.9901,
      "mean_token_accuracy": 0.18803468942642212,
      "num_tokens": 64703347.0,
      "step": 29535
    },
    {
      "entropy": 5.655773544311524,
      "epoch": 2.9317189360857485,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041506902826724234,
      "loss": 5.077,
      "mean_token_accuracy": 0.1772453546524048,
      "num_tokens": 64714115.0,
      "step": 29540
    },
    {
      "entropy": 5.631459283828735,
      "epoch": 2.932215164747916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004150412979580264,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18473396301269532,
      "num_tokens": 64724888.0,
      "step": 29545
    },
    {
      "entropy": 5.705251884460449,
      "epoch": 2.9327113934100835,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041501356417608816,
      "loss": 5.0708,
      "mean_token_accuracy": 0.18407013416290283,
      "num_tokens": 64735626.0,
      "step": 29550
    },
    {
      "entropy": 5.681424570083618,
      "epoch": 2.933207622072251,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041498582692211513,
      "loss": 4.992,
      "mean_token_accuracy": 0.1867678865790367,
      "num_tokens": 64747807.0,
      "step": 29555
    },
    {
      "entropy": 5.610938930511475,
      "epoch": 2.9337038507344184,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041495808619679525,
      "loss": 4.9742,
      "mean_token_accuracy": 0.18786593079566954,
      "num_tokens": 64756845.0,
      "step": 29560
    },
    {
      "entropy": 5.603854370117188,
      "epoch": 2.934200079396586,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041493034200081643,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18431503176689149,
      "num_tokens": 64767316.0,
      "step": 29565
    },
    {
      "entropy": 5.6465226173400875,
      "epoch": 2.9346963080587534,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004149025943348666,
      "loss": 4.9994,
      "mean_token_accuracy": 0.1872343122959137,
      "num_tokens": 64778329.0,
      "step": 29570
    },
    {
      "entropy": 5.653970098495483,
      "epoch": 2.935192536720921,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041487484319963395,
      "loss": 5.0542,
      "mean_token_accuracy": 0.1841172307729721,
      "num_tokens": 64789215.0,
      "step": 29575
    },
    {
      "entropy": 5.648893356323242,
      "epoch": 2.9356887653830883,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004148470885958066,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18210098445415496,
      "num_tokens": 64801878.0,
      "step": 29580
    },
    {
      "entropy": 5.585911417007447,
      "epoch": 2.936184994045256,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004148193305240727,
      "loss": 4.9158,
      "mean_token_accuracy": 0.1871993899345398,
      "num_tokens": 64813442.0,
      "step": 29585
    },
    {
      "entropy": 5.646575689315796,
      "epoch": 2.9366812227074237,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041479156898512076,
      "loss": 4.994,
      "mean_token_accuracy": 0.18780707120895385,
      "num_tokens": 64824095.0,
      "step": 29590
    },
    {
      "entropy": 5.65571928024292,
      "epoch": 2.937177451369591,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004147638039796391,
      "loss": 5.0331,
      "mean_token_accuracy": 0.19054831117391585,
      "num_tokens": 64833101.0,
      "step": 29595
    },
    {
      "entropy": 5.550573253631592,
      "epoch": 2.9376736800317587,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041473603550831636,
      "loss": 4.9534,
      "mean_token_accuracy": 0.18564011603593827,
      "num_tokens": 64844315.0,
      "step": 29600
    },
    {
      "entropy": 5.6785698413848875,
      "epoch": 2.938169908693926,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041470826357184096,
      "loss": 5.0543,
      "mean_token_accuracy": 0.1834315165877342,
      "num_tokens": 64855090.0,
      "step": 29605
    },
    {
      "entropy": 5.612608289718628,
      "epoch": 2.9386661373560936,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004146804881709017,
      "loss": 4.987,
      "mean_token_accuracy": 0.18278390616178514,
      "num_tokens": 64865093.0,
      "step": 29610
    },
    {
      "entropy": 5.743820238113403,
      "epoch": 2.939162366018261,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004146527093061873,
      "loss": 5.0851,
      "mean_token_accuracy": 0.17973996698856354,
      "num_tokens": 64876250.0,
      "step": 29615
    },
    {
      "entropy": 5.696481037139892,
      "epoch": 2.939658594680429,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004146249269783867,
      "loss": 5.0362,
      "mean_token_accuracy": 0.1856662541627884,
      "num_tokens": 64886339.0,
      "step": 29620
    },
    {
      "entropy": 5.690748882293701,
      "epoch": 2.9401548233425965,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004145971411881886,
      "loss": 5.1081,
      "mean_token_accuracy": 0.18095911741256715,
      "num_tokens": 64899473.0,
      "step": 29625
    },
    {
      "entropy": 5.668385028839111,
      "epoch": 2.940651052004764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041456935193628236,
      "loss": 5.0549,
      "mean_token_accuracy": 0.18000100404024125,
      "num_tokens": 64909919.0,
      "step": 29630
    },
    {
      "entropy": 5.717376804351806,
      "epoch": 2.9411472806669314,
      "grad_norm": 1.25,
      "learning_rate": 0.0004145415592233569,
      "loss": 5.0569,
      "mean_token_accuracy": 0.18056776821613313,
      "num_tokens": 64921069.0,
      "step": 29635
    },
    {
      "entropy": 5.602607250213623,
      "epoch": 2.941643509329099,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041451376305010144,
      "loss": 4.9868,
      "mean_token_accuracy": 0.1944720208644867,
      "num_tokens": 64932173.0,
      "step": 29640
    },
    {
      "entropy": 5.640494012832642,
      "epoch": 2.9421397379912664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004144859634172053,
      "loss": 4.9443,
      "mean_token_accuracy": 0.1917568176984787,
      "num_tokens": 64941650.0,
      "step": 29645
    },
    {
      "entropy": 5.588782453536988,
      "epoch": 2.942635966653434,
      "grad_norm": 1.0,
      "learning_rate": 0.00041445816032535784,
      "loss": 4.9793,
      "mean_token_accuracy": 0.19284473210573197,
      "num_tokens": 64952520.0,
      "step": 29650
    },
    {
      "entropy": 5.6212489128112795,
      "epoch": 2.9431321953156013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041443035377524844,
      "loss": 5.1219,
      "mean_token_accuracy": 0.17634470760822296,
      "num_tokens": 64965030.0,
      "step": 29655
    },
    {
      "entropy": 5.580777978897094,
      "epoch": 2.943628423977769,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004144025437675668,
      "loss": 4.9695,
      "mean_token_accuracy": 0.1873398095369339,
      "num_tokens": 64976260.0,
      "step": 29660
    },
    {
      "entropy": 5.699671030044556,
      "epoch": 2.9441246526399363,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004143747303030025,
      "loss": 5.1245,
      "mean_token_accuracy": 0.1821952521800995,
      "num_tokens": 64988049.0,
      "step": 29665
    },
    {
      "entropy": 5.627034330368042,
      "epoch": 2.9446208813021038,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041434691338224506,
      "loss": 4.9865,
      "mean_token_accuracy": 0.18782857060432434,
      "num_tokens": 64998506.0,
      "step": 29670
    },
    {
      "entropy": 5.651325702667236,
      "epoch": 2.9451171099642717,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004143190930059846,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18650370240211486,
      "num_tokens": 65010533.0,
      "step": 29675
    },
    {
      "entropy": 5.727979946136474,
      "epoch": 2.945613338626439,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142912691749107,
      "loss": 5.0646,
      "mean_token_accuracy": 0.18304260522127153,
      "num_tokens": 65021238.0,
      "step": 29680
    },
    {
      "entropy": 5.725074195861817,
      "epoch": 2.9461095672886066,
      "grad_norm": 1.125,
      "learning_rate": 0.0004142634418897135,
      "loss": 5.1079,
      "mean_token_accuracy": 0.17757664620876312,
      "num_tokens": 65033251.0,
      "step": 29685
    },
    {
      "entropy": 5.6752996921539305,
      "epoch": 2.946605795950774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004142356111510831,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18059927970170975,
      "num_tokens": 65043845.0,
      "step": 29690
    },
    {
      "entropy": 5.592424535751343,
      "epoch": 2.9471020246129416,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041420777695970945,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18948636054992676,
      "num_tokens": 65055395.0,
      "step": 29695
    },
    {
      "entropy": 5.674594593048096,
      "epoch": 2.947598253275109,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000414179939316283,
      "loss": 5.0495,
      "mean_token_accuracy": 0.18116589933633803,
      "num_tokens": 65064931.0,
      "step": 29700
    },
    {
      "entropy": 5.697499322891235,
      "epoch": 2.948094481937277,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041415209822149397,
      "loss": 5.0256,
      "mean_token_accuracy": 0.1875823050737381,
      "num_tokens": 65074678.0,
      "step": 29705
    },
    {
      "entropy": 5.6817307472229,
      "epoch": 2.9485907105994444,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004141242536760327,
      "loss": 5.0364,
      "mean_token_accuracy": 0.18774397820234298,
      "num_tokens": 65084562.0,
      "step": 29710
    },
    {
      "entropy": 5.657258415222168,
      "epoch": 2.949086939261612,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004140964056805897,
      "loss": 5.0243,
      "mean_token_accuracy": 0.18457545787096025,
      "num_tokens": 65095924.0,
      "step": 29715
    },
    {
      "entropy": 5.659221220016479,
      "epoch": 2.9495831679237794,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004140685542358556,
      "loss": 5.0513,
      "mean_token_accuracy": 0.18823993653059007,
      "num_tokens": 65106802.0,
      "step": 29720
    },
    {
      "entropy": 5.661164855957031,
      "epoch": 2.950079396585947,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004140406993425211,
      "loss": 5.0887,
      "mean_token_accuracy": 0.17946268618106842,
      "num_tokens": 65118196.0,
      "step": 29725
    },
    {
      "entropy": 5.65198130607605,
      "epoch": 2.9505756252481143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004140128410012767,
      "loss": 5.0076,
      "mean_token_accuracy": 0.19226383715867995,
      "num_tokens": 65129924.0,
      "step": 29730
    },
    {
      "entropy": 5.561973333358765,
      "epoch": 2.951071853910282,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041398497921281354,
      "loss": 4.8814,
      "mean_token_accuracy": 0.19712416976690292,
      "num_tokens": 65140994.0,
      "step": 29735
    },
    {
      "entropy": 5.6689615726470945,
      "epoch": 2.9515680825724493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041395711397782236,
      "loss": 5.1016,
      "mean_token_accuracy": 0.18157809972763062,
      "num_tokens": 65153261.0,
      "step": 29740
    },
    {
      "entropy": 5.749296188354492,
      "epoch": 2.9520643112346168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004139292452969941,
      "loss": 5.1556,
      "mean_token_accuracy": 0.1789582595229149,
      "num_tokens": 65164127.0,
      "step": 29745
    },
    {
      "entropy": 5.620865297317505,
      "epoch": 2.9525605398967842,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041390137317102,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1839592933654785,
      "num_tokens": 65175603.0,
      "step": 29750
    },
    {
      "entropy": 5.654139614105224,
      "epoch": 2.9530567685589517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004138734976005911,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18384457677602767,
      "num_tokens": 65186780.0,
      "step": 29755
    },
    {
      "entropy": 5.604316282272339,
      "epoch": 2.9535529972211196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004138456185863986,
      "loss": 4.9252,
      "mean_token_accuracy": 0.19064365178346634,
      "num_tokens": 65198883.0,
      "step": 29760
    },
    {
      "entropy": 5.620806503295898,
      "epoch": 2.954049225883287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041381773612913413,
      "loss": 5.031,
      "mean_token_accuracy": 0.1813298612833023,
      "num_tokens": 65209088.0,
      "step": 29765
    },
    {
      "entropy": 5.6225076675415036,
      "epoch": 2.9545454545454546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004137898502294889,
      "loss": 4.9963,
      "mean_token_accuracy": 0.1889519140124321,
      "num_tokens": 65220816.0,
      "step": 29770
    },
    {
      "entropy": 5.683761119842529,
      "epoch": 2.955041683207622,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004137619608881544,
      "loss": 4.9949,
      "mean_token_accuracy": 0.18435115069150926,
      "num_tokens": 65232549.0,
      "step": 29775
    },
    {
      "entropy": 5.679590511322021,
      "epoch": 2.9555379118697895,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004137340681058223,
      "loss": 5.0461,
      "mean_token_accuracy": 0.18133591562509538,
      "num_tokens": 65243691.0,
      "step": 29780
    },
    {
      "entropy": 5.6534124374389645,
      "epoch": 2.956034140531957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041370617188318424,
      "loss": 5.0544,
      "mean_token_accuracy": 0.18623193204402924,
      "num_tokens": 65254651.0,
      "step": 29785
    },
    {
      "entropy": 5.625347137451172,
      "epoch": 2.956530369194125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041367827222093195,
      "loss": 5.0022,
      "mean_token_accuracy": 0.19189587831497193,
      "num_tokens": 65264669.0,
      "step": 29790
    },
    {
      "entropy": 5.649659156799316,
      "epoch": 2.9570265978562924,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004136503691197574,
      "loss": 5.0271,
      "mean_token_accuracy": 0.18347920924425126,
      "num_tokens": 65274663.0,
      "step": 29795
    },
    {
      "entropy": 5.66598596572876,
      "epoch": 2.95752282651846,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041362246258035254,
      "loss": 5.0462,
      "mean_token_accuracy": 0.18502351492643357,
      "num_tokens": 65286383.0,
      "step": 29800
    },
    {
      "entropy": 5.5843769073486325,
      "epoch": 2.9580190551806274,
      "grad_norm": 1.125,
      "learning_rate": 0.0004135945526034091,
      "loss": 4.8821,
      "mean_token_accuracy": 0.18572946786880493,
      "num_tokens": 65297156.0,
      "step": 29805
    },
    {
      "entropy": 5.67445478439331,
      "epoch": 2.958515283842795,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00041356663918961956,
      "loss": 5.0726,
      "mean_token_accuracy": 0.18334475606679917,
      "num_tokens": 65306836.0,
      "step": 29810
    },
    {
      "entropy": 5.692247819900513,
      "epoch": 2.9590115125049623,
      "grad_norm": 1.125,
      "learning_rate": 0.00041353872233967595,
      "loss": 5.0861,
      "mean_token_accuracy": 0.17856210321187974,
      "num_tokens": 65318355.0,
      "step": 29815
    },
    {
      "entropy": 5.6693014144897464,
      "epoch": 2.95950774116713,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004135108020542705,
      "loss": 5.0634,
      "mean_token_accuracy": 0.18380436450242996,
      "num_tokens": 65330039.0,
      "step": 29820
    },
    {
      "entropy": 5.606721591949463,
      "epoch": 2.9600039698292973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041348287833409557,
      "loss": 4.9577,
      "mean_token_accuracy": 0.18798955231904985,
      "num_tokens": 65339771.0,
      "step": 29825
    },
    {
      "entropy": 5.63840389251709,
      "epoch": 2.9605001984914647,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004134549511798438,
      "loss": 5.0414,
      "mean_token_accuracy": 0.1860576093196869,
      "num_tokens": 65352108.0,
      "step": 29830
    },
    {
      "entropy": 5.64449667930603,
      "epoch": 2.960996427153632,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004134270205922075,
      "loss": 4.9898,
      "mean_token_accuracy": 0.17971233576536177,
      "num_tokens": 65364117.0,
      "step": 29835
    },
    {
      "entropy": 5.640604162216187,
      "epoch": 2.9614926558157997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041339908657187933,
      "loss": 4.9677,
      "mean_token_accuracy": 0.18828944712877274,
      "num_tokens": 65374168.0,
      "step": 29840
    },
    {
      "entropy": 5.747609233856201,
      "epoch": 2.9619888844779676,
      "grad_norm": 1.25,
      "learning_rate": 0.0004133711491195522,
      "loss": 5.1698,
      "mean_token_accuracy": 0.18060049563646316,
      "num_tokens": 65385243.0,
      "step": 29845
    },
    {
      "entropy": 5.688323307037353,
      "epoch": 2.962485113140135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041334320823591857,
      "loss": 5.0222,
      "mean_token_accuracy": 0.18187082260847093,
      "num_tokens": 65396674.0,
      "step": 29850
    },
    {
      "entropy": 5.7102080345153805,
      "epoch": 2.9629813418023025,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041331526392167155,
      "loss": 5.0855,
      "mean_token_accuracy": 0.17345467805862427,
      "num_tokens": 65408121.0,
      "step": 29855
    },
    {
      "entropy": 5.684891176223755,
      "epoch": 2.96347757046447,
      "grad_norm": 0.984375,
      "learning_rate": 0.000413287316177504,
      "loss": 5.0443,
      "mean_token_accuracy": 0.1819643944501877,
      "num_tokens": 65420093.0,
      "step": 29860
    },
    {
      "entropy": 5.645544528961182,
      "epoch": 2.9639737991266375,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041325936500410905,
      "loss": 4.9808,
      "mean_token_accuracy": 0.18311737179756166,
      "num_tokens": 65431338.0,
      "step": 29865
    },
    {
      "entropy": 5.630655241012573,
      "epoch": 2.964470027788805,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041323141040217976,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18453021347522736,
      "num_tokens": 65442305.0,
      "step": 29870
    },
    {
      "entropy": 5.5725878238677975,
      "epoch": 2.9649662564509724,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00041320345237240923,
      "loss": 4.9535,
      "mean_token_accuracy": 0.18979968577623368,
      "num_tokens": 65453028.0,
      "step": 29875
    },
    {
      "entropy": 5.63998646736145,
      "epoch": 2.9654624851131404,
      "grad_norm": 1.234375,
      "learning_rate": 0.000413175490915491,
      "loss": 5.0669,
      "mean_token_accuracy": 0.18852283507585527,
      "num_tokens": 65463374.0,
      "step": 29880
    },
    {
      "entropy": 5.633168506622314,
      "epoch": 2.965958713775308,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004131475260321183,
      "loss": 4.9823,
      "mean_token_accuracy": 0.1831694334745407,
      "num_tokens": 65474144.0,
      "step": 29885
    },
    {
      "entropy": 5.597996807098388,
      "epoch": 2.9664549424374753,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041311955772298464,
      "loss": 4.9664,
      "mean_token_accuracy": 0.20355918258428574,
      "num_tokens": 65485453.0,
      "step": 29890
    },
    {
      "entropy": 5.674776268005371,
      "epoch": 2.966951171099643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004130915859887837,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18516744077205657,
      "num_tokens": 65496400.0,
      "step": 29895
    },
    {
      "entropy": 5.68133020401001,
      "epoch": 2.9674473997618103,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004130636108302088,
      "loss": 5.0679,
      "mean_token_accuracy": 0.18042637556791305,
      "num_tokens": 65508524.0,
      "step": 29900
    },
    {
      "entropy": 5.5019166469573975,
      "epoch": 2.9679436284239777,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041303563224795393,
      "loss": 4.9139,
      "mean_token_accuracy": 0.19387867897748948,
      "num_tokens": 65519264.0,
      "step": 29905
    },
    {
      "entropy": 5.578285646438599,
      "epoch": 2.968439857086145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004130076502427128,
      "loss": 4.9244,
      "mean_token_accuracy": 0.19192849099636078,
      "num_tokens": 65531049.0,
      "step": 29910
    },
    {
      "entropy": 5.653775596618653,
      "epoch": 2.9689360857483127,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041297966481517937,
      "loss": 5.0524,
      "mean_token_accuracy": 0.1825059399008751,
      "num_tokens": 65543126.0,
      "step": 29915
    },
    {
      "entropy": 5.729743623733521,
      "epoch": 2.96943231441048,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004129516759660475,
      "loss": 5.1074,
      "mean_token_accuracy": 0.1709002286195755,
      "num_tokens": 65555050.0,
      "step": 29920
    },
    {
      "entropy": 5.706798267364502,
      "epoch": 2.9699285430726476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00041292368369601136,
      "loss": 5.0912,
      "mean_token_accuracy": 0.18939679116010666,
      "num_tokens": 65564156.0,
      "step": 29925
    },
    {
      "entropy": 5.701503705978394,
      "epoch": 2.9704247717348156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412895688005765,
      "loss": 5.1153,
      "mean_token_accuracy": 0.18079415559768677,
      "num_tokens": 65574639.0,
      "step": 29930
    },
    {
      "entropy": 5.674201440811157,
      "epoch": 2.970921000396983,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004128676888960028,
      "loss": 4.9682,
      "mean_token_accuracy": 0.18827724307775498,
      "num_tokens": 65585628.0,
      "step": 29935
    },
    {
      "entropy": 5.618797826766968,
      "epoch": 2.9714172290591505,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041283968636741903,
      "loss": 4.9915,
      "mean_token_accuracy": 0.19046924710273744,
      "num_tokens": 65595544.0,
      "step": 29940
    },
    {
      "entropy": 5.665066385269165,
      "epoch": 2.971913457721318,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004128116804207081,
      "loss": 5.0754,
      "mean_token_accuracy": 0.18459194898605347,
      "num_tokens": 65607963.0,
      "step": 29945
    },
    {
      "entropy": 5.710080146789551,
      "epoch": 2.9724096863834855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004127836710565642,
      "loss": 4.9862,
      "mean_token_accuracy": 0.19943797886371611,
      "num_tokens": 65619881.0,
      "step": 29950
    },
    {
      "entropy": 5.707682514190674,
      "epoch": 2.972905915045653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00041275565827568234,
      "loss": 5.0822,
      "mean_token_accuracy": 0.18104959428310394,
      "num_tokens": 65629967.0,
      "step": 29955
    },
    {
      "entropy": 5.6624949932098385,
      "epoch": 2.9734021437078204,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000412727642078757,
      "loss": 5.0227,
      "mean_token_accuracy": 0.18123858720064162,
      "num_tokens": 65640804.0,
      "step": 29960
    },
    {
      "entropy": 5.681999635696411,
      "epoch": 2.9738983723699883,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004126996224664829,
      "loss": 5.101,
      "mean_token_accuracy": 0.1752261370420456,
      "num_tokens": 65652355.0,
      "step": 29965
    },
    {
      "entropy": 5.679146432876587,
      "epoch": 2.974394601032156,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041267159943955483,
      "loss": 5.0638,
      "mean_token_accuracy": 0.1752949208021164,
      "num_tokens": 65664047.0,
      "step": 29970
    },
    {
      "entropy": 5.680718517303466,
      "epoch": 2.9748908296943233,
      "grad_norm": 1.125,
      "learning_rate": 0.0004126435729986678,
      "loss": 5.015,
      "mean_token_accuracy": 0.1858601450920105,
      "num_tokens": 65674942.0,
      "step": 29975
    },
    {
      "entropy": 5.640568542480469,
      "epoch": 2.9753870583564908,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041261554314451674,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1877062901854515,
      "num_tokens": 65685285.0,
      "step": 29980
    },
    {
      "entropy": 5.582769441604614,
      "epoch": 2.9758832870186582,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004125875098777967,
      "loss": 4.9533,
      "mean_token_accuracy": 0.19693800210952758,
      "num_tokens": 65695967.0,
      "step": 29985
    },
    {
      "entropy": 5.62844443321228,
      "epoch": 2.9763795156808257,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041255947319920297,
      "loss": 5.0236,
      "mean_token_accuracy": 0.18604697287082672,
      "num_tokens": 65707434.0,
      "step": 29990
    },
    {
      "entropy": 5.610231971740722,
      "epoch": 2.976875744342993,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041253143310943064,
      "loss": 4.9811,
      "mean_token_accuracy": 0.19445813447237015,
      "num_tokens": 65717836.0,
      "step": 29995
    },
    {
      "entropy": 5.674016380310059,
      "epoch": 2.9773719730051607,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00041250338960917515,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18435851335525513,
      "num_tokens": 65728527.0,
      "step": 30000
    },
    {
      "epoch": 2.9773719730051607,
      "eval_entropy": 5.4294826510287315,
      "eval_loss": 5.1319074630737305,
      "eval_mean_token_accuracy": 0.18538569771974792,
      "eval_num_tokens": 65728527.0,
      "eval_runtime": 20.4207,
      "eval_samples_per_second": 1648.771,
      "eval_steps_per_second": 206.115,
      "step": 30000
    },
    {
      "entropy": 5.622583341598511,
      "epoch": 2.977868201667328,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041247534269913197,
      "loss": 4.9837,
      "mean_token_accuracy": 0.18958968818187713,
      "num_tokens": 65739724.0,
      "step": 30005
    },
    {
      "entropy": 5.682161140441894,
      "epoch": 2.9783644303294956,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004124472923799964,
      "loss": 5.0591,
      "mean_token_accuracy": 0.17975455522537231,
      "num_tokens": 65750057.0,
      "step": 30010
    },
    {
      "entropy": 5.6726487159729,
      "epoch": 2.9788606589916635,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004124192386524642,
      "loss": 5.039,
      "mean_token_accuracy": 0.18840931206941605,
      "num_tokens": 65760534.0,
      "step": 30015
    },
    {
      "entropy": 5.642183494567871,
      "epoch": 2.979356887653831,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000412391181517231,
      "loss": 5.0824,
      "mean_token_accuracy": 0.18256707340478898,
      "num_tokens": 65772524.0,
      "step": 30020
    },
    {
      "entropy": 5.665837574005127,
      "epoch": 2.9798531163159985,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041236312097499257,
      "loss": 5.0218,
      "mean_token_accuracy": 0.18213433921337127,
      "num_tokens": 65783003.0,
      "step": 30025
    },
    {
      "entropy": 5.603643417358398,
      "epoch": 2.980349344978166,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004123350570264447,
      "loss": 5.0221,
      "mean_token_accuracy": 0.18002210408449174,
      "num_tokens": 65793219.0,
      "step": 30030
    },
    {
      "entropy": 5.70863904953003,
      "epoch": 2.9808455736403334,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041230698967228334,
      "loss": 5.1279,
      "mean_token_accuracy": 0.1842878982424736,
      "num_tokens": 65804105.0,
      "step": 30035
    },
    {
      "entropy": 5.81916823387146,
      "epoch": 2.981341802302501,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004122789189132045,
      "loss": 5.1098,
      "mean_token_accuracy": 0.18148695379495622,
      "num_tokens": 65814211.0,
      "step": 30040
    },
    {
      "entropy": 5.759814834594726,
      "epoch": 2.9818380309646684,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004122508447499043,
      "loss": 5.1235,
      "mean_token_accuracy": 0.17379141598939896,
      "num_tokens": 65825304.0,
      "step": 30045
    },
    {
      "entropy": 5.629247951507568,
      "epoch": 2.9823342596268363,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041222276718307896,
      "loss": 4.968,
      "mean_token_accuracy": 0.18933085203170777,
      "num_tokens": 65835933.0,
      "step": 30050
    },
    {
      "entropy": 5.63326473236084,
      "epoch": 2.9828304882890038,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004121946862134246,
      "loss": 5.0413,
      "mean_token_accuracy": 0.18617455661296844,
      "num_tokens": 65846479.0,
      "step": 30055
    },
    {
      "entropy": 5.62682876586914,
      "epoch": 2.9833267169511712,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041216660184163767,
      "loss": 4.9623,
      "mean_token_accuracy": 0.18845667988061904,
      "num_tokens": 65857033.0,
      "step": 30060
    },
    {
      "entropy": 5.704694128036499,
      "epoch": 2.9838229456133387,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041213851406841456,
      "loss": 5.1245,
      "mean_token_accuracy": 0.17673497498035431,
      "num_tokens": 65867983.0,
      "step": 30065
    },
    {
      "entropy": 5.685877180099487,
      "epoch": 2.984319174275506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004121104228944518,
      "loss": 5.0623,
      "mean_token_accuracy": 0.1762468621134758,
      "num_tokens": 65878535.0,
      "step": 30070
    },
    {
      "entropy": 5.702661180496216,
      "epoch": 2.9848154029376737,
      "grad_norm": 1.125,
      "learning_rate": 0.0004120823283204461,
      "loss": 5.0359,
      "mean_token_accuracy": 0.186234450340271,
      "num_tokens": 65890484.0,
      "step": 30075
    },
    {
      "entropy": 5.618539571762085,
      "epoch": 2.985311631599841,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000412054230347094,
      "loss": 4.993,
      "mean_token_accuracy": 0.18598960638046264,
      "num_tokens": 65901220.0,
      "step": 30080
    },
    {
      "entropy": 5.678935813903808,
      "epoch": 2.9858078602620086,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004120261289750923,
      "loss": 5.0369,
      "mean_token_accuracy": 0.18189896643161774,
      "num_tokens": 65911875.0,
      "step": 30085
    },
    {
      "entropy": 5.659861278533936,
      "epoch": 2.986304088924176,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041199802420513785,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18768321126699447,
      "num_tokens": 65921915.0,
      "step": 30090
    },
    {
      "entropy": 5.611316776275634,
      "epoch": 2.9868003175863436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004119699160379276,
      "loss": 4.9799,
      "mean_token_accuracy": 0.18261648863554,
      "num_tokens": 65932061.0,
      "step": 30095
    },
    {
      "entropy": 5.626504707336426,
      "epoch": 2.987296546248511,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041194180447415855,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18927524834871293,
      "num_tokens": 65942818.0,
      "step": 30100
    },
    {
      "entropy": 5.650602865219116,
      "epoch": 2.987792774910679,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004119136895145279,
      "loss": 5.008,
      "mean_token_accuracy": 0.18767224699258805,
      "num_tokens": 65953715.0,
      "step": 30105
    },
    {
      "entropy": 5.7375153541564945,
      "epoch": 2.9882890035728464,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004118855711597327,
      "loss": 5.0648,
      "mean_token_accuracy": 0.18180398195981978,
      "num_tokens": 65964235.0,
      "step": 30110
    },
    {
      "entropy": 5.632211828231812,
      "epoch": 2.988785232235014,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041185744941047033,
      "loss": 5.0378,
      "mean_token_accuracy": 0.1812919020652771,
      "num_tokens": 65976457.0,
      "step": 30115
    },
    {
      "entropy": 5.643520355224609,
      "epoch": 2.9892814608971814,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004118293242674382,
      "loss": 5.0118,
      "mean_token_accuracy": 0.18570461571216584,
      "num_tokens": 65986810.0,
      "step": 30120
    },
    {
      "entropy": 5.678421926498413,
      "epoch": 2.989777689559349,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004118011957313336,
      "loss": 5.1133,
      "mean_token_accuracy": 0.17706471383571626,
      "num_tokens": 65998019.0,
      "step": 30125
    },
    {
      "entropy": 5.714756298065185,
      "epoch": 2.9902739182215163,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041177306380285405,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18523832857608796,
      "num_tokens": 66009404.0,
      "step": 30130
    },
    {
      "entropy": 5.614530849456787,
      "epoch": 2.9907701468836843,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004117449284826974,
      "loss": 4.9659,
      "mean_token_accuracy": 0.18642411828041078,
      "num_tokens": 66020610.0,
      "step": 30135
    },
    {
      "entropy": 5.674035358428955,
      "epoch": 2.9912663755458517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041171678977156104,
      "loss": 5.0582,
      "mean_token_accuracy": 0.1787695422768593,
      "num_tokens": 66032576.0,
      "step": 30140
    },
    {
      "entropy": 5.667307186126709,
      "epoch": 2.991762604208019,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041168864767014295,
      "loss": 5.0281,
      "mean_token_accuracy": 0.1943618193268776,
      "num_tokens": 66043378.0,
      "step": 30145
    },
    {
      "entropy": 5.672971200942993,
      "epoch": 2.9922588328701867,
      "grad_norm": 1.203125,
      "learning_rate": 0.000411660502179141,
      "loss": 4.9575,
      "mean_token_accuracy": 0.1874187707901001,
      "num_tokens": 66053960.0,
      "step": 30150
    },
    {
      "entropy": 5.6484767436981205,
      "epoch": 2.992755061532354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004116323532992529,
      "loss": 5.0437,
      "mean_token_accuracy": 0.17946202903985978,
      "num_tokens": 66066069.0,
      "step": 30155
    },
    {
      "entropy": 5.63858470916748,
      "epoch": 2.9932512901945216,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041160420103117704,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18796855211257935,
      "num_tokens": 66076789.0,
      "step": 30160
    },
    {
      "entropy": 5.662926101684571,
      "epoch": 2.993747518856689,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041157604537561123,
      "loss": 4.9777,
      "mean_token_accuracy": 0.18355180472135543,
      "num_tokens": 66088267.0,
      "step": 30165
    },
    {
      "entropy": 5.601974153518677,
      "epoch": 2.9942437475188566,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041154788633325387,
      "loss": 4.961,
      "mean_token_accuracy": 0.19024408757686614,
      "num_tokens": 66098789.0,
      "step": 30170
    },
    {
      "entropy": 5.615823030471802,
      "epoch": 2.994739976181024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004115197239048031,
      "loss": 5.0161,
      "mean_token_accuracy": 0.1826931431889534,
      "num_tokens": 66110881.0,
      "step": 30175
    },
    {
      "entropy": 5.691059398651123,
      "epoch": 2.9952362048431915,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041149155809095734,
      "loss": 5.0228,
      "mean_token_accuracy": 0.17835796773433685,
      "num_tokens": 66121728.0,
      "step": 30180
    },
    {
      "entropy": 5.629473924636841,
      "epoch": 2.995732433505359,
      "grad_norm": 1.125,
      "learning_rate": 0.00041146338889241506,
      "loss": 4.9285,
      "mean_token_accuracy": 0.1896342009305954,
      "num_tokens": 66131386.0,
      "step": 30185
    },
    {
      "entropy": 5.672706556320191,
      "epoch": 2.996228662167527,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041143521630987483,
      "loss": 5.1072,
      "mean_token_accuracy": 0.17769033163785936,
      "num_tokens": 66142905.0,
      "step": 30190
    },
    {
      "entropy": 5.682532405853271,
      "epoch": 2.9967248908296944,
      "grad_norm": 1.234375,
      "learning_rate": 0.00041140704034403527,
      "loss": 5.1539,
      "mean_token_accuracy": 0.16826344132423401,
      "num_tokens": 66152708.0,
      "step": 30195
    },
    {
      "entropy": 5.773283529281616,
      "epoch": 2.997221119491862,
      "grad_norm": 1.109375,
      "learning_rate": 0.000411378860995595,
      "loss": 5.1449,
      "mean_token_accuracy": 0.17847787141799926,
      "num_tokens": 66164351.0,
      "step": 30200
    },
    {
      "entropy": 5.722042989730835,
      "epoch": 2.9977173481540293,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004113506782652528,
      "loss": 5.0679,
      "mean_token_accuracy": 0.18506928980350495,
      "num_tokens": 66176641.0,
      "step": 30205
    },
    {
      "entropy": 5.788472700119018,
      "epoch": 2.998213576816197,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004113224921537076,
      "loss": 5.119,
      "mean_token_accuracy": 0.17303275763988496,
      "num_tokens": 66187822.0,
      "step": 30210
    },
    {
      "entropy": 5.642163419723511,
      "epoch": 2.9987098054783643,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041129430266165836,
      "loss": 5.0306,
      "mean_token_accuracy": 0.17618827223777772,
      "num_tokens": 66199636.0,
      "step": 30215
    },
    {
      "entropy": 5.692378568649292,
      "epoch": 2.999206034140532,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004112661097898041,
      "loss": 5.0318,
      "mean_token_accuracy": 0.17919603288173674,
      "num_tokens": 66210649.0,
      "step": 30220
    },
    {
      "entropy": 5.68196063041687,
      "epoch": 2.9997022628026997,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041123791353884397,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18701947927474977,
      "num_tokens": 66221806.0,
      "step": 30225
    },
    {
      "entropy": 5.612423372268677,
      "epoch": 3.000198491464867,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004112097139094771,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18305500596761703,
      "num_tokens": 66232016.0,
      "step": 30230
    },
    {
      "entropy": 5.604811811447144,
      "epoch": 3.0006947201270346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041118151090240284,
      "loss": 4.9667,
      "mean_token_accuracy": 0.18448809981346131,
      "num_tokens": 66242852.0,
      "step": 30235
    },
    {
      "entropy": 5.680728816986084,
      "epoch": 3.001190948789202,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041115330451832065,
      "loss": 4.9816,
      "mean_token_accuracy": 0.19093921333551406,
      "num_tokens": 66253254.0,
      "step": 30240
    },
    {
      "entropy": 5.663672304153442,
      "epoch": 3.0016871774513696,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041112509475792983,
      "loss": 4.9679,
      "mean_token_accuracy": 0.18386608958244324,
      "num_tokens": 66262721.0,
      "step": 30245
    },
    {
      "entropy": 5.618712997436523,
      "epoch": 3.002183406113537,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041109688162193,
      "loss": 4.9413,
      "mean_token_accuracy": 0.1901197463274002,
      "num_tokens": 66273718.0,
      "step": 30250
    },
    {
      "entropy": 5.663272762298584,
      "epoch": 3.0026796347757045,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004110686651110207,
      "loss": 5.0159,
      "mean_token_accuracy": 0.1835177019238472,
      "num_tokens": 66285527.0,
      "step": 30255
    },
    {
      "entropy": 5.6792991161346436,
      "epoch": 3.003175863437872,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004110404452259017,
      "loss": 4.9807,
      "mean_token_accuracy": 0.18677655458450318,
      "num_tokens": 66295564.0,
      "step": 30260
    },
    {
      "entropy": 5.642493391036988,
      "epoch": 3.0036720921000395,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004110122219672729,
      "loss": 4.9439,
      "mean_token_accuracy": 0.1933341756463051,
      "num_tokens": 66306721.0,
      "step": 30265
    },
    {
      "entropy": 5.638416814804077,
      "epoch": 3.0041683207622074,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000410983995335834,
      "loss": 4.9591,
      "mean_token_accuracy": 0.19602895230054856,
      "num_tokens": 66316990.0,
      "step": 30270
    },
    {
      "entropy": 5.63905553817749,
      "epoch": 3.004664549424375,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000410955765332285,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18636680841445924,
      "num_tokens": 66327814.0,
      "step": 30275
    },
    {
      "entropy": 5.62063775062561,
      "epoch": 3.0051607780865424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041092753195732607,
      "loss": 4.9245,
      "mean_token_accuracy": 0.1932871624827385,
      "num_tokens": 66338373.0,
      "step": 30280
    },
    {
      "entropy": 5.6399554252624515,
      "epoch": 3.00565700674871,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041089929521165725,
      "loss": 5.013,
      "mean_token_accuracy": 0.19046285301446914,
      "num_tokens": 66349311.0,
      "step": 30285
    },
    {
      "entropy": 5.7360005378723145,
      "epoch": 3.0061532354108773,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004108710550959787,
      "loss": 4.9877,
      "mean_token_accuracy": 0.19014948904514312,
      "num_tokens": 66360771.0,
      "step": 30290
    },
    {
      "entropy": 5.677119588851928,
      "epoch": 3.006649464073045,
      "grad_norm": 1.125,
      "learning_rate": 0.00041084281161099073,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18261407762765886,
      "num_tokens": 66371348.0,
      "step": 30295
    },
    {
      "entropy": 5.661842250823975,
      "epoch": 3.0071456927352123,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004108145647573938,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1882560968399048,
      "num_tokens": 66382489.0,
      "step": 30300
    },
    {
      "entropy": 5.632399559020996,
      "epoch": 3.0076419213973797,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041078631453588834,
      "loss": 4.917,
      "mean_token_accuracy": 0.19442253261804582,
      "num_tokens": 66392825.0,
      "step": 30305
    },
    {
      "entropy": 5.681684875488282,
      "epoch": 3.0081381500595477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004107580609471748,
      "loss": 5.0067,
      "mean_token_accuracy": 0.1856796622276306,
      "num_tokens": 66403646.0,
      "step": 30310
    },
    {
      "entropy": 5.631747722625732,
      "epoch": 3.008634378721715,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041072980399195394,
      "loss": 4.9054,
      "mean_token_accuracy": 0.19617275446653365,
      "num_tokens": 66414199.0,
      "step": 30315
    },
    {
      "entropy": 5.596850967407226,
      "epoch": 3.0091306073838826,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004107015436709264,
      "loss": 4.8995,
      "mean_token_accuracy": 0.18979560136795043,
      "num_tokens": 66426331.0,
      "step": 30320
    },
    {
      "entropy": 5.616516542434693,
      "epoch": 3.00962683604605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041067327998479296,
      "loss": 4.9242,
      "mean_token_accuracy": 0.1894946202635765,
      "num_tokens": 66436854.0,
      "step": 30325
    },
    {
      "entropy": 5.641244697570801,
      "epoch": 3.0101230647082176,
      "grad_norm": 1.359375,
      "learning_rate": 0.00041064501293425453,
      "loss": 4.9974,
      "mean_token_accuracy": 0.18979464769363402,
      "num_tokens": 66447757.0,
      "step": 30330
    },
    {
      "entropy": 5.649451017379761,
      "epoch": 3.010619293370385,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004106167425200121,
      "loss": 4.9755,
      "mean_token_accuracy": 0.18936157822608948,
      "num_tokens": 66458395.0,
      "step": 30335
    },
    {
      "entropy": 5.645033121109009,
      "epoch": 3.0111155220325525,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004105884687427667,
      "loss": 4.8595,
      "mean_token_accuracy": 0.1985275223851204,
      "num_tokens": 66467728.0,
      "step": 30340
    },
    {
      "entropy": 5.6301007747650145,
      "epoch": 3.01161175069472,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004105601916032194,
      "loss": 4.9652,
      "mean_token_accuracy": 0.19401230067014694,
      "num_tokens": 66478113.0,
      "step": 30345
    },
    {
      "entropy": 5.663118028640747,
      "epoch": 3.0121079793568875,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004105319111020714,
      "loss": 4.9516,
      "mean_token_accuracy": 0.18892737925052644,
      "num_tokens": 66489055.0,
      "step": 30350
    },
    {
      "entropy": 5.611892461776733,
      "epoch": 3.0126042080190554,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041050362724002413,
      "loss": 4.9331,
      "mean_token_accuracy": 0.19184019416570663,
      "num_tokens": 66499516.0,
      "step": 30355
    },
    {
      "entropy": 5.628358316421509,
      "epoch": 3.013100436681223,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00041047534001777884,
      "loss": 4.9982,
      "mean_token_accuracy": 0.1905234530568123,
      "num_tokens": 66510832.0,
      "step": 30360
    },
    {
      "entropy": 5.732261753082275,
      "epoch": 3.0135966653433903,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004104470494360371,
      "loss": 5.0652,
      "mean_token_accuracy": 0.18165794014930725,
      "num_tokens": 66522457.0,
      "step": 30365
    },
    {
      "entropy": 5.682747554779053,
      "epoch": 3.014092894005558,
      "grad_norm": 1.375,
      "learning_rate": 0.00041041875549550043,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18874492943286897,
      "num_tokens": 66532527.0,
      "step": 30370
    },
    {
      "entropy": 5.721376228332519,
      "epoch": 3.0145891226677253,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004103904581968703,
      "loss": 4.9927,
      "mean_token_accuracy": 0.1872924044728279,
      "num_tokens": 66542437.0,
      "step": 30375
    },
    {
      "entropy": 5.59935450553894,
      "epoch": 3.0150853513298927,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041036215754084864,
      "loss": 4.9468,
      "mean_token_accuracy": 0.19122267812490462,
      "num_tokens": 66552927.0,
      "step": 30380
    },
    {
      "entropy": 5.650675916671753,
      "epoch": 3.0155815799920602,
      "grad_norm": 1.15625,
      "learning_rate": 0.00041033385352813714,
      "loss": 5.0608,
      "mean_token_accuracy": 0.1818948656320572,
      "num_tokens": 66563664.0,
      "step": 30385
    },
    {
      "entropy": 5.655502271652222,
      "epoch": 3.0160778086542277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004103055461594377,
      "loss": 4.9662,
      "mean_token_accuracy": 0.19154741764068603,
      "num_tokens": 66573968.0,
      "step": 30390
    },
    {
      "entropy": 5.680853700637817,
      "epoch": 3.0165740373163956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004102772354354522,
      "loss": 5.0488,
      "mean_token_accuracy": 0.18228618800640106,
      "num_tokens": 66585048.0,
      "step": 30395
    },
    {
      "entropy": 5.66311936378479,
      "epoch": 3.017070265978563,
      "grad_norm": 1.390625,
      "learning_rate": 0.00041024892135688287,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18872907012701035,
      "num_tokens": 66594813.0,
      "step": 30400
    },
    {
      "entropy": 5.6771458148956295,
      "epoch": 3.0175664946407306,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004102206039244316,
      "loss": 4.9678,
      "mean_token_accuracy": 0.19244375824928284,
      "num_tokens": 66607096.0,
      "step": 30405
    },
    {
      "entropy": 5.698879814147949,
      "epoch": 3.018062723302898,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004101922831388008,
      "loss": 5.003,
      "mean_token_accuracy": 0.18064044117927552,
      "num_tokens": 66617001.0,
      "step": 30410
    },
    {
      "entropy": 5.673694753646851,
      "epoch": 3.0185589519650655,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041016395900069267,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18577697575092317,
      "num_tokens": 66627036.0,
      "step": 30415
    },
    {
      "entropy": 5.678611469268799,
      "epoch": 3.019055180627233,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004101356315108097,
      "loss": 4.9887,
      "mean_token_accuracy": 0.1820354178547859,
      "num_tokens": 66637526.0,
      "step": 30420
    },
    {
      "entropy": 5.639827060699463,
      "epoch": 3.0195514092894005,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041010730066985414,
      "loss": 4.9362,
      "mean_token_accuracy": 0.19430045038461685,
      "num_tokens": 66650046.0,
      "step": 30425
    },
    {
      "entropy": 5.611819553375244,
      "epoch": 3.020047637951568,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004100789664785287,
      "loss": 4.887,
      "mean_token_accuracy": 0.19034115374088287,
      "num_tokens": 66660922.0,
      "step": 30430
    },
    {
      "entropy": 5.6650317192077635,
      "epoch": 3.0205438666137354,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004100506289375359,
      "loss": 5.0419,
      "mean_token_accuracy": 0.1849277973175049,
      "num_tokens": 66671861.0,
      "step": 30435
    },
    {
      "entropy": 5.63644757270813,
      "epoch": 3.0210400952759033,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004100222880475786,
      "loss": 4.9765,
      "mean_token_accuracy": 0.19333745092153548,
      "num_tokens": 66682834.0,
      "step": 30440
    },
    {
      "entropy": 5.628024291992188,
      "epoch": 3.021536323938071,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040999394380935943,
      "loss": 4.9901,
      "mean_token_accuracy": 0.1865541473031044,
      "num_tokens": 66693692.0,
      "step": 30445
    },
    {
      "entropy": 5.65487904548645,
      "epoch": 3.0220325526002383,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040996559622358145,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1903396189212799,
      "num_tokens": 66705035.0,
      "step": 30450
    },
    {
      "entropy": 5.570562219619751,
      "epoch": 3.0225287812624058,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004099372452909474,
      "loss": 4.8976,
      "mean_token_accuracy": 0.1951105237007141,
      "num_tokens": 66714713.0,
      "step": 30455
    },
    {
      "entropy": 5.625882530212403,
      "epoch": 3.0230250099245732,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004099088910121605,
      "loss": 5.0273,
      "mean_token_accuracy": 0.18239627331495284,
      "num_tokens": 66726454.0,
      "step": 30460
    },
    {
      "entropy": 5.614239740371704,
      "epoch": 3.0235212385867407,
      "grad_norm": 1.28125,
      "learning_rate": 0.00040988053338792367,
      "loss": 4.9701,
      "mean_token_accuracy": 0.1845447912812233,
      "num_tokens": 66736103.0,
      "step": 30465
    },
    {
      "entropy": 5.654922103881836,
      "epoch": 3.024017467248908,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004098521724189404,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18095705807209014,
      "num_tokens": 66746554.0,
      "step": 30470
    },
    {
      "entropy": 5.673551607131958,
      "epoch": 3.0245136959110757,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040982380810591386,
      "loss": 4.9385,
      "mean_token_accuracy": 0.1909753382205963,
      "num_tokens": 66756568.0,
      "step": 30475
    },
    {
      "entropy": 5.649005126953125,
      "epoch": 3.025009924573243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004097954404495473,
      "loss": 4.9524,
      "mean_token_accuracy": 0.18764752596616746,
      "num_tokens": 66769036.0,
      "step": 30480
    },
    {
      "entropy": 5.578696393966675,
      "epoch": 3.025506153235411,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040976706945054434,
      "loss": 4.8366,
      "mean_token_accuracy": 0.21511558890342714,
      "num_tokens": 66779868.0,
      "step": 30485
    },
    {
      "entropy": 5.651692152023315,
      "epoch": 3.0260023818975785,
      "grad_norm": 1.125,
      "learning_rate": 0.00040973869510960837,
      "loss": 4.993,
      "mean_token_accuracy": 0.18413075059652328,
      "num_tokens": 66791903.0,
      "step": 30490
    },
    {
      "entropy": 5.676706695556641,
      "epoch": 3.026498610559746,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040971031742744314,
      "loss": 5.0056,
      "mean_token_accuracy": 0.18704207390546798,
      "num_tokens": 66802118.0,
      "step": 30495
    },
    {
      "entropy": 5.7117688179016115,
      "epoch": 3.0269948392219135,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040968193640475236,
      "loss": 5.0265,
      "mean_token_accuracy": 0.19050508886575698,
      "num_tokens": 66813653.0,
      "step": 30500
    },
    {
      "entropy": 5.7198601245880125,
      "epoch": 3.027491067884081,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004096535520422397,
      "loss": 5.067,
      "mean_token_accuracy": 0.18083609342575074,
      "num_tokens": 66824752.0,
      "step": 30505
    },
    {
      "entropy": 5.648956966400147,
      "epoch": 3.0279872965462484,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040962516434060915,
      "loss": 4.9338,
      "mean_token_accuracy": 0.19357501417398454,
      "num_tokens": 66835483.0,
      "step": 30510
    },
    {
      "entropy": 5.701659870147705,
      "epoch": 3.028483525208416,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004095967733005646,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18904826939105987,
      "num_tokens": 66846890.0,
      "step": 30515
    },
    {
      "entropy": 5.567217302322388,
      "epoch": 3.0289797538705834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040956837892281014,
      "loss": 4.8867,
      "mean_token_accuracy": 0.1855529397726059,
      "num_tokens": 66857326.0,
      "step": 30520
    },
    {
      "entropy": 5.6608339786529545,
      "epoch": 3.0294759825327513,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040953998120804976,
      "loss": 4.9711,
      "mean_token_accuracy": 0.18970234394073487,
      "num_tokens": 66868266.0,
      "step": 30525
    },
    {
      "entropy": 5.69858832359314,
      "epoch": 3.0299722111949188,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004095115801569878,
      "loss": 5.0148,
      "mean_token_accuracy": 0.1796177953481674,
      "num_tokens": 66879121.0,
      "step": 30530
    },
    {
      "entropy": 5.655814456939697,
      "epoch": 3.0304684398570862,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004094831757703285,
      "loss": 4.941,
      "mean_token_accuracy": 0.18872552514076232,
      "num_tokens": 66889804.0,
      "step": 30535
    },
    {
      "entropy": 5.662028360366821,
      "epoch": 3.0309646685192537,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040945476804877623,
      "loss": 4.961,
      "mean_token_accuracy": 0.1923557624220848,
      "num_tokens": 66900421.0,
      "step": 30540
    },
    {
      "entropy": 5.690329313278198,
      "epoch": 3.031460897181421,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004094263569930355,
      "loss": 5.0564,
      "mean_token_accuracy": 0.18236169517040252,
      "num_tokens": 66910754.0,
      "step": 30545
    },
    {
      "entropy": 5.728973293304444,
      "epoch": 3.0319571258435887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040939794260381074,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18220025897026063,
      "num_tokens": 66921107.0,
      "step": 30550
    },
    {
      "entropy": 5.642486572265625,
      "epoch": 3.032453354505756,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004093695248818065,
      "loss": 4.946,
      "mean_token_accuracy": 0.189946648478508,
      "num_tokens": 66932851.0,
      "step": 30555
    },
    {
      "entropy": 5.65090765953064,
      "epoch": 3.0329495831679236,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004093411038277277,
      "loss": 4.9842,
      "mean_token_accuracy": 0.18599752187728882,
      "num_tokens": 66943403.0,
      "step": 30560
    },
    {
      "entropy": 5.728872585296631,
      "epoch": 3.033445811830091,
      "grad_norm": 1.203125,
      "learning_rate": 0.000409312679442279,
      "loss": 5.1274,
      "mean_token_accuracy": 0.18296850323677064,
      "num_tokens": 66955021.0,
      "step": 30565
    },
    {
      "entropy": 5.677647876739502,
      "epoch": 3.033942040492259,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00040928425172616524,
      "loss": 5.0022,
      "mean_token_accuracy": 0.18551743626594544,
      "num_tokens": 66966147.0,
      "step": 30570
    },
    {
      "entropy": 5.687159538269043,
      "epoch": 3.0344382691544265,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040925582068009145,
      "loss": 4.9468,
      "mean_token_accuracy": 0.18653012067079544,
      "num_tokens": 66977946.0,
      "step": 30575
    },
    {
      "entropy": 5.6806943893432615,
      "epoch": 3.034934497816594,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004092273863047625,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18237378150224687,
      "num_tokens": 66988014.0,
      "step": 30580
    },
    {
      "entropy": 5.727109813690186,
      "epoch": 3.0354307264787614,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040919894860088366,
      "loss": 5.01,
      "mean_token_accuracy": 0.18664720207452773,
      "num_tokens": 66999142.0,
      "step": 30585
    },
    {
      "entropy": 5.645805549621582,
      "epoch": 3.035926955140929,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004091705075691601,
      "loss": 5.0255,
      "mean_token_accuracy": 0.18041563630104065,
      "num_tokens": 67010849.0,
      "step": 30590
    },
    {
      "entropy": 5.640702581405639,
      "epoch": 3.0364231838030964,
      "grad_norm": 1.1875,
      "learning_rate": 0.000409142063210297,
      "loss": 4.9441,
      "mean_token_accuracy": 0.19483641386032105,
      "num_tokens": 67021681.0,
      "step": 30595
    },
    {
      "entropy": 5.685677194595337,
      "epoch": 3.036919412465264,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040911361552499974,
      "loss": 5.0375,
      "mean_token_accuracy": 0.17547702193260192,
      "num_tokens": 67034080.0,
      "step": 30600
    },
    {
      "entropy": 5.67418360710144,
      "epoch": 3.0374156411274313,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040908516451397395,
      "loss": 4.9055,
      "mean_token_accuracy": 0.190007184445858,
      "num_tokens": 67045528.0,
      "step": 30605
    },
    {
      "entropy": 5.650725507736206,
      "epoch": 3.0379118697895993,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004090567101779249,
      "loss": 4.9352,
      "mean_token_accuracy": 0.19377490431070327,
      "num_tokens": 67057611.0,
      "step": 30610
    },
    {
      "entropy": 5.639294719696045,
      "epoch": 3.0384080984517667,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004090282525175583,
      "loss": 4.9629,
      "mean_token_accuracy": 0.18877752274274825,
      "num_tokens": 67067843.0,
      "step": 30615
    },
    {
      "entropy": 5.723117685317993,
      "epoch": 3.038904327113934,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004089997915335799,
      "loss": 5.0607,
      "mean_token_accuracy": 0.1873531237244606,
      "num_tokens": 67079125.0,
      "step": 30620
    },
    {
      "entropy": 5.646538591384887,
      "epoch": 3.0394005557761017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004089713272266954,
      "loss": 4.9816,
      "mean_token_accuracy": 0.18770452588796616,
      "num_tokens": 67089750.0,
      "step": 30625
    },
    {
      "entropy": 5.654922866821289,
      "epoch": 3.039896784438269,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004089428595976106,
      "loss": 5.0169,
      "mean_token_accuracy": 0.19249796122312546,
      "num_tokens": 67101738.0,
      "step": 30630
    },
    {
      "entropy": 5.655565452575684,
      "epoch": 3.0403930131004366,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004089143886470316,
      "loss": 4.9427,
      "mean_token_accuracy": 0.19060348421335221,
      "num_tokens": 67111941.0,
      "step": 30635
    },
    {
      "entropy": 5.696366882324218,
      "epoch": 3.040889241762604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004088859143756643,
      "loss": 5.0805,
      "mean_token_accuracy": 0.18049140572547911,
      "num_tokens": 67124755.0,
      "step": 30640
    },
    {
      "entropy": 5.745198154449463,
      "epoch": 3.0413854704247716,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004088574367842148,
      "loss": 5.1424,
      "mean_token_accuracy": 0.17968570739030837,
      "num_tokens": 67136597.0,
      "step": 30645
    },
    {
      "entropy": 5.700341892242432,
      "epoch": 3.041881699086939,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004088289558733893,
      "loss": 5.0378,
      "mean_token_accuracy": 0.1884353920817375,
      "num_tokens": 67146790.0,
      "step": 30650
    },
    {
      "entropy": 5.655940151214599,
      "epoch": 3.042377927749107,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004088004716438941,
      "loss": 4.9851,
      "mean_token_accuracy": 0.19330414831638337,
      "num_tokens": 67158054.0,
      "step": 30655
    },
    {
      "entropy": 5.649253034591675,
      "epoch": 3.0428741564112745,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004087719840964355,
      "loss": 4.9345,
      "mean_token_accuracy": 0.19193266779184343,
      "num_tokens": 67168244.0,
      "step": 30660
    },
    {
      "entropy": 5.630404853820801,
      "epoch": 3.043370385073442,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040874349323171994,
      "loss": 4.9645,
      "mean_token_accuracy": 0.1927070513367653,
      "num_tokens": 67179165.0,
      "step": 30665
    },
    {
      "entropy": 5.613630867004394,
      "epoch": 3.0438666137356094,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004087149990504539,
      "loss": 4.9591,
      "mean_token_accuracy": 0.1849343553185463,
      "num_tokens": 67189653.0,
      "step": 30670
    },
    {
      "entropy": 5.654103088378906,
      "epoch": 3.044362842397777,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040868650155334406,
      "loss": 5.0033,
      "mean_token_accuracy": 0.18308991491794585,
      "num_tokens": 67203933.0,
      "step": 30675
    },
    {
      "entropy": 5.62613959312439,
      "epoch": 3.0448590710599444,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040865800074109713,
      "loss": 4.94,
      "mean_token_accuracy": 0.19192473143339156,
      "num_tokens": 67215590.0,
      "step": 30680
    },
    {
      "entropy": 5.651137399673462,
      "epoch": 3.045355299722112,
      "grad_norm": 1.125,
      "learning_rate": 0.0004086294966144196,
      "loss": 4.9305,
      "mean_token_accuracy": 0.19004953652620316,
      "num_tokens": 67225416.0,
      "step": 30685
    },
    {
      "entropy": 5.647174072265625,
      "epoch": 3.0458515283842793,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004086009891740187,
      "loss": 4.9174,
      "mean_token_accuracy": 0.19034214764833451,
      "num_tokens": 67234492.0,
      "step": 30690
    },
    {
      "entropy": 5.62056074142456,
      "epoch": 3.046347757046447,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040857247842060105,
      "loss": 4.9813,
      "mean_token_accuracy": 0.18456099033355713,
      "num_tokens": 67245630.0,
      "step": 30695
    },
    {
      "entropy": 5.65878586769104,
      "epoch": 3.0468439857086147,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004085439643548737,
      "loss": 5.0519,
      "mean_token_accuracy": 0.18470731973648072,
      "num_tokens": 67257014.0,
      "step": 30700
    },
    {
      "entropy": 5.598062515258789,
      "epoch": 3.047340214370782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004085154469775439,
      "loss": 4.8819,
      "mean_token_accuracy": 0.19409338533878326,
      "num_tokens": 67267165.0,
      "step": 30705
    },
    {
      "entropy": 5.679300165176391,
      "epoch": 3.0478364430329496,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004084869262893187,
      "loss": 4.9774,
      "mean_token_accuracy": 0.19214876145124435,
      "num_tokens": 67277077.0,
      "step": 30710
    },
    {
      "entropy": 5.6506763935089115,
      "epoch": 3.048332671695117,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004084584022909053,
      "loss": 4.985,
      "mean_token_accuracy": 0.18729155957698823,
      "num_tokens": 67287941.0,
      "step": 30715
    },
    {
      "entropy": 5.51153826713562,
      "epoch": 3.0488289003572846,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040842987498301115,
      "loss": 4.8817,
      "mean_token_accuracy": 0.19190887063741685,
      "num_tokens": 67299764.0,
      "step": 30720
    },
    {
      "entropy": 5.649044990539551,
      "epoch": 3.049325129019452,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004084013443663437,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18504853546619415,
      "num_tokens": 67310527.0,
      "step": 30725
    },
    {
      "entropy": 5.686164617538452,
      "epoch": 3.0498213576816195,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004083728104416102,
      "loss": 4.9729,
      "mean_token_accuracy": 0.19134342819452285,
      "num_tokens": 67321369.0,
      "step": 30730
    },
    {
      "entropy": 5.6377113342285154,
      "epoch": 3.050317586343787,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004083442732095185,
      "loss": 4.915,
      "mean_token_accuracy": 0.19666786789894103,
      "num_tokens": 67332215.0,
      "step": 30735
    },
    {
      "entropy": 5.648122978210449,
      "epoch": 3.050813815005955,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004083157326707762,
      "loss": 5.0926,
      "mean_token_accuracy": 0.17992788255214692,
      "num_tokens": 67344407.0,
      "step": 30740
    },
    {
      "entropy": 5.657682991027832,
      "epoch": 3.0513100436681224,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040828718882609085,
      "loss": 4.9547,
      "mean_token_accuracy": 0.19166303128004075,
      "num_tokens": 67355513.0,
      "step": 30745
    },
    {
      "entropy": 5.64826512336731,
      "epoch": 3.05180627233029,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040825864167617047,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18309032320976257,
      "num_tokens": 67366515.0,
      "step": 30750
    },
    {
      "entropy": 5.662742280960083,
      "epoch": 3.0523025009924574,
      "grad_norm": 1.125,
      "learning_rate": 0.000408230091221723,
      "loss": 4.9657,
      "mean_token_accuracy": 0.19456444829702377,
      "num_tokens": 67376568.0,
      "step": 30755
    },
    {
      "entropy": 5.698182535171509,
      "epoch": 3.052798729654625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004082015374634563,
      "loss": 5.062,
      "mean_token_accuracy": 0.17575045228004454,
      "num_tokens": 67388639.0,
      "step": 30760
    },
    {
      "entropy": 5.68499436378479,
      "epoch": 3.0532949583167923,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004081729804020785,
      "loss": 4.9867,
      "mean_token_accuracy": 0.1811544418334961,
      "num_tokens": 67400330.0,
      "step": 30765
    },
    {
      "entropy": 5.6393317699432375,
      "epoch": 3.05379118697896,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004081444200382978,
      "loss": 4.9594,
      "mean_token_accuracy": 0.1915320187807083,
      "num_tokens": 67411067.0,
      "step": 30770
    },
    {
      "entropy": 5.60473666191101,
      "epoch": 3.0542874156411273,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040811585637282234,
      "loss": 4.9204,
      "mean_token_accuracy": 0.19148870259523393,
      "num_tokens": 67421142.0,
      "step": 30775
    },
    {
      "entropy": 5.60412917137146,
      "epoch": 3.0547836443032947,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004080872894063605,
      "loss": 4.9495,
      "mean_token_accuracy": 0.19226493686437607,
      "num_tokens": 67433609.0,
      "step": 30780
    },
    {
      "entropy": 5.776165199279785,
      "epoch": 3.0552798729654627,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040805871913962056,
      "loss": 5.1292,
      "mean_token_accuracy": 0.17447586506605148,
      "num_tokens": 67445281.0,
      "step": 30785
    },
    {
      "entropy": 5.62522668838501,
      "epoch": 3.05577610162763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004080301455733112,
      "loss": 4.9444,
      "mean_token_accuracy": 0.18697623014450074,
      "num_tokens": 67455831.0,
      "step": 30790
    },
    {
      "entropy": 5.682117891311646,
      "epoch": 3.0562723302897976,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040800156870814087,
      "loss": 5.0622,
      "mean_token_accuracy": 0.18766324818134308,
      "num_tokens": 67467195.0,
      "step": 30795
    },
    {
      "entropy": 5.575071859359741,
      "epoch": 3.056768558951965,
      "grad_norm": 1.125,
      "learning_rate": 0.00040797298854481826,
      "loss": 4.9239,
      "mean_token_accuracy": 0.18884676694869995,
      "num_tokens": 67478332.0,
      "step": 30800
    },
    {
      "entropy": 5.622537136077881,
      "epoch": 3.0572647876141326,
      "grad_norm": 1.421875,
      "learning_rate": 0.00040794440508405195,
      "loss": 4.9516,
      "mean_token_accuracy": 0.19840733259916304,
      "num_tokens": 67489600.0,
      "step": 30805
    },
    {
      "entropy": 5.716244125366211,
      "epoch": 3.0577610162763,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000407915818326551,
      "loss": 5.0622,
      "mean_token_accuracy": 0.1791688933968544,
      "num_tokens": 67498837.0,
      "step": 30810
    },
    {
      "entropy": 5.723680925369263,
      "epoch": 3.0582572449384675,
      "grad_norm": 1.109375,
      "learning_rate": 0.000407887228273024,
      "loss": 5.0109,
      "mean_token_accuracy": 0.19159072041511535,
      "num_tokens": 67509118.0,
      "step": 30815
    },
    {
      "entropy": 5.5719757080078125,
      "epoch": 3.058753473600635,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004078586349241802,
      "loss": 4.9351,
      "mean_token_accuracy": 0.1874098539352417,
      "num_tokens": 67520577.0,
      "step": 30820
    },
    {
      "entropy": 5.655934190750122,
      "epoch": 3.059249702262803,
      "grad_norm": 1.125,
      "learning_rate": 0.00040783003828072855,
      "loss": 4.9174,
      "mean_token_accuracy": 0.19709863364696503,
      "num_tokens": 67532020.0,
      "step": 30825
    },
    {
      "entropy": 5.626074171066284,
      "epoch": 3.0597459309249704,
      "grad_norm": 1.25,
      "learning_rate": 0.0004078014383433782,
      "loss": 4.9661,
      "mean_token_accuracy": 0.18665089458227158,
      "num_tokens": 67540845.0,
      "step": 30830
    },
    {
      "entropy": 5.607986497879028,
      "epoch": 3.060242159587138,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004077728351128383,
      "loss": 4.9791,
      "mean_token_accuracy": 0.19091403782367705,
      "num_tokens": 67552698.0,
      "step": 30835
    },
    {
      "entropy": 5.596750497817993,
      "epoch": 3.0607383882493053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004077442285898181,
      "loss": 4.9636,
      "mean_token_accuracy": 0.185807503759861,
      "num_tokens": 67563772.0,
      "step": 30840
    },
    {
      "entropy": 5.6590758800506595,
      "epoch": 3.061234616911473,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040771561877502727,
      "loss": 4.984,
      "mean_token_accuracy": 0.1893924966454506,
      "num_tokens": 67574955.0,
      "step": 30845
    },
    {
      "entropy": 5.642880916595459,
      "epoch": 3.0617308455736403,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004076870056691749,
      "loss": 4.9342,
      "mean_token_accuracy": 0.19338614493608475,
      "num_tokens": 67586086.0,
      "step": 30850
    },
    {
      "entropy": 5.716059398651123,
      "epoch": 3.0622270742358078,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040765838927297077,
      "loss": 5.0251,
      "mean_token_accuracy": 0.1862015336751938,
      "num_tokens": 67597036.0,
      "step": 30855
    },
    {
      "entropy": 5.676837301254272,
      "epoch": 3.0627233028979752,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004076297695871244,
      "loss": 4.8993,
      "mean_token_accuracy": 0.20651559233665467,
      "num_tokens": 67606499.0,
      "step": 30860
    },
    {
      "entropy": 5.677365732192993,
      "epoch": 3.0632195315601427,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040760114661234553,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18432625830173494,
      "num_tokens": 67617820.0,
      "step": 30865
    },
    {
      "entropy": 5.624530553817749,
      "epoch": 3.0637157602223106,
      "grad_norm": 1.15625,
      "learning_rate": 0.000407572520349344,
      "loss": 4.9095,
      "mean_token_accuracy": 0.1969657152891159,
      "num_tokens": 67628122.0,
      "step": 30870
    },
    {
      "entropy": 5.608409976959228,
      "epoch": 3.064211988884478,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004075438907988296,
      "loss": 4.9946,
      "mean_token_accuracy": 0.18866120129823685,
      "num_tokens": 67639127.0,
      "step": 30875
    },
    {
      "entropy": 5.620789241790772,
      "epoch": 3.0647082175466456,
      "grad_norm": 1.125,
      "learning_rate": 0.0004075152579615124,
      "loss": 4.9029,
      "mean_token_accuracy": 0.1946312665939331,
      "num_tokens": 67650122.0,
      "step": 30880
    },
    {
      "entropy": 5.684666633605957,
      "epoch": 3.065204446208813,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004074866218381023,
      "loss": 5.0296,
      "mean_token_accuracy": 0.1829001009464264,
      "num_tokens": 67660427.0,
      "step": 30885
    },
    {
      "entropy": 5.628302526473999,
      "epoch": 3.0657006748709805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040745798242930947,
      "loss": 5.0659,
      "mean_token_accuracy": 0.18897947371006013,
      "num_tokens": 67671151.0,
      "step": 30890
    },
    {
      "entropy": 5.678963136672974,
      "epoch": 3.066196903533148,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000407429339735844,
      "loss": 5.021,
      "mean_token_accuracy": 0.18188706934452056,
      "num_tokens": 67681338.0,
      "step": 30895
    },
    {
      "entropy": 5.64892611503601,
      "epoch": 3.0666931321953155,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040740069375841637,
      "loss": 4.9282,
      "mean_token_accuracy": 0.18925302475690842,
      "num_tokens": 67691884.0,
      "step": 30900
    },
    {
      "entropy": 5.628232717514038,
      "epoch": 3.067189360857483,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004073720444977368,
      "loss": 4.9911,
      "mean_token_accuracy": 0.1915970414876938,
      "num_tokens": 67703011.0,
      "step": 30905
    },
    {
      "entropy": 5.746575546264649,
      "epoch": 3.067685589519651,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004073433919545157,
      "loss": 5.1554,
      "mean_token_accuracy": 0.177664415538311,
      "num_tokens": 67714042.0,
      "step": 30910
    },
    {
      "entropy": 5.670121955871582,
      "epoch": 3.0681818181818183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004073147361294638,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18773265033960343,
      "num_tokens": 67725449.0,
      "step": 30915
    },
    {
      "entropy": 5.683173513412475,
      "epoch": 3.068678046843986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004072860770232915,
      "loss": 5.076,
      "mean_token_accuracy": 0.18235631585121154,
      "num_tokens": 67737334.0,
      "step": 30920
    },
    {
      "entropy": 5.617729520797729,
      "epoch": 3.0691742755061533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004072574146367095,
      "loss": 4.9665,
      "mean_token_accuracy": 0.18884899467229843,
      "num_tokens": 67747777.0,
      "step": 30925
    },
    {
      "entropy": 5.637265539169311,
      "epoch": 3.0696705041683208,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040722874897042863,
      "loss": 4.9351,
      "mean_token_accuracy": 0.19624906927347183,
      "num_tokens": 67758584.0,
      "step": 30930
    },
    {
      "entropy": 5.624196624755859,
      "epoch": 3.0701667328304882,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004072000800251597,
      "loss": 4.9669,
      "mean_token_accuracy": 0.1915315717458725,
      "num_tokens": 67769168.0,
      "step": 30935
    },
    {
      "entropy": 5.6329797267913815,
      "epoch": 3.0706629614926557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040717140780161367,
      "loss": 4.981,
      "mean_token_accuracy": 0.18874119967222214,
      "num_tokens": 67779680.0,
      "step": 30940
    },
    {
      "entropy": 5.631940126419067,
      "epoch": 3.071159190154823,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040714273230050153,
      "loss": 4.9921,
      "mean_token_accuracy": 0.18251799643039704,
      "num_tokens": 67789678.0,
      "step": 30945
    },
    {
      "entropy": 5.637035179138183,
      "epoch": 3.0716554188169907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004071140535225343,
      "loss": 5.0178,
      "mean_token_accuracy": 0.19091658443212509,
      "num_tokens": 67800991.0,
      "step": 30950
    },
    {
      "entropy": 5.734177207946777,
      "epoch": 3.0721516474791586,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004070853714684233,
      "loss": 5.1086,
      "mean_token_accuracy": 0.18209696859121322,
      "num_tokens": 67811319.0,
      "step": 30955
    },
    {
      "entropy": 5.740793752670288,
      "epoch": 3.072647876141326,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004070566861388797,
      "loss": 5.0989,
      "mean_token_accuracy": 0.18144431561231614,
      "num_tokens": 67823704.0,
      "step": 30960
    },
    {
      "entropy": 5.68080325126648,
      "epoch": 3.0731441048034935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040702799753461477,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18551439344882964,
      "num_tokens": 67835460.0,
      "step": 30965
    },
    {
      "entropy": 5.628761339187622,
      "epoch": 3.073640333465661,
      "grad_norm": 1.125,
      "learning_rate": 0.0004069993056563401,
      "loss": 4.9402,
      "mean_token_accuracy": 0.19642819762229918,
      "num_tokens": 67845464.0,
      "step": 30970
    },
    {
      "entropy": 5.749191951751709,
      "epoch": 3.0741365621278285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000406970610504767,
      "loss": 5.0664,
      "mean_token_accuracy": 0.17735726684331893,
      "num_tokens": 67856718.0,
      "step": 30975
    },
    {
      "entropy": 5.6805401802062985,
      "epoch": 3.074632790789996,
      "grad_norm": 1.125,
      "learning_rate": 0.0004069419120806072,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1887628898024559,
      "num_tokens": 67867066.0,
      "step": 30980
    },
    {
      "entropy": 5.648596906661988,
      "epoch": 3.0751290194521634,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040691321038457224,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18408591002225877,
      "num_tokens": 67878213.0,
      "step": 30985
    },
    {
      "entropy": 5.644204711914062,
      "epoch": 3.075625248114331,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004068845054173739,
      "loss": 5.0217,
      "mean_token_accuracy": 0.1836567535996437,
      "num_tokens": 67888503.0,
      "step": 30990
    },
    {
      "entropy": 5.712199783325195,
      "epoch": 3.076121476776499,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004068557971797241,
      "loss": 5.0098,
      "mean_token_accuracy": 0.18296639770269393,
      "num_tokens": 67899035.0,
      "step": 30995
    },
    {
      "entropy": 5.612288188934326,
      "epoch": 3.0766177054386663,
      "grad_norm": 1.125,
      "learning_rate": 0.00040682708567233464,
      "loss": 4.9833,
      "mean_token_accuracy": 0.18978845626115798,
      "num_tokens": 67909137.0,
      "step": 31000
    },
    {
      "entropy": 5.6582191467285154,
      "epoch": 3.077113934100834,
      "grad_norm": 1.21875,
      "learning_rate": 0.00040679837089591743,
      "loss": 5.0004,
      "mean_token_accuracy": 0.18303195834159852,
      "num_tokens": 67919026.0,
      "step": 31005
    },
    {
      "entropy": 5.732039165496826,
      "epoch": 3.0776101627630013,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040676965285118465,
      "loss": 5.0676,
      "mean_token_accuracy": 0.18603979349136351,
      "num_tokens": 67930333.0,
      "step": 31010
    },
    {
      "entropy": 5.661561727523804,
      "epoch": 3.0781063914251687,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004067409315388484,
      "loss": 4.9376,
      "mean_token_accuracy": 0.19197780787944793,
      "num_tokens": 67940448.0,
      "step": 31015
    },
    {
      "entropy": 5.6462915420532225,
      "epoch": 3.078602620087336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040671220695962104,
      "loss": 4.9849,
      "mean_token_accuracy": 0.18918138295412062,
      "num_tokens": 67951398.0,
      "step": 31020
    },
    {
      "entropy": 5.691695213317871,
      "epoch": 3.0790988487495037,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004066834791142146,
      "loss": 4.9936,
      "mean_token_accuracy": 0.1874009355902672,
      "num_tokens": 67962741.0,
      "step": 31025
    },
    {
      "entropy": 5.657364702224731,
      "epoch": 3.079595077411671,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004066547480033417,
      "loss": 4.9706,
      "mean_token_accuracy": 0.1929507374763489,
      "num_tokens": 67975015.0,
      "step": 31030
    },
    {
      "entropy": 5.719782924652099,
      "epoch": 3.0800913060738386,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004066260136277148,
      "loss": 5.047,
      "mean_token_accuracy": 0.1826100915670395,
      "num_tokens": 67985918.0,
      "step": 31035
    },
    {
      "entropy": 5.653482294082641,
      "epoch": 3.0805875347360065,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040659727598804635,
      "loss": 5.0149,
      "mean_token_accuracy": 0.18587248474359513,
      "num_tokens": 67997419.0,
      "step": 31040
    },
    {
      "entropy": 5.639696550369263,
      "epoch": 3.081083763398174,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040656853508504897,
      "loss": 4.9545,
      "mean_token_accuracy": 0.18833935856819153,
      "num_tokens": 68008739.0,
      "step": 31045
    },
    {
      "entropy": 5.648548030853272,
      "epoch": 3.0815799920603415,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040653979091943555,
      "loss": 5.0169,
      "mean_token_accuracy": 0.1842658445239067,
      "num_tokens": 68020707.0,
      "step": 31050
    },
    {
      "entropy": 5.715845394134521,
      "epoch": 3.082076220722509,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040651104349191864,
      "loss": 4.9377,
      "mean_token_accuracy": 0.19135227799415588,
      "num_tokens": 68030825.0,
      "step": 31055
    },
    {
      "entropy": 5.649228525161743,
      "epoch": 3.0825724493846764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040648229280321126,
      "loss": 4.9906,
      "mean_token_accuracy": 0.18837181031703948,
      "num_tokens": 68041385.0,
      "step": 31060
    },
    {
      "entropy": 5.677231025695801,
      "epoch": 3.083068678046844,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004064535388540264,
      "loss": 5.0747,
      "mean_token_accuracy": 0.1840759336948395,
      "num_tokens": 68053123.0,
      "step": 31065
    },
    {
      "entropy": 5.631771373748779,
      "epoch": 3.0835649067090114,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004064247816450769,
      "loss": 4.9691,
      "mean_token_accuracy": 0.18523749560117722,
      "num_tokens": 68064529.0,
      "step": 31070
    },
    {
      "entropy": 5.756842231750488,
      "epoch": 3.084061135371179,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004063960211770761,
      "loss": 5.0719,
      "mean_token_accuracy": 0.18582841753959656,
      "num_tokens": 68074091.0,
      "step": 31075
    },
    {
      "entropy": 5.661604118347168,
      "epoch": 3.084557364033347,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040636725745073714,
      "loss": 4.9765,
      "mean_token_accuracy": 0.19099530577659607,
      "num_tokens": 68085185.0,
      "step": 31080
    },
    {
      "entropy": 5.606393814086914,
      "epoch": 3.0850535926955143,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004063384904667732,
      "loss": 4.9092,
      "mean_token_accuracy": 0.19470038563013076,
      "num_tokens": 68095977.0,
      "step": 31085
    },
    {
      "entropy": 5.6523267269134525,
      "epoch": 3.0855498213576817,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00040630972022589774,
      "loss": 5.0054,
      "mean_token_accuracy": 0.18737533837556838,
      "num_tokens": 68107469.0,
      "step": 31090
    },
    {
      "entropy": 5.7805016040802,
      "epoch": 3.086046050019849,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040628094672882415,
      "loss": 5.1582,
      "mean_token_accuracy": 0.17048311233520508,
      "num_tokens": 68119153.0,
      "step": 31095
    },
    {
      "entropy": 5.666574192047119,
      "epoch": 3.0865422786820167,
      "grad_norm": 1.078125,
      "learning_rate": 0.000406252169976266,
      "loss": 4.9361,
      "mean_token_accuracy": 0.1887947842478752,
      "num_tokens": 68130336.0,
      "step": 31100
    },
    {
      "entropy": 5.70192813873291,
      "epoch": 3.087038507344184,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004062233899689368,
      "loss": 5.0013,
      "mean_token_accuracy": 0.18621270209550858,
      "num_tokens": 68141231.0,
      "step": 31105
    },
    {
      "entropy": 5.640956068038941,
      "epoch": 3.0875347360063516,
      "grad_norm": 1.375,
      "learning_rate": 0.0004061946067075503,
      "loss": 4.974,
      "mean_token_accuracy": 0.19510236233472825,
      "num_tokens": 68151891.0,
      "step": 31110
    },
    {
      "entropy": 5.656244373321533,
      "epoch": 3.088030964668519,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004061658201928203,
      "loss": 4.9579,
      "mean_token_accuracy": 0.18933019191026687,
      "num_tokens": 68162537.0,
      "step": 31115
    },
    {
      "entropy": 5.718591260910034,
      "epoch": 3.0885271933306866,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004061370304254606,
      "loss": 5.1515,
      "mean_token_accuracy": 0.18129466027021407,
      "num_tokens": 68173393.0,
      "step": 31120
    },
    {
      "entropy": 5.688375568389892,
      "epoch": 3.0890234219928545,
      "grad_norm": 1.140625,
      "learning_rate": 0.000406108237406185,
      "loss": 5.0326,
      "mean_token_accuracy": 0.1903962656855583,
      "num_tokens": 68184937.0,
      "step": 31125
    },
    {
      "entropy": 5.6625996112823485,
      "epoch": 3.089519650655022,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040607944113570777,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18870149701833724,
      "num_tokens": 68196962.0,
      "step": 31130
    },
    {
      "entropy": 5.737184715270996,
      "epoch": 3.0900158793171895,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004060506416147428,
      "loss": 5.0324,
      "mean_token_accuracy": 0.18622016310691833,
      "num_tokens": 68208603.0,
      "step": 31135
    },
    {
      "entropy": 5.640028285980224,
      "epoch": 3.090512107979357,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004060218388440043,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19323102831840516,
      "num_tokens": 68219424.0,
      "step": 31140
    },
    {
      "entropy": 5.616947889328003,
      "epoch": 3.0910083366415244,
      "grad_norm": 1.25,
      "learning_rate": 0.0004059930328242065,
      "loss": 4.942,
      "mean_token_accuracy": 0.1951662302017212,
      "num_tokens": 68229758.0,
      "step": 31145
    },
    {
      "entropy": 5.581995296478271,
      "epoch": 3.091504565303692,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040596422355606383,
      "loss": 4.9289,
      "mean_token_accuracy": 0.19312549829483033,
      "num_tokens": 68239146.0,
      "step": 31150
    },
    {
      "entropy": 5.654583930969238,
      "epoch": 3.0920007939658594,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004059354110402905,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18426240086555482,
      "num_tokens": 68249339.0,
      "step": 31155
    },
    {
      "entropy": 5.657814025878906,
      "epoch": 3.092497022628027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040590659527760123,
      "loss": 4.9785,
      "mean_token_accuracy": 0.1907879024744034,
      "num_tokens": 68260136.0,
      "step": 31160
    },
    {
      "entropy": 5.666217374801636,
      "epoch": 3.0929932512901943,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040587777626871046,
      "loss": 5.0074,
      "mean_token_accuracy": 0.1832693949341774,
      "num_tokens": 68270033.0,
      "step": 31165
    },
    {
      "entropy": 5.699298620223999,
      "epoch": 3.0934894799523622,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004058489540143328,
      "loss": 5.0457,
      "mean_token_accuracy": 0.18229165375232698,
      "num_tokens": 68283344.0,
      "step": 31170
    },
    {
      "entropy": 5.677957534790039,
      "epoch": 3.0939857086145297,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004058201285151831,
      "loss": 4.959,
      "mean_token_accuracy": 0.18833765834569932,
      "num_tokens": 68294604.0,
      "step": 31175
    },
    {
      "entropy": 5.676319885253906,
      "epoch": 3.094481937276697,
      "grad_norm": 1.25,
      "learning_rate": 0.0004057912997719761,
      "loss": 4.9485,
      "mean_token_accuracy": 0.19758192598819732,
      "num_tokens": 68304436.0,
      "step": 31180
    },
    {
      "entropy": 5.602096796035767,
      "epoch": 3.0949781659388647,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004057624677854266,
      "loss": 4.9223,
      "mean_token_accuracy": 0.1908154159784317,
      "num_tokens": 68315593.0,
      "step": 31185
    },
    {
      "entropy": 5.650812864303589,
      "epoch": 3.095474394601032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040573363255624976,
      "loss": 5.0075,
      "mean_token_accuracy": 0.19112136960029602,
      "num_tokens": 68327540.0,
      "step": 31190
    },
    {
      "entropy": 5.774739980697632,
      "epoch": 3.0959706232631996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004057047940851605,
      "loss": 5.0805,
      "mean_token_accuracy": 0.18404752910137176,
      "num_tokens": 68339784.0,
      "step": 31195
    },
    {
      "entropy": 5.671247386932373,
      "epoch": 3.096466851925367,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004056759523728739,
      "loss": 4.9907,
      "mean_token_accuracy": 0.18371029794216157,
      "num_tokens": 68351182.0,
      "step": 31200
    },
    {
      "entropy": 5.68141736984253,
      "epoch": 3.0969630805875346,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040564710742010545,
      "loss": 5.0359,
      "mean_token_accuracy": 0.18079383969306945,
      "num_tokens": 68361726.0,
      "step": 31205
    },
    {
      "entropy": 5.667634201049805,
      "epoch": 3.0974593092497025,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004056182592275702,
      "loss": 5.0197,
      "mean_token_accuracy": 0.18558653593063354,
      "num_tokens": 68373167.0,
      "step": 31210
    },
    {
      "entropy": 5.716386556625366,
      "epoch": 3.09795553791187,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004055894077959835,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18268266022205354,
      "num_tokens": 68384646.0,
      "step": 31215
    },
    {
      "entropy": 5.62832612991333,
      "epoch": 3.0984517665740374,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004055605531260609,
      "loss": 5.0045,
      "mean_token_accuracy": 0.1872977703809738,
      "num_tokens": 68395953.0,
      "step": 31220
    },
    {
      "entropy": 5.70998249053955,
      "epoch": 3.098947995236205,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040553169521851794,
      "loss": 5.0379,
      "mean_token_accuracy": 0.1875109627842903,
      "num_tokens": 68407167.0,
      "step": 31225
    },
    {
      "entropy": 5.683544683456421,
      "epoch": 3.0994442238983724,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040550283407407025,
      "loss": 5.0537,
      "mean_token_accuracy": 0.18453404754400254,
      "num_tokens": 68419530.0,
      "step": 31230
    },
    {
      "entropy": 5.69152684211731,
      "epoch": 3.09994045256054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004054739696934334,
      "loss": 5.0164,
      "mean_token_accuracy": 0.18083010166883468,
      "num_tokens": 68430970.0,
      "step": 31235
    },
    {
      "entropy": 5.642843914031983,
      "epoch": 3.1004366812227073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040544510207732337,
      "loss": 4.9446,
      "mean_token_accuracy": 0.19591667652130126,
      "num_tokens": 68441257.0,
      "step": 31240
    },
    {
      "entropy": 5.596414995193482,
      "epoch": 3.100932909884875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040541623122645576,
      "loss": 4.9794,
      "mean_token_accuracy": 0.18694092333316803,
      "num_tokens": 68452493.0,
      "step": 31245
    },
    {
      "entropy": 5.604236221313476,
      "epoch": 3.1014291385470423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040538735714154675,
      "loss": 4.9273,
      "mean_token_accuracy": 0.20247305035591126,
      "num_tokens": 68464023.0,
      "step": 31250
    },
    {
      "entropy": 5.679601335525513,
      "epoch": 3.10192536720921,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004053584798233122,
      "loss": 4.9982,
      "mean_token_accuracy": 0.19445858746767045,
      "num_tokens": 68474519.0,
      "step": 31255
    },
    {
      "entropy": 5.702237129211426,
      "epoch": 3.1024215958713777,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040532959927246814,
      "loss": 4.997,
      "mean_token_accuracy": 0.19143179059028625,
      "num_tokens": 68486239.0,
      "step": 31260
    },
    {
      "entropy": 5.7017566680908205,
      "epoch": 3.102917824533545,
      "grad_norm": 1.140625,
      "learning_rate": 0.000405300715489731,
      "loss": 5.0456,
      "mean_token_accuracy": 0.18209685385227203,
      "num_tokens": 68495921.0,
      "step": 31265
    },
    {
      "entropy": 5.692885065078736,
      "epoch": 3.1034140531957126,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004052718284758169,
      "loss": 4.9555,
      "mean_token_accuracy": 0.18921525180339813,
      "num_tokens": 68505773.0,
      "step": 31270
    },
    {
      "entropy": 5.7495733261108395,
      "epoch": 3.10391028185788,
      "grad_norm": 1.15625,
      "learning_rate": 0.000405242938231442,
      "loss": 5.0733,
      "mean_token_accuracy": 0.18054611086845399,
      "num_tokens": 68516563.0,
      "step": 31275
    },
    {
      "entropy": 5.68655743598938,
      "epoch": 3.1044065105200476,
      "grad_norm": 1.140625,
      "learning_rate": 0.000405214044757323,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19363916367292405,
      "num_tokens": 68527952.0,
      "step": 31280
    },
    {
      "entropy": 5.671428632736206,
      "epoch": 3.104902739182215,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004051851480541763,
      "loss": 5.004,
      "mean_token_accuracy": 0.18341573923826218,
      "num_tokens": 68538450.0,
      "step": 31285
    },
    {
      "entropy": 5.614675092697143,
      "epoch": 3.1053989678443825,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004051562481227184,
      "loss": 4.9648,
      "mean_token_accuracy": 0.18965734988451005,
      "num_tokens": 68547941.0,
      "step": 31290
    },
    {
      "entropy": 5.640803289413452,
      "epoch": 3.10589519650655,
      "grad_norm": 1.15625,
      "learning_rate": 0.000405127344963666,
      "loss": 4.9493,
      "mean_token_accuracy": 0.19166860580444336,
      "num_tokens": 68558525.0,
      "step": 31295
    },
    {
      "entropy": 5.616433238983154,
      "epoch": 3.106391425168718,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004050984385777359,
      "loss": 4.9508,
      "mean_token_accuracy": 0.18338238298892975,
      "num_tokens": 68569487.0,
      "step": 31300
    },
    {
      "entropy": 5.653242826461792,
      "epoch": 3.1068876538308854,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004050695289656448,
      "loss": 5.0029,
      "mean_token_accuracy": 0.18735502958297728,
      "num_tokens": 68580069.0,
      "step": 31305
    },
    {
      "entropy": 5.660795593261719,
      "epoch": 3.107383882493053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040504061612810964,
      "loss": 4.9552,
      "mean_token_accuracy": 0.19046802818775177,
      "num_tokens": 68591511.0,
      "step": 31310
    },
    {
      "entropy": 5.554008293151855,
      "epoch": 3.1078801111552203,
      "grad_norm": 1.25,
      "learning_rate": 0.0004050117000658475,
      "loss": 4.9939,
      "mean_token_accuracy": 0.18752927780151368,
      "num_tokens": 68602464.0,
      "step": 31315
    },
    {
      "entropy": 5.649525213241577,
      "epoch": 3.108376339817388,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040498278077957534,
      "loss": 4.9565,
      "mean_token_accuracy": 0.18925299048423766,
      "num_tokens": 68613492.0,
      "step": 31320
    },
    {
      "entropy": 5.648590040206909,
      "epoch": 3.1088725684795553,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004049538582700102,
      "loss": 4.9248,
      "mean_token_accuracy": 0.19526780843734742,
      "num_tokens": 68625259.0,
      "step": 31325
    },
    {
      "entropy": 5.700690746307373,
      "epoch": 3.1093687971417228,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004049249325378695,
      "loss": 5.0681,
      "mean_token_accuracy": 0.17897527664899826,
      "num_tokens": 68635455.0,
      "step": 31330
    },
    {
      "entropy": 5.662363719940186,
      "epoch": 3.1098650258038902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040489600358387036,
      "loss": 5.0585,
      "mean_token_accuracy": 0.18422080874443053,
      "num_tokens": 68648102.0,
      "step": 31335
    },
    {
      "entropy": 5.698928928375244,
      "epoch": 3.110361254466058,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004048670714087303,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19174565821886064,
      "num_tokens": 68660854.0,
      "step": 31340
    },
    {
      "entropy": 5.571126461029053,
      "epoch": 3.1108574831282256,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004048381360131667,
      "loss": 4.8555,
      "mean_token_accuracy": 0.2023005872964859,
      "num_tokens": 68671839.0,
      "step": 31345
    },
    {
      "entropy": 5.709979248046875,
      "epoch": 3.111353711790393,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004048091973978971,
      "loss": 5.0151,
      "mean_token_accuracy": 0.18824114203453063,
      "num_tokens": 68682535.0,
      "step": 31350
    },
    {
      "entropy": 5.64224681854248,
      "epoch": 3.1118499404525606,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004047802555636391,
      "loss": 4.972,
      "mean_token_accuracy": 0.18816229999065398,
      "num_tokens": 68694283.0,
      "step": 31355
    },
    {
      "entropy": 5.649995946884156,
      "epoch": 3.112346169114728,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004047513105111104,
      "loss": 4.9895,
      "mean_token_accuracy": 0.1871285170316696,
      "num_tokens": 68704268.0,
      "step": 31360
    },
    {
      "entropy": 5.633577728271485,
      "epoch": 3.1128423977768955,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040472236224102883,
      "loss": 4.9189,
      "mean_token_accuracy": 0.1849453866481781,
      "num_tokens": 68714762.0,
      "step": 31365
    },
    {
      "entropy": 5.619634771347046,
      "epoch": 3.113338626439063,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040469341075411217,
      "loss": 4.9328,
      "mean_token_accuracy": 0.18592132478952408,
      "num_tokens": 68725302.0,
      "step": 31370
    },
    {
      "entropy": 5.679748630523681,
      "epoch": 3.1138348551012305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004046644560510783,
      "loss": 5.0502,
      "mean_token_accuracy": 0.1781301349401474,
      "num_tokens": 68736893.0,
      "step": 31375
    },
    {
      "entropy": 5.644285583496094,
      "epoch": 3.114331083763398,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040463549813264557,
      "loss": 4.9642,
      "mean_token_accuracy": 0.1907876744866371,
      "num_tokens": 68748202.0,
      "step": 31380
    },
    {
      "entropy": 5.690695142745971,
      "epoch": 3.114827312425566,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040460653699953166,
      "loss": 4.9629,
      "mean_token_accuracy": 0.1962425634264946,
      "num_tokens": 68759033.0,
      "step": 31385
    },
    {
      "entropy": 5.6603381633758545,
      "epoch": 3.1153235410877333,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040457757265245485,
      "loss": 4.9562,
      "mean_token_accuracy": 0.18949244618415834,
      "num_tokens": 68770879.0,
      "step": 31390
    },
    {
      "entropy": 5.6289506435394285,
      "epoch": 3.115819769749901,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004045486050921336,
      "loss": 4.9406,
      "mean_token_accuracy": 0.1929953634738922,
      "num_tokens": 68782087.0,
      "step": 31395
    },
    {
      "entropy": 5.637213611602784,
      "epoch": 3.1163159984120683,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000404519634319286,
      "loss": 4.9608,
      "mean_token_accuracy": 0.19232148677110672,
      "num_tokens": 68792355.0,
      "step": 31400
    },
    {
      "entropy": 5.637286281585693,
      "epoch": 3.1168122270742358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040449066033463053,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18210525959730148,
      "num_tokens": 68804255.0,
      "step": 31405
    },
    {
      "entropy": 5.648334169387818,
      "epoch": 3.1173084557364032,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004044616831388858,
      "loss": 4.9571,
      "mean_token_accuracy": 0.19373418092727662,
      "num_tokens": 68816102.0,
      "step": 31410
    },
    {
      "entropy": 5.675658178329468,
      "epoch": 3.1178046843985707,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004044327027327703,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1855076402425766,
      "num_tokens": 68827303.0,
      "step": 31415
    },
    {
      "entropy": 5.658487176895141,
      "epoch": 3.118300913060738,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004044037191170026,
      "loss": 4.9777,
      "mean_token_accuracy": 0.1885417655110359,
      "num_tokens": 68838418.0,
      "step": 31420
    },
    {
      "entropy": 5.72223482131958,
      "epoch": 3.118797141722906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040437473229230155,
      "loss": 5.0173,
      "mean_token_accuracy": 0.18468344509601592,
      "num_tokens": 68848431.0,
      "step": 31425
    },
    {
      "entropy": 5.661756992340088,
      "epoch": 3.1192933703850736,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004043457422593859,
      "loss": 4.925,
      "mean_token_accuracy": 0.18679533153772354,
      "num_tokens": 68858983.0,
      "step": 31430
    },
    {
      "entropy": 5.650056076049805,
      "epoch": 3.119789599047241,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004043167490189746,
      "loss": 4.9903,
      "mean_token_accuracy": 0.19247959703207015,
      "num_tokens": 68870170.0,
      "step": 31435
    },
    {
      "entropy": 5.632922744750976,
      "epoch": 3.1202858277094085,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040428775257178655,
      "loss": 5.019,
      "mean_token_accuracy": 0.1952331468462944,
      "num_tokens": 68880242.0,
      "step": 31440
    },
    {
      "entropy": 5.707369613647461,
      "epoch": 3.120782056371576,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040425875291854083,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17780957967042924,
      "num_tokens": 68892150.0,
      "step": 31445
    },
    {
      "entropy": 5.755414962768555,
      "epoch": 3.1212782850337435,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040422975005995656,
      "loss": 5.0692,
      "mean_token_accuracy": 0.17868634462356567,
      "num_tokens": 68903164.0,
      "step": 31450
    },
    {
      "entropy": 5.653294944763184,
      "epoch": 3.121774513695911,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004042007439967529,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18473294526338577,
      "num_tokens": 68914514.0,
      "step": 31455
    },
    {
      "entropy": 5.650970411300659,
      "epoch": 3.1222707423580784,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004041717347296493,
      "loss": 4.9816,
      "mean_token_accuracy": 0.18884287923574447,
      "num_tokens": 68925366.0,
      "step": 31460
    },
    {
      "entropy": 5.637800168991089,
      "epoch": 3.122766971020246,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004041427222593649,
      "loss": 4.9293,
      "mean_token_accuracy": 0.18909592628479005,
      "num_tokens": 68936019.0,
      "step": 31465
    },
    {
      "entropy": 5.557557106018066,
      "epoch": 3.123263199682414,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040411370658661935,
      "loss": 4.8959,
      "mean_token_accuracy": 0.19620292335748674,
      "num_tokens": 68947405.0,
      "step": 31470
    },
    {
      "entropy": 5.643517780303955,
      "epoch": 3.1237594283445813,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040408468771213206,
      "loss": 4.9847,
      "mean_token_accuracy": 0.1890711233019829,
      "num_tokens": 68958822.0,
      "step": 31475
    },
    {
      "entropy": 5.677580070495606,
      "epoch": 3.124255657006749,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040405566563662265,
      "loss": 5.0155,
      "mean_token_accuracy": 0.1871380925178528,
      "num_tokens": 68969992.0,
      "step": 31480
    },
    {
      "entropy": 5.640973520278931,
      "epoch": 3.1247518856689163,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040402664036081086,
      "loss": 4.9614,
      "mean_token_accuracy": 0.188765087723732,
      "num_tokens": 68982081.0,
      "step": 31485
    },
    {
      "entropy": 5.688490009307861,
      "epoch": 3.1252481143310837,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004039976118854164,
      "loss": 5.0093,
      "mean_token_accuracy": 0.18459469676017762,
      "num_tokens": 68993430.0,
      "step": 31490
    },
    {
      "entropy": 5.735746431350708,
      "epoch": 3.125744342993251,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004039685802111591,
      "loss": 5.0762,
      "mean_token_accuracy": 0.18182710707187652,
      "num_tokens": 69004939.0,
      "step": 31495
    },
    {
      "entropy": 5.628640460968017,
      "epoch": 3.1262405716554187,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040393954533875904,
      "loss": 4.9861,
      "mean_token_accuracy": 0.1880466967821121,
      "num_tokens": 69016293.0,
      "step": 31500
    },
    {
      "entropy": 5.713147926330566,
      "epoch": 3.126736800317586,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040391050726893595,
      "loss": 5.0828,
      "mean_token_accuracy": 0.1827217623591423,
      "num_tokens": 69029087.0,
      "step": 31505
    },
    {
      "entropy": 5.6774121761322025,
      "epoch": 3.127233028979754,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004038814660024101,
      "loss": 4.99,
      "mean_token_accuracy": 0.18386918753385545,
      "num_tokens": 69040937.0,
      "step": 31510
    },
    {
      "entropy": 5.697520208358765,
      "epoch": 3.1277292576419216,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004038524215399016,
      "loss": 5.0705,
      "mean_token_accuracy": 0.1840050533413887,
      "num_tokens": 69052212.0,
      "step": 31515
    },
    {
      "entropy": 5.708924913406372,
      "epoch": 3.128225486304089,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004038233738821308,
      "loss": 4.9889,
      "mean_token_accuracy": 0.19216507375240327,
      "num_tokens": 69061951.0,
      "step": 31520
    },
    {
      "entropy": 5.702224445343018,
      "epoch": 3.1287217149662565,
      "grad_norm": 1.125,
      "learning_rate": 0.0004037943230298178,
      "loss": 4.9293,
      "mean_token_accuracy": 0.20211723297834397,
      "num_tokens": 69072156.0,
      "step": 31525
    },
    {
      "entropy": 5.656062173843384,
      "epoch": 3.129217943628424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040376526898368314,
      "loss": 4.986,
      "mean_token_accuracy": 0.18745361417531967,
      "num_tokens": 69084032.0,
      "step": 31530
    },
    {
      "entropy": 5.67871618270874,
      "epoch": 3.1297141722905915,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040373621174444736,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18423738330602646,
      "num_tokens": 69096436.0,
      "step": 31535
    },
    {
      "entropy": 5.655202579498291,
      "epoch": 3.130210400952759,
      "grad_norm": 1.1875,
      "learning_rate": 0.00040370715131283076,
      "loss": 4.9373,
      "mean_token_accuracy": 0.1928894951939583,
      "num_tokens": 69107090.0,
      "step": 31540
    },
    {
      "entropy": 5.695382356643677,
      "epoch": 3.1307066296149264,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004036780876895543,
      "loss": 4.9133,
      "mean_token_accuracy": 0.1975613936781883,
      "num_tokens": 69117644.0,
      "step": 31545
    },
    {
      "entropy": 5.623253011703492,
      "epoch": 3.131202858277094,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004036490208753385,
      "loss": 4.9789,
      "mean_token_accuracy": 0.18527060002088547,
      "num_tokens": 69129639.0,
      "step": 31550
    },
    {
      "entropy": 5.575646018981933,
      "epoch": 3.131699086939262,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040361995087090407,
      "loss": 4.9282,
      "mean_token_accuracy": 0.19076600521802903,
      "num_tokens": 69139429.0,
      "step": 31555
    },
    {
      "entropy": 5.548508882522583,
      "epoch": 3.1321953156014293,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040359087767697223,
      "loss": 4.8912,
      "mean_token_accuracy": 0.19981607496738435,
      "num_tokens": 69150777.0,
      "step": 31560
    },
    {
      "entropy": 5.672869110107422,
      "epoch": 3.1326915442635968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040356180129426356,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18364103883504868,
      "num_tokens": 69161970.0,
      "step": 31565
    },
    {
      "entropy": 5.610935974121094,
      "epoch": 3.1331877729257642,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004035327217234993,
      "loss": 4.9449,
      "mean_token_accuracy": 0.19080401062965394,
      "num_tokens": 69172990.0,
      "step": 31570
    },
    {
      "entropy": 5.621453523635864,
      "epoch": 3.1336840015879317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040350363896540054,
      "loss": 4.9734,
      "mean_token_accuracy": 0.1918248787522316,
      "num_tokens": 69184044.0,
      "step": 31575
    },
    {
      "entropy": 5.6009608745574955,
      "epoch": 3.134180230250099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040347455302068835,
      "loss": 4.9386,
      "mean_token_accuracy": 0.1817751184105873,
      "num_tokens": 69194841.0,
      "step": 31580
    },
    {
      "entropy": 5.664725828170776,
      "epoch": 3.1346764589122667,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040344546389008406,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18606604933738707,
      "num_tokens": 69205465.0,
      "step": 31585
    },
    {
      "entropy": 5.69821457862854,
      "epoch": 3.135172687574434,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040341637157430916,
      "loss": 5.0358,
      "mean_token_accuracy": 0.1852526918053627,
      "num_tokens": 69215831.0,
      "step": 31590
    },
    {
      "entropy": 5.682521390914917,
      "epoch": 3.135668916236602,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004033872760740849,
      "loss": 4.9835,
      "mean_token_accuracy": 0.18943016827106476,
      "num_tokens": 69225156.0,
      "step": 31595
    },
    {
      "entropy": 5.671972846984863,
      "epoch": 3.1361651448987695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004033581773901328,
      "loss": 5.0162,
      "mean_token_accuracy": 0.18231598734855653,
      "num_tokens": 69237590.0,
      "step": 31600
    },
    {
      "entropy": 5.7044836521148685,
      "epoch": 3.136661373560937,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040332907552317453,
      "loss": 4.9763,
      "mean_token_accuracy": 0.19141660630702972,
      "num_tokens": 69246904.0,
      "step": 31605
    },
    {
      "entropy": 5.787760972976685,
      "epoch": 3.1371576022231045,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004032999704739317,
      "loss": 5.0901,
      "mean_token_accuracy": 0.18221694380044937,
      "num_tokens": 69258289.0,
      "step": 31610
    },
    {
      "entropy": 5.689143705368042,
      "epoch": 3.137653830885272,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004032708622431261,
      "loss": 4.9851,
      "mean_token_accuracy": 0.18679921180009842,
      "num_tokens": 69268232.0,
      "step": 31615
    },
    {
      "entropy": 5.65430383682251,
      "epoch": 3.1381500595474394,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004032417508314795,
      "loss": 4.969,
      "mean_token_accuracy": 0.18240108639001845,
      "num_tokens": 69278596.0,
      "step": 31620
    },
    {
      "entropy": 5.659956932067871,
      "epoch": 3.138646288209607,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040321263623971376,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18337202221155166,
      "num_tokens": 69290675.0,
      "step": 31625
    },
    {
      "entropy": 5.701989698410034,
      "epoch": 3.1391425168717744,
      "grad_norm": 1.0,
      "learning_rate": 0.000403183518468551,
      "loss": 5.0522,
      "mean_token_accuracy": 0.18908022344112396,
      "num_tokens": 69303600.0,
      "step": 31630
    },
    {
      "entropy": 5.71670446395874,
      "epoch": 3.139638745533942,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004031543975187132,
      "loss": 5.0813,
      "mean_token_accuracy": 0.18382801860570908,
      "num_tokens": 69314672.0,
      "step": 31635
    },
    {
      "entropy": 5.622894716262818,
      "epoch": 3.1401349741961098,
      "grad_norm": 1.125,
      "learning_rate": 0.00040312527339092237,
      "loss": 4.9406,
      "mean_token_accuracy": 0.19510395377874373,
      "num_tokens": 69325574.0,
      "step": 31640
    },
    {
      "entropy": 5.6487995147705075,
      "epoch": 3.1406312028582772,
      "grad_norm": 1.125,
      "learning_rate": 0.00040309614608590095,
      "loss": 4.9343,
      "mean_token_accuracy": 0.18713704347610474,
      "num_tokens": 69336430.0,
      "step": 31645
    },
    {
      "entropy": 5.5874176025390625,
      "epoch": 3.1411274315204447,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004030670156043711,
      "loss": 4.8595,
      "mean_token_accuracy": 0.19878055453300475,
      "num_tokens": 69347007.0,
      "step": 31650
    },
    {
      "entropy": 5.660831356048584,
      "epoch": 3.141623660182612,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040303788194705523,
      "loss": 4.9957,
      "mean_token_accuracy": 0.18728003948926925,
      "num_tokens": 69357796.0,
      "step": 31655
    },
    {
      "entropy": 5.67830171585083,
      "epoch": 3.1421198888447797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040300874511467574,
      "loss": 5.0154,
      "mean_token_accuracy": 0.1883057951927185,
      "num_tokens": 69370098.0,
      "step": 31660
    },
    {
      "entropy": 5.678920125961303,
      "epoch": 3.142616117506947,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040297960510795526,
      "loss": 4.984,
      "mean_token_accuracy": 0.17964295148849488,
      "num_tokens": 69380991.0,
      "step": 31665
    },
    {
      "entropy": 5.696601676940918,
      "epoch": 3.1431123461691146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004029504619276163,
      "loss": 4.988,
      "mean_token_accuracy": 0.18553187549114228,
      "num_tokens": 69391748.0,
      "step": 31670
    },
    {
      "entropy": 5.7340068340301515,
      "epoch": 3.143608574831282,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040292131557438166,
      "loss": 5.0351,
      "mean_token_accuracy": 0.18140330016613007,
      "num_tokens": 69404964.0,
      "step": 31675
    },
    {
      "entropy": 5.62988247871399,
      "epoch": 3.14410480349345,
      "grad_norm": 1.171875,
      "learning_rate": 0.000402892166048974,
      "loss": 4.9615,
      "mean_token_accuracy": 0.18808364123106003,
      "num_tokens": 69416045.0,
      "step": 31680
    },
    {
      "entropy": 5.640719223022461,
      "epoch": 3.1446010321556175,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004028630133521162,
      "loss": 4.9511,
      "mean_token_accuracy": 0.18830210566520691,
      "num_tokens": 69427489.0,
      "step": 31685
    },
    {
      "entropy": 5.616275501251221,
      "epoch": 3.145097260817785,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004028338574845312,
      "loss": 4.9644,
      "mean_token_accuracy": 0.1881508082151413,
      "num_tokens": 69437929.0,
      "step": 31690
    },
    {
      "entropy": 5.6512950420379635,
      "epoch": 3.1455934894799524,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040280469844694203,
      "loss": 5.0208,
      "mean_token_accuracy": 0.18915736973285674,
      "num_tokens": 69449449.0,
      "step": 31695
    },
    {
      "entropy": 5.644449472427368,
      "epoch": 3.14608971814212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004027755362400716,
      "loss": 5.023,
      "mean_token_accuracy": 0.1836494594812393,
      "num_tokens": 69459451.0,
      "step": 31700
    },
    {
      "entropy": 5.7120890617370605,
      "epoch": 3.1465859468042874,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004027463708646434,
      "loss": 5.0342,
      "mean_token_accuracy": 0.1786272332072258,
      "num_tokens": 69470615.0,
      "step": 31705
    },
    {
      "entropy": 5.70830774307251,
      "epoch": 3.147082175466455,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004027172023213803,
      "loss": 4.9874,
      "mean_token_accuracy": 0.18835387527942657,
      "num_tokens": 69481123.0,
      "step": 31710
    },
    {
      "entropy": 5.694554710388184,
      "epoch": 3.1475784041286223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004026880306110058,
      "loss": 5.0219,
      "mean_token_accuracy": 0.1829851046204567,
      "num_tokens": 69491590.0,
      "step": 31715
    },
    {
      "entropy": 5.602756500244141,
      "epoch": 3.14807463279079,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00040265885573424336,
      "loss": 4.9899,
      "mean_token_accuracy": 0.18708622604608535,
      "num_tokens": 69501883.0,
      "step": 31720
    },
    {
      "entropy": 5.613150262832642,
      "epoch": 3.1485708614529577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040262967769181634,
      "loss": 4.9478,
      "mean_token_accuracy": 0.18603047877550125,
      "num_tokens": 69514556.0,
      "step": 31725
    },
    {
      "entropy": 5.723021078109741,
      "epoch": 3.149067090115125,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040260049648444836,
      "loss": 5.0761,
      "mean_token_accuracy": 0.18295336663722991,
      "num_tokens": 69525483.0,
      "step": 31730
    },
    {
      "entropy": 5.692237472534179,
      "epoch": 3.1495633187772927,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00040257131211286307,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18897990137338638,
      "num_tokens": 69535681.0,
      "step": 31735
    },
    {
      "entropy": 5.721411514282226,
      "epoch": 3.15005954743946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000402542124577784,
      "loss": 5.1175,
      "mean_token_accuracy": 0.1696932926774025,
      "num_tokens": 69547825.0,
      "step": 31740
    },
    {
      "entropy": 5.65129599571228,
      "epoch": 3.1505557761016276,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040251293387993513,
      "loss": 4.9731,
      "mean_token_accuracy": 0.18929697275161744,
      "num_tokens": 69557922.0,
      "step": 31745
    },
    {
      "entropy": 5.681573915481567,
      "epoch": 3.151052004763795,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004024837400200403,
      "loss": 5.0529,
      "mean_token_accuracy": 0.1816951870918274,
      "num_tokens": 69568165.0,
      "step": 31750
    },
    {
      "entropy": 5.620519781112671,
      "epoch": 3.1515482334259626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040245454299882343,
      "loss": 4.8387,
      "mean_token_accuracy": 0.19805925190448762,
      "num_tokens": 69578828.0,
      "step": 31755
    },
    {
      "entropy": 5.665142345428467,
      "epoch": 3.15204446208813,
      "grad_norm": 1.125,
      "learning_rate": 0.00040242534281700843,
      "loss": 5.0335,
      "mean_token_accuracy": 0.18510703444480897,
      "num_tokens": 69590527.0,
      "step": 31760
    },
    {
      "entropy": 5.672067546844483,
      "epoch": 3.152540690750298,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040239613947531965,
      "loss": 5.0047,
      "mean_token_accuracy": 0.1877628356218338,
      "num_tokens": 69602032.0,
      "step": 31765
    },
    {
      "entropy": 5.6354179859161375,
      "epoch": 3.1530369194124654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040236693297448107,
      "loss": 5.0073,
      "mean_token_accuracy": 0.181374229490757,
      "num_tokens": 69613490.0,
      "step": 31770
    },
    {
      "entropy": 5.7123208999633786,
      "epoch": 3.153533148074633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004023377233152169,
      "loss": 5.0941,
      "mean_token_accuracy": 0.1798441454768181,
      "num_tokens": 69624623.0,
      "step": 31775
    },
    {
      "entropy": 5.69431438446045,
      "epoch": 3.1540293767368004,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040230851049825173,
      "loss": 5.05,
      "mean_token_accuracy": 0.18392330408096313,
      "num_tokens": 69634956.0,
      "step": 31780
    },
    {
      "entropy": 5.682680416107178,
      "epoch": 3.154525605398968,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040227929452430974,
      "loss": 4.9716,
      "mean_token_accuracy": 0.1953323721885681,
      "num_tokens": 69645370.0,
      "step": 31785
    },
    {
      "entropy": 5.713923692703247,
      "epoch": 3.1550218340611353,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040225007539411553,
      "loss": 5.0418,
      "mean_token_accuracy": 0.17513469010591506,
      "num_tokens": 69656328.0,
      "step": 31790
    },
    {
      "entropy": 5.638432931900025,
      "epoch": 3.155518062723303,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004022208531083936,
      "loss": 4.9934,
      "mean_token_accuracy": 0.1827361613512039,
      "num_tokens": 69668584.0,
      "step": 31795
    },
    {
      "entropy": 5.642061948776245,
      "epoch": 3.1560142913854703,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004021916276678688,
      "loss": 5.025,
      "mean_token_accuracy": 0.18129991888999938,
      "num_tokens": 69677977.0,
      "step": 31800
    },
    {
      "entropy": 5.640947961807251,
      "epoch": 3.1565105200476378,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040216239907326555,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18687686324119568,
      "num_tokens": 69688979.0,
      "step": 31805
    },
    {
      "entropy": 5.663319492340088,
      "epoch": 3.1570067487098052,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004021331673253089,
      "loss": 5.0327,
      "mean_token_accuracy": 0.18767621666193007,
      "num_tokens": 69698567.0,
      "step": 31810
    },
    {
      "entropy": 5.691408348083496,
      "epoch": 3.157502977371973,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004021039324247236,
      "loss": 4.9513,
      "mean_token_accuracy": 0.18615829646587373,
      "num_tokens": 69708623.0,
      "step": 31815
    },
    {
      "entropy": 5.626473522186279,
      "epoch": 3.1579992060341406,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040207469437223466,
      "loss": 5.0145,
      "mean_token_accuracy": 0.18558835834264756,
      "num_tokens": 69719613.0,
      "step": 31820
    },
    {
      "entropy": 5.641735935211182,
      "epoch": 3.158495434696308,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040204545316856714,
      "loss": 5.0048,
      "mean_token_accuracy": 0.18268322199583054,
      "num_tokens": 69731407.0,
      "step": 31825
    },
    {
      "entropy": 5.6412327766418455,
      "epoch": 3.1589916633584756,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004020162088144461,
      "loss": 4.9492,
      "mean_token_accuracy": 0.18899863362312316,
      "num_tokens": 69742363.0,
      "step": 31830
    },
    {
      "entropy": 5.629723072052002,
      "epoch": 3.159487892020643,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040198696131059684,
      "loss": 4.9577,
      "mean_token_accuracy": 0.19204029440879822,
      "num_tokens": 69754878.0,
      "step": 31835
    },
    {
      "entropy": 5.636578798294067,
      "epoch": 3.1599841206828105,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040195771065774455,
      "loss": 4.932,
      "mean_token_accuracy": 0.18974061012268068,
      "num_tokens": 69764986.0,
      "step": 31840
    },
    {
      "entropy": 5.6310406684875485,
      "epoch": 3.160480349344978,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004019284568566146,
      "loss": 4.9756,
      "mean_token_accuracy": 0.1919396236538887,
      "num_tokens": 69775762.0,
      "step": 31845
    },
    {
      "entropy": 5.653202295303345,
      "epoch": 3.1609765780071455,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040189919990793243,
      "loss": 5.0388,
      "mean_token_accuracy": 0.1869576469063759,
      "num_tokens": 69786411.0,
      "step": 31850
    },
    {
      "entropy": 5.701563024520874,
      "epoch": 3.1614728066693134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00040186993981242353,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18190045952796935,
      "num_tokens": 69797140.0,
      "step": 31855
    },
    {
      "entropy": 5.627822828292847,
      "epoch": 3.161969035331481,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004018406765708135,
      "loss": 4.8904,
      "mean_token_accuracy": 0.19725941121578217,
      "num_tokens": 69807452.0,
      "step": 31860
    },
    {
      "entropy": 5.618134784698486,
      "epoch": 3.1624652639936484,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040181141018382806,
      "loss": 4.972,
      "mean_token_accuracy": 0.1891648456454277,
      "num_tokens": 69819650.0,
      "step": 31865
    },
    {
      "entropy": 5.6309511184692385,
      "epoch": 3.162961492655816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040178214065219285,
      "loss": 4.9777,
      "mean_token_accuracy": 0.18816299140453338,
      "num_tokens": 69830411.0,
      "step": 31870
    },
    {
      "entropy": 5.647471618652344,
      "epoch": 3.1634577213179833,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004017528679766337,
      "loss": 4.958,
      "mean_token_accuracy": 0.1977603852748871,
      "num_tokens": 69841795.0,
      "step": 31875
    },
    {
      "entropy": 5.610672521591186,
      "epoch": 3.163953949980151,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004017235921578767,
      "loss": 4.9798,
      "mean_token_accuracy": 0.18598906695842743,
      "num_tokens": 69851633.0,
      "step": 31880
    },
    {
      "entropy": 5.613911437988281,
      "epoch": 3.1644501786423183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040169431319664756,
      "loss": 4.9013,
      "mean_token_accuracy": 0.19655187875032426,
      "num_tokens": 69862398.0,
      "step": 31885
    },
    {
      "entropy": 5.641408491134643,
      "epoch": 3.1649464073044857,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004016650310936725,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1882606789469719,
      "num_tokens": 69874397.0,
      "step": 31890
    },
    {
      "entropy": 5.671711587905884,
      "epoch": 3.165442635966653,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004016357458496776,
      "loss": 4.9308,
      "mean_token_accuracy": 0.1991926833987236,
      "num_tokens": 69883950.0,
      "step": 31895
    },
    {
      "entropy": 5.636527633666992,
      "epoch": 3.165938864628821,
      "grad_norm": 1.125,
      "learning_rate": 0.000401606457465389,
      "loss": 4.9228,
      "mean_token_accuracy": 0.19185782819986344,
      "num_tokens": 69895586.0,
      "step": 31900
    },
    {
      "entropy": 5.6608318328857425,
      "epoch": 3.1664350932909886,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004015771659415332,
      "loss": 4.9638,
      "mean_token_accuracy": 0.19635046124458314,
      "num_tokens": 69906093.0,
      "step": 31905
    },
    {
      "entropy": 5.609019136428833,
      "epoch": 3.166931321953156,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004015478712788364,
      "loss": 4.9396,
      "mean_token_accuracy": 0.18798002004623413,
      "num_tokens": 69916381.0,
      "step": 31910
    },
    {
      "entropy": 5.634754085540772,
      "epoch": 3.1674275506153236,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040151857347802506,
      "loss": 4.9567,
      "mean_token_accuracy": 0.19256277680397033,
      "num_tokens": 69926059.0,
      "step": 31915
    },
    {
      "entropy": 5.62621579170227,
      "epoch": 3.167923779277491,
      "grad_norm": 1.125,
      "learning_rate": 0.00040148927253982577,
      "loss": 4.9847,
      "mean_token_accuracy": 0.18847695887088775,
      "num_tokens": 69937677.0,
      "step": 31920
    },
    {
      "entropy": 5.663166570663452,
      "epoch": 3.1684200079396585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040145996846496497,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18527337163686752,
      "num_tokens": 69949399.0,
      "step": 31925
    },
    {
      "entropy": 5.67116060256958,
      "epoch": 3.168916236601826,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004014306612541695,
      "loss": 5.0233,
      "mean_token_accuracy": 0.19621526449918747,
      "num_tokens": 69960595.0,
      "step": 31930
    },
    {
      "entropy": 5.649838876724243,
      "epoch": 3.1694124652639935,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004014013509081661,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18764977008104325,
      "num_tokens": 69971018.0,
      "step": 31935
    },
    {
      "entropy": 5.688103437423706,
      "epoch": 3.1699086939261614,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00040137203742768155,
      "loss": 4.979,
      "mean_token_accuracy": 0.18440223336219788,
      "num_tokens": 69982507.0,
      "step": 31940
    },
    {
      "entropy": 5.654143476486206,
      "epoch": 3.170404922588329,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004013427208134427,
      "loss": 4.9474,
      "mean_token_accuracy": 0.19822347909212112,
      "num_tokens": 69994190.0,
      "step": 31945
    },
    {
      "entropy": 5.68930344581604,
      "epoch": 3.1709011512504963,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040131340106617673,
      "loss": 5.1552,
      "mean_token_accuracy": 0.17992264330387114,
      "num_tokens": 70005522.0,
      "step": 31950
    },
    {
      "entropy": 5.753449726104736,
      "epoch": 3.171397379912664,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004012840781866106,
      "loss": 5.0407,
      "mean_token_accuracy": 0.1925917312502861,
      "num_tokens": 70016706.0,
      "step": 31955
    },
    {
      "entropy": 5.676651906967163,
      "epoch": 3.1718936085748313,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040125475217547135,
      "loss": 4.9508,
      "mean_token_accuracy": 0.1889388844370842,
      "num_tokens": 70028577.0,
      "step": 31960
    },
    {
      "entropy": 5.608208799362183,
      "epoch": 3.1723898372369987,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040122542303348634,
      "loss": 4.9694,
      "mean_token_accuracy": 0.18972748816013335,
      "num_tokens": 70040258.0,
      "step": 31965
    },
    {
      "entropy": 5.686326265335083,
      "epoch": 3.172886065899166,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004011960907613828,
      "loss": 5.0695,
      "mean_token_accuracy": 0.1820374235510826,
      "num_tokens": 70052105.0,
      "step": 31970
    },
    {
      "entropy": 5.672727298736572,
      "epoch": 3.1733822945613337,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040116675535988825,
      "loss": 4.9634,
      "mean_token_accuracy": 0.19029318243265153,
      "num_tokens": 70062272.0,
      "step": 31975
    },
    {
      "entropy": 5.653440523147583,
      "epoch": 3.173878523223501,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040113741682972994,
      "loss": 4.9882,
      "mean_token_accuracy": 0.19032527655363082,
      "num_tokens": 70073614.0,
      "step": 31980
    },
    {
      "entropy": 5.609822559356689,
      "epoch": 3.174374751885669,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004011080751716356,
      "loss": 4.9854,
      "mean_token_accuracy": 0.1807964935898781,
      "num_tokens": 70085191.0,
      "step": 31985
    },
    {
      "entropy": 5.686111116409302,
      "epoch": 3.1748709805478366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040107873038633255,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1840754896402359,
      "num_tokens": 70097610.0,
      "step": 31990
    },
    {
      "entropy": 5.63992977142334,
      "epoch": 3.175367209210004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004010493824745487,
      "loss": 4.9252,
      "mean_token_accuracy": 0.18903598040342331,
      "num_tokens": 70109040.0,
      "step": 31995
    },
    {
      "entropy": 5.646824312210083,
      "epoch": 3.1758634378721715,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040102003143701194,
      "loss": 5.009,
      "mean_token_accuracy": 0.18117888867855073,
      "num_tokens": 70120472.0,
      "step": 32000
    },
    {
      "entropy": 5.645576906204224,
      "epoch": 3.176359666534339,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040099067727444984,
      "loss": 4.905,
      "mean_token_accuracy": 0.19197281897068025,
      "num_tokens": 70130481.0,
      "step": 32005
    },
    {
      "entropy": 5.656555032730102,
      "epoch": 3.1768558951965065,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004009613199875904,
      "loss": 4.9265,
      "mean_token_accuracy": 0.1913239672780037,
      "num_tokens": 70140783.0,
      "step": 32010
    },
    {
      "entropy": 5.6781645774841305,
      "epoch": 3.177352123858674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004009319595771617,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18329784721136094,
      "num_tokens": 70151624.0,
      "step": 32015
    },
    {
      "entropy": 5.687900114059448,
      "epoch": 3.1778483525208414,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004009025960438918,
      "loss": 5.008,
      "mean_token_accuracy": 0.18372306674718858,
      "num_tokens": 70162600.0,
      "step": 32020
    },
    {
      "entropy": 5.697344398498535,
      "epoch": 3.1783445811830093,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040087322938850867,
      "loss": 5.0389,
      "mean_token_accuracy": 0.18474187850952148,
      "num_tokens": 70175563.0,
      "step": 32025
    },
    {
      "entropy": 5.659175443649292,
      "epoch": 3.178840809845177,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004008438596117408,
      "loss": 4.9319,
      "mean_token_accuracy": 0.1830044284462929,
      "num_tokens": 70186358.0,
      "step": 32030
    },
    {
      "entropy": 5.707951927185059,
      "epoch": 3.1793370385073443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040081448671431635,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18539282232522963,
      "num_tokens": 70196872.0,
      "step": 32035
    },
    {
      "entropy": 5.64186053276062,
      "epoch": 3.1798332671695118,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040078511069696377,
      "loss": 4.9636,
      "mean_token_accuracy": 0.18301561772823333,
      "num_tokens": 70208136.0,
      "step": 32040
    },
    {
      "entropy": 5.724837160110473,
      "epoch": 3.1803294958316792,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040075573156041146,
      "loss": 5.2035,
      "mean_token_accuracy": 0.17640179842710496,
      "num_tokens": 70219716.0,
      "step": 32045
    },
    {
      "entropy": 5.691476917266845,
      "epoch": 3.1808257244938467,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040072634930538796,
      "loss": 5.0283,
      "mean_token_accuracy": 0.1801580384373665,
      "num_tokens": 70230502.0,
      "step": 32050
    },
    {
      "entropy": 5.709350442886352,
      "epoch": 3.181321953156014,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040069696393262196,
      "loss": 5.029,
      "mean_token_accuracy": 0.1835069939494133,
      "num_tokens": 70241820.0,
      "step": 32055
    },
    {
      "entropy": 5.681621408462524,
      "epoch": 3.1818181818181817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004006675754428421,
      "loss": 4.9437,
      "mean_token_accuracy": 0.1885718286037445,
      "num_tokens": 70252243.0,
      "step": 32060
    },
    {
      "entropy": 5.630124282836914,
      "epoch": 3.182314410480349,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040063818383677716,
      "loss": 4.8874,
      "mean_token_accuracy": 0.1887117236852646,
      "num_tokens": 70261835.0,
      "step": 32065
    },
    {
      "entropy": 5.704461097717285,
      "epoch": 3.182810639142517,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000400608789115156,
      "loss": 5.037,
      "mean_token_accuracy": 0.18789269477128984,
      "num_tokens": 70273210.0,
      "step": 32070
    },
    {
      "entropy": 5.657616472244262,
      "epoch": 3.1833068678046845,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040057939127870754,
      "loss": 5.0059,
      "mean_token_accuracy": 0.1832543820142746,
      "num_tokens": 70283904.0,
      "step": 32075
    },
    {
      "entropy": 5.734863901138306,
      "epoch": 3.183803096466852,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004005499903281607,
      "loss": 4.9919,
      "mean_token_accuracy": 0.19408566057682036,
      "num_tokens": 70295199.0,
      "step": 32080
    },
    {
      "entropy": 5.6327211380004885,
      "epoch": 3.1842993251290195,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040052058626424464,
      "loss": 4.8997,
      "mean_token_accuracy": 0.19090789705514907,
      "num_tokens": 70305418.0,
      "step": 32085
    },
    {
      "entropy": 5.65957670211792,
      "epoch": 3.184795553791187,
      "grad_norm": 1.125,
      "learning_rate": 0.0004004911790876886,
      "loss": 5.0605,
      "mean_token_accuracy": 0.1808157354593277,
      "num_tokens": 70316131.0,
      "step": 32090
    },
    {
      "entropy": 5.63316240310669,
      "epoch": 3.1852917824533544,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00040046176879922156,
      "loss": 4.909,
      "mean_token_accuracy": 0.19204880446195602,
      "num_tokens": 70326804.0,
      "step": 32095
    },
    {
      "entropy": 5.763708257675171,
      "epoch": 3.185788011115522,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040043235539957315,
      "loss": 5.1188,
      "mean_token_accuracy": 0.17943723499774933,
      "num_tokens": 70338125.0,
      "step": 32100
    },
    {
      "entropy": 5.677971410751343,
      "epoch": 3.1862842397776894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004004029388894726,
      "loss": 4.9549,
      "mean_token_accuracy": 0.18815961629152297,
      "num_tokens": 70348614.0,
      "step": 32105
    },
    {
      "entropy": 5.670093822479248,
      "epoch": 3.1867804684398573,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004003735192696494,
      "loss": 5.0293,
      "mean_token_accuracy": 0.18431285321712493,
      "num_tokens": 70358735.0,
      "step": 32110
    },
    {
      "entropy": 5.612734746932984,
      "epoch": 3.1872766971020248,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000400344096540833,
      "loss": 4.8988,
      "mean_token_accuracy": 0.19576579034328462,
      "num_tokens": 70369448.0,
      "step": 32115
    },
    {
      "entropy": 5.6623640060424805,
      "epoch": 3.1877729257641922,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040031467070375315,
      "loss": 4.9533,
      "mean_token_accuracy": 0.1891726866364479,
      "num_tokens": 70379462.0,
      "step": 32120
    },
    {
      "entropy": 5.668935394287109,
      "epoch": 3.1882691544263597,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004002852417591395,
      "loss": 4.9981,
      "mean_token_accuracy": 0.1845574587583542,
      "num_tokens": 70389283.0,
      "step": 32125
    },
    {
      "entropy": 5.632357454299926,
      "epoch": 3.188765383088527,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004002558097077218,
      "loss": 5.0218,
      "mean_token_accuracy": 0.1871291443705559,
      "num_tokens": 70400058.0,
      "step": 32130
    },
    {
      "entropy": 5.624167013168335,
      "epoch": 3.1892616117506947,
      "grad_norm": 1.125,
      "learning_rate": 0.0004002263745502299,
      "loss": 4.9083,
      "mean_token_accuracy": 0.19576812386512757,
      "num_tokens": 70411614.0,
      "step": 32135
    },
    {
      "entropy": 5.6672014713287355,
      "epoch": 3.189757840412862,
      "grad_norm": 1.109375,
      "learning_rate": 0.00040019693628739377,
      "loss": 4.9674,
      "mean_token_accuracy": 0.18533118218183517,
      "num_tokens": 70422404.0,
      "step": 32140
    },
    {
      "entropy": 5.742796230316162,
      "epoch": 3.1902540690750296,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004001674949199434,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18524208217859267,
      "num_tokens": 70432219.0,
      "step": 32145
    },
    {
      "entropy": 5.660440111160279,
      "epoch": 3.190750297737197,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004001380504486089,
      "loss": 4.9954,
      "mean_token_accuracy": 0.18850134313106537,
      "num_tokens": 70443129.0,
      "step": 32150
    },
    {
      "entropy": 5.669123744964599,
      "epoch": 3.191246526399365,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004001086028741204,
      "loss": 5.0069,
      "mean_token_accuracy": 0.18875257819890975,
      "num_tokens": 70453225.0,
      "step": 32155
    },
    {
      "entropy": 5.616152620315551,
      "epoch": 3.1917427550615325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004000791521972081,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18563943207263947,
      "num_tokens": 70464909.0,
      "step": 32160
    },
    {
      "entropy": 5.67049503326416,
      "epoch": 3.1922389837237,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00040004969841860233,
      "loss": 4.9989,
      "mean_token_accuracy": 0.18848163336515428,
      "num_tokens": 70475211.0,
      "step": 32165
    },
    {
      "entropy": 5.748235130310059,
      "epoch": 3.1927352123858674,
      "grad_norm": 1.25,
      "learning_rate": 0.00040002024153903347,
      "loss": 5.0075,
      "mean_token_accuracy": 0.18782986104488372,
      "num_tokens": 70485603.0,
      "step": 32170
    },
    {
      "entropy": 5.652606105804443,
      "epoch": 3.193231441048035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003999907815592321,
      "loss": 4.9204,
      "mean_token_accuracy": 0.1949745759367943,
      "num_tokens": 70497163.0,
      "step": 32175
    },
    {
      "entropy": 5.660182237625122,
      "epoch": 3.1937276697102024,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039996131847992863,
      "loss": 4.9702,
      "mean_token_accuracy": 0.1933565229177475,
      "num_tokens": 70509323.0,
      "step": 32180
    },
    {
      "entropy": 5.686785125732422,
      "epoch": 3.19422389837237,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003999318523018537,
      "loss": 5.0907,
      "mean_token_accuracy": 0.18070805817842484,
      "num_tokens": 70520767.0,
      "step": 32185
    },
    {
      "entropy": 5.6167387008667,
      "epoch": 3.1947201270345373,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003999023830257382,
      "loss": 4.9352,
      "mean_token_accuracy": 0.18266466110944748,
      "num_tokens": 70531671.0,
      "step": 32190
    },
    {
      "entropy": 5.640625524520874,
      "epoch": 3.1952163556967053,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003998729106523126,
      "loss": 4.9698,
      "mean_token_accuracy": 0.1872660383582115,
      "num_tokens": 70542309.0,
      "step": 32195
    },
    {
      "entropy": 5.708175420761108,
      "epoch": 3.1957125843588727,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003998434351823078,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18456467986106873,
      "num_tokens": 70553003.0,
      "step": 32200
    },
    {
      "entropy": 5.635298109054565,
      "epoch": 3.19620881302104,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000399813956616455,
      "loss": 4.9933,
      "mean_token_accuracy": 0.18530155718326569,
      "num_tokens": 70564161.0,
      "step": 32205
    },
    {
      "entropy": 5.681760931015015,
      "epoch": 3.1967050416832077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039978447495548497,
      "loss": 4.9551,
      "mean_token_accuracy": 0.19225687086582183,
      "num_tokens": 70575891.0,
      "step": 32210
    },
    {
      "entropy": 5.705931997299194,
      "epoch": 3.197201270345375,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039975499020012886,
      "loss": 5.0971,
      "mean_token_accuracy": 0.18773213475942613,
      "num_tokens": 70586691.0,
      "step": 32215
    },
    {
      "entropy": 5.634438991546631,
      "epoch": 3.1976974990075426,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039972550235111776,
      "loss": 4.9285,
      "mean_token_accuracy": 0.18916135430335998,
      "num_tokens": 70597423.0,
      "step": 32220
    },
    {
      "entropy": 5.650142574310303,
      "epoch": 3.19819372766971,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000399696011409183,
      "loss": 4.9856,
      "mean_token_accuracy": 0.18725373595952988,
      "num_tokens": 70609828.0,
      "step": 32225
    },
    {
      "entropy": 5.600301933288574,
      "epoch": 3.1986899563318776,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039966651737505585,
      "loss": 4.9216,
      "mean_token_accuracy": 0.188231985270977,
      "num_tokens": 70620480.0,
      "step": 32230
    },
    {
      "entropy": 5.6638885021209715,
      "epoch": 3.199186184994045,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003996370202494677,
      "loss": 4.9883,
      "mean_token_accuracy": 0.19092340171337127,
      "num_tokens": 70630045.0,
      "step": 32235
    },
    {
      "entropy": 5.641228437423706,
      "epoch": 3.199682413656213,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039960752003315,
      "loss": 4.9063,
      "mean_token_accuracy": 0.1950401172041893,
      "num_tokens": 70640142.0,
      "step": 32240
    },
    {
      "entropy": 5.601183748245239,
      "epoch": 3.2001786423183805,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003995780167268343,
      "loss": 4.9692,
      "mean_token_accuracy": 0.18778800070285798,
      "num_tokens": 70652062.0,
      "step": 32245
    },
    {
      "entropy": 5.617107963562011,
      "epoch": 3.200674870980548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003995485103312523,
      "loss": 4.9965,
      "mean_token_accuracy": 0.1853554680943489,
      "num_tokens": 70666187.0,
      "step": 32250
    },
    {
      "entropy": 5.73950161933899,
      "epoch": 3.2011710996427154,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039951900084713555,
      "loss": 5.062,
      "mean_token_accuracy": 0.18776675164699555,
      "num_tokens": 70677432.0,
      "step": 32255
    },
    {
      "entropy": 5.682144832611084,
      "epoch": 3.201667328304883,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003994894882752159,
      "loss": 4.9548,
      "mean_token_accuracy": 0.18847829401493071,
      "num_tokens": 70686963.0,
      "step": 32260
    },
    {
      "entropy": 5.7156154155731205,
      "epoch": 3.2021635569670504,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039945997261622526,
      "loss": 5.1351,
      "mean_token_accuracy": 0.17764229476451873,
      "num_tokens": 70698345.0,
      "step": 32265
    },
    {
      "entropy": 5.656388998031616,
      "epoch": 3.202659785629218,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003994304538708954,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18904411643743516,
      "num_tokens": 70709695.0,
      "step": 32270
    },
    {
      "entropy": 5.6398993015289305,
      "epoch": 3.2031560142913853,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003994009320399584,
      "loss": 4.9254,
      "mean_token_accuracy": 0.18826599419116974,
      "num_tokens": 70720843.0,
      "step": 32275
    },
    {
      "entropy": 5.64176378250122,
      "epoch": 3.203652242953553,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039937140712414644,
      "loss": 5.027,
      "mean_token_accuracy": 0.1847139224410057,
      "num_tokens": 70731886.0,
      "step": 32280
    },
    {
      "entropy": 5.664373445510864,
      "epoch": 3.2041484716157207,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003993418791241915,
      "loss": 5.0038,
      "mean_token_accuracy": 0.18980195373296738,
      "num_tokens": 70741838.0,
      "step": 32285
    },
    {
      "entropy": 5.646849489212036,
      "epoch": 3.204644700277888,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003993123480408259,
      "loss": 4.9203,
      "mean_token_accuracy": 0.19901324212551116,
      "num_tokens": 70751454.0,
      "step": 32290
    },
    {
      "entropy": 5.695271158218384,
      "epoch": 3.2051409289400556,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039928281387478207,
      "loss": 5.015,
      "mean_token_accuracy": 0.18060576766729355,
      "num_tokens": 70762294.0,
      "step": 32295
    },
    {
      "entropy": 5.631837701797485,
      "epoch": 3.205637157602223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003992532766267921,
      "loss": 4.9476,
      "mean_token_accuracy": 0.1921502485871315,
      "num_tokens": 70773186.0,
      "step": 32300
    },
    {
      "entropy": 5.667434167861939,
      "epoch": 3.2061333862643906,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003992237362975887,
      "loss": 4.9801,
      "mean_token_accuracy": 0.18408251702785491,
      "num_tokens": 70783984.0,
      "step": 32305
    },
    {
      "entropy": 5.633081912994385,
      "epoch": 3.206629614926558,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039919419288790423,
      "loss": 4.9545,
      "mean_token_accuracy": 0.18804371803998948,
      "num_tokens": 70795779.0,
      "step": 32310
    },
    {
      "entropy": 5.650979709625244,
      "epoch": 3.2071258435887255,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039916464639847145,
      "loss": 4.9422,
      "mean_token_accuracy": 0.18892617970705033,
      "num_tokens": 70807247.0,
      "step": 32315
    },
    {
      "entropy": 5.687059307098389,
      "epoch": 3.207622072250893,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003991350968300229,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18110007792711258,
      "num_tokens": 70818507.0,
      "step": 32320
    },
    {
      "entropy": 5.667143630981445,
      "epoch": 3.208118300913061,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003991055441832916,
      "loss": 5.0114,
      "mean_token_accuracy": 0.1828716829419136,
      "num_tokens": 70830055.0,
      "step": 32325
    },
    {
      "entropy": 5.714956474304199,
      "epoch": 3.2086145295752284,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039907598845901016,
      "loss": 5.0946,
      "mean_token_accuracy": 0.18310026228427886,
      "num_tokens": 70841052.0,
      "step": 32330
    },
    {
      "entropy": 5.727729606628418,
      "epoch": 3.209110758237396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003990464296579114,
      "loss": 5.0326,
      "mean_token_accuracy": 0.18777697235345842,
      "num_tokens": 70852148.0,
      "step": 32335
    },
    {
      "entropy": 5.634010934829712,
      "epoch": 3.2096069868995634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003990168677807287,
      "loss": 5.011,
      "mean_token_accuracy": 0.18340205103158952,
      "num_tokens": 70863998.0,
      "step": 32340
    },
    {
      "entropy": 5.666911983489991,
      "epoch": 3.210103215561731,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003989873028281948,
      "loss": 5.0827,
      "mean_token_accuracy": 0.17986435443162918,
      "num_tokens": 70874013.0,
      "step": 32345
    },
    {
      "entropy": 5.649129438400268,
      "epoch": 3.2105994442238983,
      "grad_norm": 1.140625,
      "learning_rate": 0.000398957734801043,
      "loss": 4.9019,
      "mean_token_accuracy": 0.19523212164640427,
      "num_tokens": 70885600.0,
      "step": 32350
    },
    {
      "entropy": 5.588941431045532,
      "epoch": 3.211095672886066,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039892816370000657,
      "loss": 4.8906,
      "mean_token_accuracy": 0.19690209925174712,
      "num_tokens": 70896447.0,
      "step": 32355
    },
    {
      "entropy": 5.717120599746704,
      "epoch": 3.2115919015482333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003988985895258186,
      "loss": 5.0562,
      "mean_token_accuracy": 0.18256372809410096,
      "num_tokens": 70907558.0,
      "step": 32360
    },
    {
      "entropy": 5.676843452453613,
      "epoch": 3.212088130210401,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039886901227921263,
      "loss": 4.9849,
      "mean_token_accuracy": 0.18708329051733016,
      "num_tokens": 70919687.0,
      "step": 32365
    },
    {
      "entropy": 5.653896999359131,
      "epoch": 3.2125843588725687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003988394319609221,
      "loss": 4.9622,
      "mean_token_accuracy": 0.18825570940971376,
      "num_tokens": 70930417.0,
      "step": 32370
    },
    {
      "entropy": 5.629243326187134,
      "epoch": 3.213080587534736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003988098485716804,
      "loss": 4.9536,
      "mean_token_accuracy": 0.18844211250543594,
      "num_tokens": 70941721.0,
      "step": 32375
    },
    {
      "entropy": 5.663503932952881,
      "epoch": 3.2135768161969036,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039878026211222136,
      "loss": 4.9384,
      "mean_token_accuracy": 0.1931072875857353,
      "num_tokens": 70952616.0,
      "step": 32380
    },
    {
      "entropy": 5.601870584487915,
      "epoch": 3.214073044859071,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039875067258327853,
      "loss": 4.9455,
      "mean_token_accuracy": 0.1931648626923561,
      "num_tokens": 70963236.0,
      "step": 32385
    },
    {
      "entropy": 5.641867542266846,
      "epoch": 3.2145692735212386,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003987210799855856,
      "loss": 5.0272,
      "mean_token_accuracy": 0.1837293952703476,
      "num_tokens": 70973376.0,
      "step": 32390
    },
    {
      "entropy": 5.658045673370362,
      "epoch": 3.215065502183406,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039869148431987653,
      "loss": 5.0086,
      "mean_token_accuracy": 0.1847960829734802,
      "num_tokens": 70984937.0,
      "step": 32395
    },
    {
      "entropy": 5.772065448760986,
      "epoch": 3.2155617308455735,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003986618855868852,
      "loss": 5.0743,
      "mean_token_accuracy": 0.18665525168180466,
      "num_tokens": 70997023.0,
      "step": 32400
    },
    {
      "entropy": 5.675120067596436,
      "epoch": 3.216057959507741,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003986322837873455,
      "loss": 4.9324,
      "mean_token_accuracy": 0.19339810609817504,
      "num_tokens": 71008304.0,
      "step": 32405
    },
    {
      "entropy": 5.621506309509277,
      "epoch": 3.216554188169909,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039860267892199167,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18632873743772507,
      "num_tokens": 71020160.0,
      "step": 32410
    },
    {
      "entropy": 5.709780883789063,
      "epoch": 3.2170504168320764,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039857307099155773,
      "loss": 5.031,
      "mean_token_accuracy": 0.19206561297178268,
      "num_tokens": 71031701.0,
      "step": 32415
    },
    {
      "entropy": 5.677289247512817,
      "epoch": 3.217546645494244,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039854345999677777,
      "loss": 5.0373,
      "mean_token_accuracy": 0.1917920485138893,
      "num_tokens": 71041955.0,
      "step": 32420
    },
    {
      "entropy": 5.676652193069458,
      "epoch": 3.2180428741564113,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039851384593838637,
      "loss": 5.0879,
      "mean_token_accuracy": 0.18572293668985368,
      "num_tokens": 71052884.0,
      "step": 32425
    },
    {
      "entropy": 5.655351018905639,
      "epoch": 3.218539102818579,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039848422881711765,
      "loss": 4.9551,
      "mean_token_accuracy": 0.18736945688724518,
      "num_tokens": 71065091.0,
      "step": 32430
    },
    {
      "entropy": 5.699371814727783,
      "epoch": 3.2190353314807463,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003984546086337061,
      "loss": 5.0526,
      "mean_token_accuracy": 0.18379932790994644,
      "num_tokens": 71075823.0,
      "step": 32435
    },
    {
      "entropy": 5.7047347068786625,
      "epoch": 3.2195315601429138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039842498538888646,
      "loss": 4.9917,
      "mean_token_accuracy": 0.19368127435445787,
      "num_tokens": 71087408.0,
      "step": 32440
    },
    {
      "entropy": 5.665345525741577,
      "epoch": 3.2200277888050812,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039839535908339297,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18649232536554336,
      "num_tokens": 71097474.0,
      "step": 32445
    },
    {
      "entropy": 5.60338773727417,
      "epoch": 3.2205240174672487,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039836572971796045,
      "loss": 4.9479,
      "mean_token_accuracy": 0.1952408328652382,
      "num_tokens": 71107064.0,
      "step": 32450
    },
    {
      "entropy": 5.639913129806518,
      "epoch": 3.2210202461294166,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003983360972933239,
      "loss": 4.9999,
      "mean_token_accuracy": 0.18483529090881348,
      "num_tokens": 71118233.0,
      "step": 32455
    },
    {
      "entropy": 5.647645092010498,
      "epoch": 3.221516474791584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003983064618102177,
      "loss": 4.9086,
      "mean_token_accuracy": 0.19595470130443574,
      "num_tokens": 71129922.0,
      "step": 32460
    },
    {
      "entropy": 5.671589326858521,
      "epoch": 3.2220127034537516,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003982768232693769,
      "loss": 5.0104,
      "mean_token_accuracy": 0.19053741693496704,
      "num_tokens": 71139073.0,
      "step": 32465
    },
    {
      "entropy": 5.625487518310547,
      "epoch": 3.222508932115919,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039824718167153666,
      "loss": 4.9073,
      "mean_token_accuracy": 0.19347963780164718,
      "num_tokens": 71152240.0,
      "step": 32470
    },
    {
      "entropy": 5.655756092071533,
      "epoch": 3.2230051607780865,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003982175370174319,
      "loss": 4.9838,
      "mean_token_accuracy": 0.19017771780490875,
      "num_tokens": 71163001.0,
      "step": 32475
    },
    {
      "entropy": 5.628947305679321,
      "epoch": 3.223501389440254,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003981878893077976,
      "loss": 4.9686,
      "mean_token_accuracy": 0.19098616689443587,
      "num_tokens": 71172811.0,
      "step": 32480
    },
    {
      "entropy": 5.6264650344848635,
      "epoch": 3.2239976181024215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039815823854336916,
      "loss": 4.8899,
      "mean_token_accuracy": 0.20509113073349,
      "num_tokens": 71182276.0,
      "step": 32485
    },
    {
      "entropy": 5.623444223403931,
      "epoch": 3.224493846764589,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003981285847248818,
      "loss": 4.8826,
      "mean_token_accuracy": 0.19744085818529128,
      "num_tokens": 71193154.0,
      "step": 32490
    },
    {
      "entropy": 5.687218713760376,
      "epoch": 3.2249900754267564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003980989278530709,
      "loss": 5.0519,
      "mean_token_accuracy": 0.18248777985572814,
      "num_tokens": 71204394.0,
      "step": 32495
    },
    {
      "entropy": 5.632289218902588,
      "epoch": 3.2254863040889243,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003980692679286718,
      "loss": 4.8778,
      "mean_token_accuracy": 0.19721261262893677,
      "num_tokens": 71215311.0,
      "step": 32500
    },
    {
      "entropy": 5.652625274658203,
      "epoch": 3.225982532751092,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039803960495241996,
      "loss": 4.9899,
      "mean_token_accuracy": 0.19553496241569518,
      "num_tokens": 71225800.0,
      "step": 32505
    },
    {
      "entropy": 5.649616527557373,
      "epoch": 3.2264787614132593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003980099389250511,
      "loss": 4.9684,
      "mean_token_accuracy": 0.18565160185098648,
      "num_tokens": 71237846.0,
      "step": 32510
    },
    {
      "entropy": 5.641085386276245,
      "epoch": 3.2269749900754268,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039798026984730084,
      "loss": 4.9677,
      "mean_token_accuracy": 0.19153966754674911,
      "num_tokens": 71248221.0,
      "step": 32515
    },
    {
      "entropy": 5.651960849761963,
      "epoch": 3.2274712187375942,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003979505977199049,
      "loss": 5.0,
      "mean_token_accuracy": 0.1866389259696007,
      "num_tokens": 71259950.0,
      "step": 32520
    },
    {
      "entropy": 5.674655199050903,
      "epoch": 3.2279674473997617,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039792092254359903,
      "loss": 5.0055,
      "mean_token_accuracy": 0.18776329159736632,
      "num_tokens": 71270541.0,
      "step": 32525
    },
    {
      "entropy": 5.66222071647644,
      "epoch": 3.228463676061929,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00039789124431911923,
      "loss": 4.9803,
      "mean_token_accuracy": 0.18548719584941864,
      "num_tokens": 71280545.0,
      "step": 32530
    },
    {
      "entropy": 5.6910241603851315,
      "epoch": 3.2289599047240967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039786156304720126,
      "loss": 5.0675,
      "mean_token_accuracy": 0.1803793802857399,
      "num_tokens": 71292368.0,
      "step": 32535
    },
    {
      "entropy": 5.693319082260132,
      "epoch": 3.2294561333862646,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003978318787285814,
      "loss": 4.9783,
      "mean_token_accuracy": 0.18620551973581315,
      "num_tokens": 71302570.0,
      "step": 32540
    },
    {
      "entropy": 5.695431089401245,
      "epoch": 3.229952362048432,
      "grad_norm": 1.28125,
      "learning_rate": 0.00039780219136399556,
      "loss": 4.9833,
      "mean_token_accuracy": 0.19014817774295806,
      "num_tokens": 71312629.0,
      "step": 32545
    },
    {
      "entropy": 5.6999836444854735,
      "epoch": 3.2304485907105995,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039777250095418,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18325173556804658,
      "num_tokens": 71324313.0,
      "step": 32550
    },
    {
      "entropy": 5.687304210662842,
      "epoch": 3.230944819372767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003977428074998709,
      "loss": 5.054,
      "mean_token_accuracy": 0.19116578847169877,
      "num_tokens": 71336136.0,
      "step": 32555
    },
    {
      "entropy": 5.617856073379516,
      "epoch": 3.2314410480349345,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039771311100180476,
      "loss": 4.9366,
      "mean_token_accuracy": 0.19444328248500825,
      "num_tokens": 71347267.0,
      "step": 32560
    },
    {
      "entropy": 5.596669483184814,
      "epoch": 3.231937276697102,
      "grad_norm": 1.40625,
      "learning_rate": 0.00039768341146071784,
      "loss": 4.9616,
      "mean_token_accuracy": 0.19537723511457444,
      "num_tokens": 71357268.0,
      "step": 32565
    },
    {
      "entropy": 5.732271003723144,
      "epoch": 3.2324335053592694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003976537088773467,
      "loss": 5.0803,
      "mean_token_accuracy": 0.18247534334659576,
      "num_tokens": 71368365.0,
      "step": 32570
    },
    {
      "entropy": 5.771731424331665,
      "epoch": 3.232929734021437,
      "grad_norm": 1.125,
      "learning_rate": 0.00039762400325242785,
      "loss": 5.0761,
      "mean_token_accuracy": 0.1826181009411812,
      "num_tokens": 71379623.0,
      "step": 32575
    },
    {
      "entropy": 5.642438650131226,
      "epoch": 3.2334259626836044,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039759429458669805,
      "loss": 4.9865,
      "mean_token_accuracy": 0.1941439613699913,
      "num_tokens": 71388522.0,
      "step": 32580
    },
    {
      "entropy": 5.590571594238281,
      "epoch": 3.2339221913457723,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039756458288089377,
      "loss": 4.957,
      "mean_token_accuracy": 0.19378824383020402,
      "num_tokens": 71399127.0,
      "step": 32585
    },
    {
      "entropy": 5.749866008758545,
      "epoch": 3.2344184200079398,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000397534868135752,
      "loss": 5.0773,
      "mean_token_accuracy": 0.18580615669488906,
      "num_tokens": 71410780.0,
      "step": 32590
    },
    {
      "entropy": 5.7409158706665036,
      "epoch": 3.2349146486701073,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039750515035200947,
      "loss": 5.0964,
      "mean_token_accuracy": 0.17680185288190842,
      "num_tokens": 71422887.0,
      "step": 32595
    },
    {
      "entropy": 5.670735120773315,
      "epoch": 3.2354108773322747,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039747542953040327,
      "loss": 5.0053,
      "mean_token_accuracy": 0.1882051035761833,
      "num_tokens": 71432525.0,
      "step": 32600
    },
    {
      "entropy": 5.638076591491699,
      "epoch": 3.235907105994442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003974457056716703,
      "loss": 5.0136,
      "mean_token_accuracy": 0.19019887447357178,
      "num_tokens": 71444802.0,
      "step": 32605
    },
    {
      "entropy": 5.631924295425415,
      "epoch": 3.2364033346566097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003974159787765477,
      "loss": 4.9244,
      "mean_token_accuracy": 0.19118320494890212,
      "num_tokens": 71455978.0,
      "step": 32610
    },
    {
      "entropy": 5.595856809616089,
      "epoch": 3.236899563318777,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003973862488457726,
      "loss": 4.8631,
      "mean_token_accuracy": 0.1932399407029152,
      "num_tokens": 71466981.0,
      "step": 32615
    },
    {
      "entropy": 5.679254627227783,
      "epoch": 3.2373957919809446,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003973565158800822,
      "loss": 5.0442,
      "mean_token_accuracy": 0.181380732357502,
      "num_tokens": 71478856.0,
      "step": 32620
    },
    {
      "entropy": 5.724590396881103,
      "epoch": 3.2378920206431125,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039732677988021394,
      "loss": 5.066,
      "mean_token_accuracy": 0.18099587559700012,
      "num_tokens": 71489449.0,
      "step": 32625
    },
    {
      "entropy": 5.698863744735718,
      "epoch": 3.23838824930528,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003972970408469051,
      "loss": 5.0425,
      "mean_token_accuracy": 0.18169956356287004,
      "num_tokens": 71499687.0,
      "step": 32630
    },
    {
      "entropy": 5.697513198852539,
      "epoch": 3.2388844779674475,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003972672987808932,
      "loss": 4.9861,
      "mean_token_accuracy": 0.19333878457546233,
      "num_tokens": 71510087.0,
      "step": 32635
    },
    {
      "entropy": 5.680063199996948,
      "epoch": 3.239380706629615,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039723755368291564,
      "loss": 4.9743,
      "mean_token_accuracy": 0.18853683173656463,
      "num_tokens": 71520504.0,
      "step": 32640
    },
    {
      "entropy": 5.727236747741699,
      "epoch": 3.2398769352917824,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003972078055537102,
      "loss": 5.0493,
      "mean_token_accuracy": 0.18118114322423934,
      "num_tokens": 71532342.0,
      "step": 32645
    },
    {
      "entropy": 5.729109764099121,
      "epoch": 3.24037316395395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003971780543940146,
      "loss": 5.012,
      "mean_token_accuracy": 0.1797688752412796,
      "num_tokens": 71544523.0,
      "step": 32650
    },
    {
      "entropy": 5.655762004852295,
      "epoch": 3.2408693926161174,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003971483002045665,
      "loss": 4.9961,
      "mean_token_accuracy": 0.18901816308498381,
      "num_tokens": 71555162.0,
      "step": 32655
    },
    {
      "entropy": 5.669670677185058,
      "epoch": 3.241365621278285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003971185429861037,
      "loss": 4.9714,
      "mean_token_accuracy": 0.19419633448123932,
      "num_tokens": 71566407.0,
      "step": 32660
    },
    {
      "entropy": 5.708176469802856,
      "epoch": 3.2418618499404523,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039708878273936424,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18960823863744736,
      "num_tokens": 71577328.0,
      "step": 32665
    },
    {
      "entropy": 5.7140788555145265,
      "epoch": 3.2423580786026203,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003970590194650859,
      "loss": 5.1014,
      "mean_token_accuracy": 0.17881713807582855,
      "num_tokens": 71587716.0,
      "step": 32670
    },
    {
      "entropy": 5.6558925151824955,
      "epoch": 3.2428543072647877,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003970292531640071,
      "loss": 4.9713,
      "mean_token_accuracy": 0.1876208260655403,
      "num_tokens": 71599575.0,
      "step": 32675
    },
    {
      "entropy": 5.692007112503052,
      "epoch": 3.243350535926955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039699948383686563,
      "loss": 5.0386,
      "mean_token_accuracy": 0.185214065015316,
      "num_tokens": 71610925.0,
      "step": 32680
    },
    {
      "entropy": 5.674749851226807,
      "epoch": 3.2438467645891227,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003969697114843998,
      "loss": 4.9911,
      "mean_token_accuracy": 0.18792439252138138,
      "num_tokens": 71620250.0,
      "step": 32685
    },
    {
      "entropy": 5.7173058032989506,
      "epoch": 3.24434299325129,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039693993610734807,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1830026000738144,
      "num_tokens": 71632633.0,
      "step": 32690
    },
    {
      "entropy": 5.711906290054321,
      "epoch": 3.2448392219134576,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039691015770644875,
      "loss": 5.0573,
      "mean_token_accuracy": 0.18414560556411744,
      "num_tokens": 71644031.0,
      "step": 32695
    },
    {
      "entropy": 5.7155050277709964,
      "epoch": 3.245335450575625,
      "grad_norm": 1.296875,
      "learning_rate": 0.00039688037628244,
      "loss": 5.0528,
      "mean_token_accuracy": 0.1946704775094986,
      "num_tokens": 71654785.0,
      "step": 32700
    },
    {
      "entropy": 5.670738554000854,
      "epoch": 3.2458316792377926,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039685059183606065,
      "loss": 4.9735,
      "mean_token_accuracy": 0.1919000640511513,
      "num_tokens": 71667014.0,
      "step": 32705
    },
    {
      "entropy": 5.69461727142334,
      "epoch": 3.2463279078999605,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039682080436804926,
      "loss": 5.0092,
      "mean_token_accuracy": 0.19617632180452346,
      "num_tokens": 71677157.0,
      "step": 32710
    },
    {
      "entropy": 5.732504796981812,
      "epoch": 3.246824136562128,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003967910138791442,
      "loss": 5.0545,
      "mean_token_accuracy": 0.18115280866622924,
      "num_tokens": 71688542.0,
      "step": 32715
    },
    {
      "entropy": 5.613659048080445,
      "epoch": 3.2473203652242955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039676122037008473,
      "loss": 4.9937,
      "mean_token_accuracy": 0.18317411094903946,
      "num_tokens": 71699364.0,
      "step": 32720
    },
    {
      "entropy": 5.705100679397583,
      "epoch": 3.247816593886463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003967314238416092,
      "loss": 5.0269,
      "mean_token_accuracy": 0.182632240653038,
      "num_tokens": 71710248.0,
      "step": 32725
    },
    {
      "entropy": 5.710696458816528,
      "epoch": 3.2483128225486304,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003967016242944567,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1830674797296524,
      "num_tokens": 71720616.0,
      "step": 32730
    },
    {
      "entropy": 5.7477912425994875,
      "epoch": 3.248809051210798,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003966718217293661,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18919406086206436,
      "num_tokens": 71731536.0,
      "step": 32735
    },
    {
      "entropy": 5.712781953811645,
      "epoch": 3.2493052798729654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039664201614707657,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18635061532258987,
      "num_tokens": 71742459.0,
      "step": 32740
    },
    {
      "entropy": 5.5710631847381595,
      "epoch": 3.249801508535133,
      "grad_norm": 1.125,
      "learning_rate": 0.0003966122075483271,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18207916170358657,
      "num_tokens": 71753611.0,
      "step": 32745
    },
    {
      "entropy": 5.64806981086731,
      "epoch": 3.2502977371973003,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039658239593385697,
      "loss": 4.9626,
      "mean_token_accuracy": 0.18968242555856704,
      "num_tokens": 71764924.0,
      "step": 32750
    },
    {
      "entropy": 5.68782639503479,
      "epoch": 3.2507939658594682,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039655258130440543,
      "loss": 5.0509,
      "mean_token_accuracy": 0.17505000084638594,
      "num_tokens": 71776078.0,
      "step": 32755
    },
    {
      "entropy": 5.7178308963775635,
      "epoch": 3.2512901945216357,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003965227636607117,
      "loss": 4.9827,
      "mean_token_accuracy": 0.18659722656011582,
      "num_tokens": 71785988.0,
      "step": 32760
    },
    {
      "entropy": 5.6446452140808105,
      "epoch": 3.251786423183803,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00039649294300351547,
      "loss": 4.9844,
      "mean_token_accuracy": 0.1883395493030548,
      "num_tokens": 71796382.0,
      "step": 32765
    },
    {
      "entropy": 5.620907735824585,
      "epoch": 3.2522826518459707,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003964631193335559,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18805274069309236,
      "num_tokens": 71806533.0,
      "step": 32770
    },
    {
      "entropy": 5.6400984764099125,
      "epoch": 3.252778880508138,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039643329265157275,
      "loss": 5.0042,
      "mean_token_accuracy": 0.1894126355648041,
      "num_tokens": 71816893.0,
      "step": 32775
    },
    {
      "entropy": 5.704699325561523,
      "epoch": 3.2532751091703056,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039640346295830557,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18782349973917006,
      "num_tokens": 71827366.0,
      "step": 32780
    },
    {
      "entropy": 5.7531708717346195,
      "epoch": 3.253771337832473,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003963736302544942,
      "loss": 5.0367,
      "mean_token_accuracy": 0.18011904805898665,
      "num_tokens": 71837896.0,
      "step": 32785
    },
    {
      "entropy": 5.666146564483642,
      "epoch": 3.2542675664946406,
      "grad_norm": 1.125,
      "learning_rate": 0.00039634379454087814,
      "loss": 5.0128,
      "mean_token_accuracy": 0.1810763955116272,
      "num_tokens": 71850583.0,
      "step": 32790
    },
    {
      "entropy": 5.6443578720092775,
      "epoch": 3.2547637951568085,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003963139558181977,
      "loss": 5.0349,
      "mean_token_accuracy": 0.18930251747369767,
      "num_tokens": 71860392.0,
      "step": 32795
    },
    {
      "entropy": 5.663821125030518,
      "epoch": 3.255260023818976,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003962841140871924,
      "loss": 5.0062,
      "mean_token_accuracy": 0.18795392364263536,
      "num_tokens": 71871323.0,
      "step": 32800
    },
    {
      "entropy": 5.66532940864563,
      "epoch": 3.2557562524811434,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003962542693486024,
      "loss": 4.9541,
      "mean_token_accuracy": 0.19078368693590164,
      "num_tokens": 71881612.0,
      "step": 32805
    },
    {
      "entropy": 5.686474132537842,
      "epoch": 3.256252481143311,
      "grad_norm": 1.125,
      "learning_rate": 0.0003962244216031679,
      "loss": 5.0784,
      "mean_token_accuracy": 0.18116381168365478,
      "num_tokens": 71893266.0,
      "step": 32810
    },
    {
      "entropy": 5.649704217910767,
      "epoch": 3.2567487098054784,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000396194570851629,
      "loss": 4.9747,
      "mean_token_accuracy": 0.19299719482660294,
      "num_tokens": 71904978.0,
      "step": 32815
    },
    {
      "entropy": 5.7225175380706785,
      "epoch": 3.257244938467646,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039616471709472573,
      "loss": 5.0531,
      "mean_token_accuracy": 0.17888737022876738,
      "num_tokens": 71915097.0,
      "step": 32820
    },
    {
      "entropy": 5.728493356704712,
      "epoch": 3.2577411671298133,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003961348603331987,
      "loss": 4.9947,
      "mean_token_accuracy": 0.18224958777427674,
      "num_tokens": 71924672.0,
      "step": 32825
    },
    {
      "entropy": 5.595474433898926,
      "epoch": 3.258237395791981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039610500056778817,
      "loss": 4.9156,
      "mean_token_accuracy": 0.18842033594846724,
      "num_tokens": 71936122.0,
      "step": 32830
    },
    {
      "entropy": 5.614183855056763,
      "epoch": 3.2587336244541483,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003960751377992344,
      "loss": 4.9216,
      "mean_token_accuracy": 0.19660517126321791,
      "num_tokens": 71946601.0,
      "step": 32835
    },
    {
      "entropy": 5.714119720458984,
      "epoch": 3.259229853116316,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039604527202827825,
      "loss": 5.0607,
      "mean_token_accuracy": 0.18213890194892884,
      "num_tokens": 71958892.0,
      "step": 32840
    },
    {
      "entropy": 5.6638421535491945,
      "epoch": 3.2597260817784837,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039601540325566014,
      "loss": 4.9607,
      "mean_token_accuracy": 0.18937372863292695,
      "num_tokens": 71969183.0,
      "step": 32845
    },
    {
      "entropy": 5.685330963134765,
      "epoch": 3.260222310440651,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003959855314821208,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18937092870473862,
      "num_tokens": 71981115.0,
      "step": 32850
    },
    {
      "entropy": 5.6879314422607425,
      "epoch": 3.2607185391028186,
      "grad_norm": 1.1875,
      "learning_rate": 0.000395955656708401,
      "loss": 5.0472,
      "mean_token_accuracy": 0.189069864153862,
      "num_tokens": 71991677.0,
      "step": 32855
    },
    {
      "entropy": 5.687830066680908,
      "epoch": 3.261214767764986,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039592577893524155,
      "loss": 5.023,
      "mean_token_accuracy": 0.1923421651124954,
      "num_tokens": 72003995.0,
      "step": 32860
    },
    {
      "entropy": 5.70494589805603,
      "epoch": 3.2617109964271536,
      "grad_norm": 1.34375,
      "learning_rate": 0.00039589589816338335,
      "loss": 4.999,
      "mean_token_accuracy": 0.18800686597824096,
      "num_tokens": 72014052.0,
      "step": 32865
    },
    {
      "entropy": 5.6557207107543945,
      "epoch": 3.262207225089321,
      "grad_norm": 1.125,
      "learning_rate": 0.0003958660143935674,
      "loss": 4.9513,
      "mean_token_accuracy": 0.1854252487421036,
      "num_tokens": 72024756.0,
      "step": 32870
    },
    {
      "entropy": 5.592964935302734,
      "epoch": 3.2627034537514885,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003958361276265347,
      "loss": 4.9686,
      "mean_token_accuracy": 0.19074191749095917,
      "num_tokens": 72035883.0,
      "step": 32875
    },
    {
      "entropy": 5.6469048976898195,
      "epoch": 3.2631996824136564,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039580623786302643,
      "loss": 4.9715,
      "mean_token_accuracy": 0.1889067620038986,
      "num_tokens": 72047296.0,
      "step": 32880
    },
    {
      "entropy": 5.670968532562256,
      "epoch": 3.263695911075824,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039577634510378374,
      "loss": 4.9936,
      "mean_token_accuracy": 0.18295124918222427,
      "num_tokens": 72060163.0,
      "step": 32885
    },
    {
      "entropy": 5.704800081253052,
      "epoch": 3.2641921397379914,
      "grad_norm": 1.171875,
      "learning_rate": 0.000395746449349548,
      "loss": 5.0559,
      "mean_token_accuracy": 0.18217405676841736,
      "num_tokens": 72070384.0,
      "step": 32890
    },
    {
      "entropy": 5.678947114944458,
      "epoch": 3.264688368400159,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003957165506010605,
      "loss": 4.9594,
      "mean_token_accuracy": 0.18247588574886323,
      "num_tokens": 72080997.0,
      "step": 32895
    },
    {
      "entropy": 5.720320320129394,
      "epoch": 3.2651845970623263,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003956866488590626,
      "loss": 5.0023,
      "mean_token_accuracy": 0.19282065480947494,
      "num_tokens": 72091161.0,
      "step": 32900
    },
    {
      "entropy": 5.644020080566406,
      "epoch": 3.265680825724494,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003956567441242959,
      "loss": 4.9876,
      "mean_token_accuracy": 0.1876245453953743,
      "num_tokens": 72101807.0,
      "step": 32905
    },
    {
      "entropy": 5.665662145614624,
      "epoch": 3.2661770543866613,
      "grad_norm": 1.09375,
      "learning_rate": 0.000395626836397502,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18390080630779265,
      "num_tokens": 72112800.0,
      "step": 32910
    },
    {
      "entropy": 5.630118274688721,
      "epoch": 3.2666732830488288,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003955969256794224,
      "loss": 4.9749,
      "mean_token_accuracy": 0.18705268055200577,
      "num_tokens": 72122391.0,
      "step": 32915
    },
    {
      "entropy": 5.642094373703003,
      "epoch": 3.2671695117109962,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039556701197079896,
      "loss": 5.0193,
      "mean_token_accuracy": 0.1873021885752678,
      "num_tokens": 72133145.0,
      "step": 32920
    },
    {
      "entropy": 5.7992809295654295,
      "epoch": 3.2676657403731637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039553709527237344,
      "loss": 5.0338,
      "mean_token_accuracy": 0.18160320073366165,
      "num_tokens": 72146126.0,
      "step": 32925
    },
    {
      "entropy": 5.690984153747559,
      "epoch": 3.2681619690353316,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003955071755848876,
      "loss": 4.9545,
      "mean_token_accuracy": 0.19437672942876816,
      "num_tokens": 72157325.0,
      "step": 32930
    },
    {
      "entropy": 5.591590833663941,
      "epoch": 3.268658197697499,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039547725290908347,
      "loss": 4.8704,
      "mean_token_accuracy": 0.19238672107458116,
      "num_tokens": 72168601.0,
      "step": 32935
    },
    {
      "entropy": 5.624763441085816,
      "epoch": 3.2691544263596666,
      "grad_norm": 2.296875,
      "learning_rate": 0.00039544732724570313,
      "loss": 4.9447,
      "mean_token_accuracy": 0.19632379859685897,
      "num_tokens": 72179845.0,
      "step": 32940
    },
    {
      "entropy": 5.618891048431396,
      "epoch": 3.269650655021834,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003954173985954887,
      "loss": 4.9055,
      "mean_token_accuracy": 0.18850270807743072,
      "num_tokens": 72191135.0,
      "step": 32945
    },
    {
      "entropy": 5.651686286926269,
      "epoch": 3.2701468836840015,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003953874669591822,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18939662724733353,
      "num_tokens": 72201551.0,
      "step": 32950
    },
    {
      "entropy": 5.660107946395874,
      "epoch": 3.270643112346169,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039535753233752583,
      "loss": 4.9782,
      "mean_token_accuracy": 0.1875036135315895,
      "num_tokens": 72211983.0,
      "step": 32955
    },
    {
      "entropy": 5.7044360637664795,
      "epoch": 3.2711393410083365,
      "grad_norm": 1.25,
      "learning_rate": 0.0003953275947312621,
      "loss": 5.0182,
      "mean_token_accuracy": 0.1850027471780777,
      "num_tokens": 72223374.0,
      "step": 32960
    },
    {
      "entropy": 5.611225032806397,
      "epoch": 3.2716355696705044,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003952976541411332,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19269566833972931,
      "num_tokens": 72233878.0,
      "step": 32965
    },
    {
      "entropy": 5.6792549133300785,
      "epoch": 3.272131798332672,
      "grad_norm": 1.125,
      "learning_rate": 0.00039526771056788186,
      "loss": 5.0588,
      "mean_token_accuracy": 0.18488863855600357,
      "num_tokens": 72246157.0,
      "step": 32970
    },
    {
      "entropy": 5.661624240875244,
      "epoch": 3.2726280269948393,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039523776401225043,
      "loss": 4.9774,
      "mean_token_accuracy": 0.19524696767330169,
      "num_tokens": 72256805.0,
      "step": 32975
    },
    {
      "entropy": 5.674283170700074,
      "epoch": 3.273124255657007,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039520781447498145,
      "loss": 5.011,
      "mean_token_accuracy": 0.17980253994464873,
      "num_tokens": 72267026.0,
      "step": 32980
    },
    {
      "entropy": 5.695543479919434,
      "epoch": 3.2736204843191743,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003951778619568177,
      "loss": 4.9741,
      "mean_token_accuracy": 0.1937365487217903,
      "num_tokens": 72277989.0,
      "step": 32985
    },
    {
      "entropy": 5.638600015640259,
      "epoch": 3.2741167129813418,
      "grad_norm": 1.28125,
      "learning_rate": 0.000395147906458502,
      "loss": 4.9218,
      "mean_token_accuracy": 0.1876099392771721,
      "num_tokens": 72287989.0,
      "step": 32990
    },
    {
      "entropy": 5.609688806533813,
      "epoch": 3.2746129416435092,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039511794798077705,
      "loss": 4.9586,
      "mean_token_accuracy": 0.1843106582760811,
      "num_tokens": 72298718.0,
      "step": 32995
    },
    {
      "entropy": 5.625388669967651,
      "epoch": 3.2751091703056767,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003950879865243858,
      "loss": 4.9481,
      "mean_token_accuracy": 0.19389324635267258,
      "num_tokens": 72309217.0,
      "step": 33000
    },
    {
      "epoch": 3.2751091703056767,
      "eval_entropy": 5.43529351019865,
      "eval_loss": 5.127163410186768,
      "eval_mean_token_accuracy": 0.18708836612393195,
      "eval_num_tokens": 72309217.0,
      "eval_runtime": 20.221,
      "eval_samples_per_second": 1665.055,
      "eval_steps_per_second": 208.15,
      "step": 33000
    },
    {
      "entropy": 5.695728445053101,
      "epoch": 3.275605398967844,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039505802209007136,
      "loss": 5.0073,
      "mean_token_accuracy": 0.18538760095834733,
      "num_tokens": 72319809.0,
      "step": 33005
    },
    {
      "entropy": 5.677643203735352,
      "epoch": 3.2761016276300117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003950280546785766,
      "loss": 5.0447,
      "mean_token_accuracy": 0.18457311391830444,
      "num_tokens": 72330616.0,
      "step": 33010
    },
    {
      "entropy": 5.6606217384338375,
      "epoch": 3.2765978562921796,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003949980842906447,
      "loss": 5.0261,
      "mean_token_accuracy": 0.1880222499370575,
      "num_tokens": 72339749.0,
      "step": 33015
    },
    {
      "entropy": 5.646131372451782,
      "epoch": 3.277094084954347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000394968110927019,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19137380570173262,
      "num_tokens": 72350653.0,
      "step": 33020
    },
    {
      "entropy": 5.673167943954468,
      "epoch": 3.2775903136165145,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039493813458844256,
      "loss": 4.9834,
      "mean_token_accuracy": 0.19059329181909562,
      "num_tokens": 72360261.0,
      "step": 33025
    },
    {
      "entropy": 5.612597703933716,
      "epoch": 3.278086542278682,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039490815527565883,
      "loss": 4.988,
      "mean_token_accuracy": 0.18597551435232162,
      "num_tokens": 72372180.0,
      "step": 33030
    },
    {
      "entropy": 5.624655818939209,
      "epoch": 3.2785827709408495,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039487817298941124,
      "loss": 4.9715,
      "mean_token_accuracy": 0.17984236627817154,
      "num_tokens": 72383324.0,
      "step": 33035
    },
    {
      "entropy": 5.648533153533935,
      "epoch": 3.279078999603017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003948481877304433,
      "loss": 4.9931,
      "mean_token_accuracy": 0.18422985672950745,
      "num_tokens": 72395014.0,
      "step": 33040
    },
    {
      "entropy": 5.724872398376465,
      "epoch": 3.2795752282651844,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003948181994994985,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18656937032938004,
      "num_tokens": 72406806.0,
      "step": 33045
    },
    {
      "entropy": 5.7127766609191895,
      "epoch": 3.2800714569273524,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003947882082973206,
      "loss": 4.9587,
      "mean_token_accuracy": 0.18602796345949174,
      "num_tokens": 72417002.0,
      "step": 33050
    },
    {
      "entropy": 5.651697444915771,
      "epoch": 3.28056768558952,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039475821412465317,
      "loss": 5.0216,
      "mean_token_accuracy": 0.18570436388254166,
      "num_tokens": 72427107.0,
      "step": 33055
    },
    {
      "entropy": 5.608659791946411,
      "epoch": 3.2810639142516873,
      "grad_norm": 1.25,
      "learning_rate": 0.00039472821698224023,
      "loss": 4.9022,
      "mean_token_accuracy": 0.20143126398324968,
      "num_tokens": 72438737.0,
      "step": 33060
    },
    {
      "entropy": 5.624824953079224,
      "epoch": 3.281560142913855,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003946982168708254,
      "loss": 4.9241,
      "mean_token_accuracy": 0.1869140088558197,
      "num_tokens": 72449070.0,
      "step": 33065
    },
    {
      "entropy": 5.670595026016235,
      "epoch": 3.2820563715760223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003946682137911527,
      "loss": 5.0253,
      "mean_token_accuracy": 0.18619106709957123,
      "num_tokens": 72459437.0,
      "step": 33070
    },
    {
      "entropy": 5.701854133605957,
      "epoch": 3.2825526002381897,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039463820774396623,
      "loss": 4.998,
      "mean_token_accuracy": 0.18337791562080383,
      "num_tokens": 72470804.0,
      "step": 33075
    },
    {
      "entropy": 5.703625583648682,
      "epoch": 3.283048828900357,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039460819873000995,
      "loss": 5.0099,
      "mean_token_accuracy": 0.18225862681865693,
      "num_tokens": 72482400.0,
      "step": 33080
    },
    {
      "entropy": 5.605824899673462,
      "epoch": 3.2835450575625247,
      "grad_norm": 1.125,
      "learning_rate": 0.00039457818675002807,
      "loss": 4.8914,
      "mean_token_accuracy": 0.19455617517232895,
      "num_tokens": 72492697.0,
      "step": 33085
    },
    {
      "entropy": 5.6105828285217285,
      "epoch": 3.284041286224692,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003945481718047648,
      "loss": 4.9833,
      "mean_token_accuracy": 0.18893595337867736,
      "num_tokens": 72503913.0,
      "step": 33090
    },
    {
      "entropy": 5.710871458053589,
      "epoch": 3.2845375148868596,
      "grad_norm": 1.125,
      "learning_rate": 0.0003945181538949646,
      "loss": 4.987,
      "mean_token_accuracy": 0.1895302265882492,
      "num_tokens": 72515018.0,
      "step": 33095
    },
    {
      "entropy": 5.661000776290893,
      "epoch": 3.2850337435490276,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003944881330213716,
      "loss": 5.0196,
      "mean_token_accuracy": 0.18421826064586638,
      "num_tokens": 72525256.0,
      "step": 33100
    },
    {
      "entropy": 5.608473682403565,
      "epoch": 3.285529972211195,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003944581091847304,
      "loss": 4.9396,
      "mean_token_accuracy": 0.1908977970480919,
      "num_tokens": 72535839.0,
      "step": 33105
    },
    {
      "entropy": 5.674889421463012,
      "epoch": 3.2860262008733625,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039442808238578553,
      "loss": 5.0542,
      "mean_token_accuracy": 0.191722509264946,
      "num_tokens": 72546808.0,
      "step": 33110
    },
    {
      "entropy": 5.724297618865966,
      "epoch": 3.28652242953553,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039439805262528153,
      "loss": 5.0468,
      "mean_token_accuracy": 0.18804130256175994,
      "num_tokens": 72559268.0,
      "step": 33115
    },
    {
      "entropy": 5.7318198680877686,
      "epoch": 3.2870186581976975,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003943680199039631,
      "loss": 5.0146,
      "mean_token_accuracy": 0.18143411576747895,
      "num_tokens": 72570466.0,
      "step": 33120
    },
    {
      "entropy": 5.692963790893555,
      "epoch": 3.287514886859865,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039433798422257505,
      "loss": 4.9551,
      "mean_token_accuracy": 0.18811022788286208,
      "num_tokens": 72581373.0,
      "step": 33125
    },
    {
      "entropy": 5.748071384429932,
      "epoch": 3.2880111155220324,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003943079455818621,
      "loss": 5.0663,
      "mean_token_accuracy": 0.18263172656297683,
      "num_tokens": 72592661.0,
      "step": 33130
    },
    {
      "entropy": 5.5950616836547855,
      "epoch": 3.2885073441842003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039427790398256925,
      "loss": 4.9273,
      "mean_token_accuracy": 0.19563668221235275,
      "num_tokens": 72603387.0,
      "step": 33135
    },
    {
      "entropy": 5.6417638778686525,
      "epoch": 3.289003572846368,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039424785942544126,
      "loss": 4.9701,
      "mean_token_accuracy": 0.18864423036575317,
      "num_tokens": 72614006.0,
      "step": 33140
    },
    {
      "entropy": 5.614690971374512,
      "epoch": 3.2894998015085353,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003942178119112234,
      "loss": 4.9527,
      "mean_token_accuracy": 0.1893549606204033,
      "num_tokens": 72625447.0,
      "step": 33145
    },
    {
      "entropy": 5.622698163986206,
      "epoch": 3.2899960301707027,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003941877614406607,
      "loss": 4.9631,
      "mean_token_accuracy": 0.18840062767267227,
      "num_tokens": 72636965.0,
      "step": 33150
    },
    {
      "entropy": 5.680086755752564,
      "epoch": 3.29049225883287,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003941577080144984,
      "loss": 4.9504,
      "mean_token_accuracy": 0.19371943026781083,
      "num_tokens": 72648161.0,
      "step": 33155
    },
    {
      "entropy": 5.669582366943359,
      "epoch": 3.2909884874950377,
      "grad_norm": 1.125,
      "learning_rate": 0.0003941276516334816,
      "loss": 4.9306,
      "mean_token_accuracy": 0.19315018951892854,
      "num_tokens": 72658253.0,
      "step": 33160
    },
    {
      "entropy": 5.604725790023804,
      "epoch": 3.291484716157205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039409759229835586,
      "loss": 4.939,
      "mean_token_accuracy": 0.19313877075910568,
      "num_tokens": 72669583.0,
      "step": 33165
    },
    {
      "entropy": 5.573974752426148,
      "epoch": 3.2919809448193726,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003940675300098663,
      "loss": 4.8937,
      "mean_token_accuracy": 0.1906040996313095,
      "num_tokens": 72680367.0,
      "step": 33170
    },
    {
      "entropy": 5.634705400466919,
      "epoch": 3.29247717348154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003940374647687586,
      "loss": 5.0014,
      "mean_token_accuracy": 0.18758548349142073,
      "num_tokens": 72692437.0,
      "step": 33175
    },
    {
      "entropy": 5.64965591430664,
      "epoch": 3.2929734021437076,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003940073965757784,
      "loss": 5.0258,
      "mean_token_accuracy": 0.18509383648633956,
      "num_tokens": 72704286.0,
      "step": 33180
    },
    {
      "entropy": 5.7177375793457035,
      "epoch": 3.2934696308058755,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003939773254316711,
      "loss": 5.053,
      "mean_token_accuracy": 0.18389751613140107,
      "num_tokens": 72714767.0,
      "step": 33185
    },
    {
      "entropy": 5.715953683853149,
      "epoch": 3.293965859468043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003939472513371826,
      "loss": 5.0498,
      "mean_token_accuracy": 0.1854212313890457,
      "num_tokens": 72725110.0,
      "step": 33190
    },
    {
      "entropy": 5.6588630199432375,
      "epoch": 3.2944620881302105,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003939171742930584,
      "loss": 4.9871,
      "mean_token_accuracy": 0.19068311303853988,
      "num_tokens": 72737150.0,
      "step": 33195
    },
    {
      "entropy": 5.695828437805176,
      "epoch": 3.294958316792378,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039388709430004465,
      "loss": 4.953,
      "mean_token_accuracy": 0.19124603271484375,
      "num_tokens": 72747523.0,
      "step": 33200
    },
    {
      "entropy": 5.699077177047729,
      "epoch": 3.2954545454545454,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039385701135888715,
      "loss": 5.0526,
      "mean_token_accuracy": 0.192521134018898,
      "num_tokens": 72760163.0,
      "step": 33205
    },
    {
      "entropy": 5.728883457183838,
      "epoch": 3.295950774116713,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003938269254703318,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18178825527429582,
      "num_tokens": 72770031.0,
      "step": 33210
    },
    {
      "entropy": 5.697427892684937,
      "epoch": 3.2964470027788804,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039379683663512475,
      "loss": 4.9117,
      "mean_token_accuracy": 0.1945084735751152,
      "num_tokens": 72780187.0,
      "step": 33215
    },
    {
      "entropy": 5.638015413284302,
      "epoch": 3.2969432314410483,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003937667448540123,
      "loss": 5.0141,
      "mean_token_accuracy": 0.18610692024230957,
      "num_tokens": 72790486.0,
      "step": 33220
    },
    {
      "entropy": 5.580369091033935,
      "epoch": 3.2974394601032158,
      "grad_norm": 1.125,
      "learning_rate": 0.00039373665012774027,
      "loss": 4.8696,
      "mean_token_accuracy": 0.1947270005941391,
      "num_tokens": 72800086.0,
      "step": 33225
    },
    {
      "entropy": 5.700982618331909,
      "epoch": 3.2979356887653832,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039370655245705527,
      "loss": 5.082,
      "mean_token_accuracy": 0.18274524211883544,
      "num_tokens": 72811018.0,
      "step": 33230
    },
    {
      "entropy": 5.668225717544556,
      "epoch": 3.2984319174275507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039367645184270364,
      "loss": 4.9721,
      "mean_token_accuracy": 0.19191589206457138,
      "num_tokens": 72821792.0,
      "step": 33235
    },
    {
      "entropy": 5.6958095073699955,
      "epoch": 3.298928146089718,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003936463482854317,
      "loss": 5.0868,
      "mean_token_accuracy": 0.1892849862575531,
      "num_tokens": 72833782.0,
      "step": 33240
    },
    {
      "entropy": 5.690759897232056,
      "epoch": 3.2994243747518857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003936162417859858,
      "loss": 4.9986,
      "mean_token_accuracy": 0.1930977448821068,
      "num_tokens": 72844480.0,
      "step": 33245
    },
    {
      "entropy": 5.733338356018066,
      "epoch": 3.299920603414053,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039358613234511286,
      "loss": 5.1327,
      "mean_token_accuracy": 0.18112690299749373,
      "num_tokens": 72855855.0,
      "step": 33250
    },
    {
      "entropy": 5.666411924362182,
      "epoch": 3.3004168320762206,
      "grad_norm": 1.125,
      "learning_rate": 0.00039355601996355944,
      "loss": 4.9644,
      "mean_token_accuracy": 0.18480539619922637,
      "num_tokens": 72866444.0,
      "step": 33255
    },
    {
      "entropy": 5.6650755405426025,
      "epoch": 3.300913060738388,
      "grad_norm": 1.203125,
      "learning_rate": 0.000393525904642072,
      "loss": 4.9945,
      "mean_token_accuracy": 0.18746765702962875,
      "num_tokens": 72877083.0,
      "step": 33260
    },
    {
      "entropy": 5.594808101654053,
      "epoch": 3.3014092894005556,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039349578638139766,
      "loss": 4.948,
      "mean_token_accuracy": 0.19275020211935043,
      "num_tokens": 72888183.0,
      "step": 33265
    },
    {
      "entropy": 5.747305583953858,
      "epoch": 3.3019055180627235,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003934656651822832,
      "loss": 5.1301,
      "mean_token_accuracy": 0.1781978890299797,
      "num_tokens": 72899171.0,
      "step": 33270
    },
    {
      "entropy": 5.687900161743164,
      "epoch": 3.302401746724891,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003934355410454754,
      "loss": 4.9849,
      "mean_token_accuracy": 0.18292107284069062,
      "num_tokens": 72909362.0,
      "step": 33275
    },
    {
      "entropy": 5.678633975982666,
      "epoch": 3.3028979753870584,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039340541397172137,
      "loss": 4.9648,
      "mean_token_accuracy": 0.19061854630708694,
      "num_tokens": 72920110.0,
      "step": 33280
    },
    {
      "entropy": 5.685652780532837,
      "epoch": 3.303394204049226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003933752839617682,
      "loss": 5.0172,
      "mean_token_accuracy": 0.18232827484607697,
      "num_tokens": 72931006.0,
      "step": 33285
    },
    {
      "entropy": 5.694718074798584,
      "epoch": 3.3038904327113934,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003933451510163632,
      "loss": 5.1127,
      "mean_token_accuracy": 0.17770932763814926,
      "num_tokens": 72941371.0,
      "step": 33290
    },
    {
      "entropy": 5.660941410064697,
      "epoch": 3.304386661373561,
      "grad_norm": 1.125,
      "learning_rate": 0.0003933150151362534,
      "loss": 4.9655,
      "mean_token_accuracy": 0.1897570699453354,
      "num_tokens": 72953044.0,
      "step": 33295
    },
    {
      "entropy": 5.667482376098633,
      "epoch": 3.3048828900357283,
      "grad_norm": 1.328125,
      "learning_rate": 0.00039328487632218613,
      "loss": 5.0052,
      "mean_token_accuracy": 0.1950204163789749,
      "num_tokens": 72963581.0,
      "step": 33300
    },
    {
      "entropy": 5.612351751327514,
      "epoch": 3.305379118697896,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003932547345749088,
      "loss": 4.9442,
      "mean_token_accuracy": 0.19285104423761368,
      "num_tokens": 72974153.0,
      "step": 33305
    },
    {
      "entropy": 5.67050952911377,
      "epoch": 3.3058753473600637,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039322458989516897,
      "loss": 4.9864,
      "mean_token_accuracy": 0.18997064232826233,
      "num_tokens": 72985071.0,
      "step": 33310
    },
    {
      "entropy": 5.663695049285889,
      "epoch": 3.306371576022231,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039319444228371397,
      "loss": 4.9826,
      "mean_token_accuracy": 0.19271438121795653,
      "num_tokens": 72995592.0,
      "step": 33315
    },
    {
      "entropy": 5.628513622283935,
      "epoch": 3.3068678046843987,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003931642917412916,
      "loss": 4.9191,
      "mean_token_accuracy": 0.19628411084413527,
      "num_tokens": 73005431.0,
      "step": 33320
    },
    {
      "entropy": 5.574351787567139,
      "epoch": 3.307364033346566,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003931341382686493,
      "loss": 4.8877,
      "mean_token_accuracy": 0.19998364597558976,
      "num_tokens": 73015212.0,
      "step": 33325
    },
    {
      "entropy": 5.651284170150757,
      "epoch": 3.3078602620087336,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00039310398186653497,
      "loss": 4.9561,
      "mean_token_accuracy": 0.18378959447145463,
      "num_tokens": 73026585.0,
      "step": 33330
    },
    {
      "entropy": 5.6966859817504885,
      "epoch": 3.308356490670901,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039307382253569637,
      "loss": 5.0427,
      "mean_token_accuracy": 0.1844810962677002,
      "num_tokens": 73036844.0,
      "step": 33335
    },
    {
      "entropy": 5.631027126312256,
      "epoch": 3.3088527193330686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003930436602768814,
      "loss": 4.9458,
      "mean_token_accuracy": 0.1899826243519783,
      "num_tokens": 73047881.0,
      "step": 33340
    },
    {
      "entropy": 5.63448748588562,
      "epoch": 3.309348947995236,
      "grad_norm": 1.171875,
      "learning_rate": 0.000393013495090838,
      "loss": 4.971,
      "mean_token_accuracy": 0.19012175351381302,
      "num_tokens": 73058541.0,
      "step": 33345
    },
    {
      "entropy": 5.587987661361694,
      "epoch": 3.3098451766574035,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003929833269783143,
      "loss": 4.8743,
      "mean_token_accuracy": 0.19985547810792922,
      "num_tokens": 73069208.0,
      "step": 33350
    },
    {
      "entropy": 5.6062867641448975,
      "epoch": 3.3103414053195714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003929531559400582,
      "loss": 4.9279,
      "mean_token_accuracy": 0.19250885546207427,
      "num_tokens": 73079581.0,
      "step": 33355
    },
    {
      "entropy": 5.630061197280884,
      "epoch": 3.310837633981739,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039292298197681803,
      "loss": 4.9525,
      "mean_token_accuracy": 0.1925985261797905,
      "num_tokens": 73090126.0,
      "step": 33360
    },
    {
      "entropy": 5.627453088760376,
      "epoch": 3.3113338626439064,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039289280508934203,
      "loss": 4.9515,
      "mean_token_accuracy": 0.1876961335539818,
      "num_tokens": 73101421.0,
      "step": 33365
    },
    {
      "entropy": 5.696677494049072,
      "epoch": 3.311830091306074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039286262527837856,
      "loss": 4.9479,
      "mean_token_accuracy": 0.19345304667949675,
      "num_tokens": 73111826.0,
      "step": 33370
    },
    {
      "entropy": 5.697140502929687,
      "epoch": 3.3123263199682413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003928324425446759,
      "loss": 4.9796,
      "mean_token_accuracy": 0.19258675277233123,
      "num_tokens": 73124195.0,
      "step": 33375
    },
    {
      "entropy": 5.603406667709351,
      "epoch": 3.312822548630409,
      "grad_norm": 1.171875,
      "learning_rate": 0.00039280225688898256,
      "loss": 4.9526,
      "mean_token_accuracy": 0.18963079452514647,
      "num_tokens": 73135590.0,
      "step": 33380
    },
    {
      "entropy": 5.6378966808319095,
      "epoch": 3.3133187772925763,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039277206831204715,
      "loss": 4.9119,
      "mean_token_accuracy": 0.18933289647102355,
      "num_tokens": 73145577.0,
      "step": 33385
    },
    {
      "entropy": 5.548543167114258,
      "epoch": 3.3138150059547438,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003927418768146182,
      "loss": 4.9215,
      "mean_token_accuracy": 0.20206085443496705,
      "num_tokens": 73156717.0,
      "step": 33390
    },
    {
      "entropy": 5.642420721054077,
      "epoch": 3.3143112346169117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039271168239744434,
      "loss": 4.9645,
      "mean_token_accuracy": 0.18650400638580322,
      "num_tokens": 73166534.0,
      "step": 33395
    },
    {
      "entropy": 5.6658549308776855,
      "epoch": 3.314807463279079,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003926814850612745,
      "loss": 4.9341,
      "mean_token_accuracy": 0.1967325523495674,
      "num_tokens": 73176182.0,
      "step": 33400
    },
    {
      "entropy": 5.690871715545654,
      "epoch": 3.3153036919412466,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003926512848068573,
      "loss": 5.0102,
      "mean_token_accuracy": 0.18435982018709182,
      "num_tokens": 73186837.0,
      "step": 33405
    },
    {
      "entropy": 5.625494956970215,
      "epoch": 3.315799920603414,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003926210816349418,
      "loss": 4.9409,
      "mean_token_accuracy": 0.19337667226791383,
      "num_tokens": 73197786.0,
      "step": 33410
    },
    {
      "entropy": 5.59560866355896,
      "epoch": 3.3162961492655816,
      "grad_norm": 1.171875,
      "learning_rate": 0.000392590875546277,
      "loss": 4.8953,
      "mean_token_accuracy": 0.19246049374341964,
      "num_tokens": 73208408.0,
      "step": 33415
    },
    {
      "entropy": 5.642316675186157,
      "epoch": 3.316792377927749,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039256066654161175,
      "loss": 5.0081,
      "mean_token_accuracy": 0.19063785821199417,
      "num_tokens": 73219596.0,
      "step": 33420
    },
    {
      "entropy": 5.700966739654541,
      "epoch": 3.3172886065899165,
      "grad_norm": 1.125,
      "learning_rate": 0.0003925304546216954,
      "loss": 5.038,
      "mean_token_accuracy": 0.18705653101205827,
      "num_tokens": 73232301.0,
      "step": 33425
    },
    {
      "entropy": 5.742330646514892,
      "epoch": 3.317784835252084,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000392500239787277,
      "loss": 5.0288,
      "mean_token_accuracy": 0.1860410675406456,
      "num_tokens": 73243447.0,
      "step": 33430
    },
    {
      "entropy": 5.697584533691407,
      "epoch": 3.3182810639142515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003924700220391058,
      "loss": 5.0582,
      "mean_token_accuracy": 0.18202148675918578,
      "num_tokens": 73254336.0,
      "step": 33435
    },
    {
      "entropy": 5.737822866439819,
      "epoch": 3.3187772925764194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003924398013779312,
      "loss": 5.0252,
      "mean_token_accuracy": 0.19278808534145356,
      "num_tokens": 73266098.0,
      "step": 33440
    },
    {
      "entropy": 5.615671014785766,
      "epoch": 3.319273521238587,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003924095778045026,
      "loss": 4.9983,
      "mean_token_accuracy": 0.1913063958287239,
      "num_tokens": 73276026.0,
      "step": 33445
    },
    {
      "entropy": 5.657993745803833,
      "epoch": 3.3197697499007544,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003923793513195695,
      "loss": 4.9611,
      "mean_token_accuracy": 0.1946242168545723,
      "num_tokens": 73286151.0,
      "step": 33450
    },
    {
      "entropy": 5.722117948532104,
      "epoch": 3.320265978562922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039234912192388137,
      "loss": 5.0846,
      "mean_token_accuracy": 0.17912802696228028,
      "num_tokens": 73298086.0,
      "step": 33455
    },
    {
      "entropy": 5.716272020339966,
      "epoch": 3.3207622072250893,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003923188896181879,
      "loss": 5.118,
      "mean_token_accuracy": 0.1771721914410591,
      "num_tokens": 73308732.0,
      "step": 33460
    },
    {
      "entropy": 5.68450026512146,
      "epoch": 3.3212584358872568,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003922886544032388,
      "loss": 4.9869,
      "mean_token_accuracy": 0.18693410158157348,
      "num_tokens": 73318979.0,
      "step": 33465
    },
    {
      "entropy": 5.731170034408569,
      "epoch": 3.3217546645494243,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00039225841627978374,
      "loss": 5.0592,
      "mean_token_accuracy": 0.18314978927373887,
      "num_tokens": 73329724.0,
      "step": 33470
    },
    {
      "entropy": 5.623520708084106,
      "epoch": 3.3222508932115917,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003922281752485727,
      "loss": 4.9852,
      "mean_token_accuracy": 0.18334131985902785,
      "num_tokens": 73340463.0,
      "step": 33475
    },
    {
      "entropy": 5.592458438873291,
      "epoch": 3.3227471218737596,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039219793131035544,
      "loss": 4.8807,
      "mean_token_accuracy": 0.20323915034532547,
      "num_tokens": 73351089.0,
      "step": 33480
    },
    {
      "entropy": 5.681954050064087,
      "epoch": 3.323243350535927,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003921676844658821,
      "loss": 5.0761,
      "mean_token_accuracy": 0.18385167866945268,
      "num_tokens": 73361747.0,
      "step": 33485
    },
    {
      "entropy": 5.676303863525391,
      "epoch": 3.3237395791980946,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003921374347159026,
      "loss": 5.0135,
      "mean_token_accuracy": 0.1892842948436737,
      "num_tokens": 73372340.0,
      "step": 33490
    },
    {
      "entropy": 5.6614158153533936,
      "epoch": 3.324235807860262,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003921071820611671,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18628250658512116,
      "num_tokens": 73383972.0,
      "step": 33495
    },
    {
      "entropy": 5.626462459564209,
      "epoch": 3.3247320365224295,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003920769265024259,
      "loss": 4.9701,
      "mean_token_accuracy": 0.18832410722970963,
      "num_tokens": 73394346.0,
      "step": 33500
    },
    {
      "entropy": 5.681484699249268,
      "epoch": 3.325228265184597,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039204666804042913,
      "loss": 5.0226,
      "mean_token_accuracy": 0.18199969679117203,
      "num_tokens": 73404964.0,
      "step": 33505
    },
    {
      "entropy": 5.688396263122558,
      "epoch": 3.3257244938467645,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003920164066759273,
      "loss": 5.0439,
      "mean_token_accuracy": 0.17996495515108107,
      "num_tokens": 73415487.0,
      "step": 33510
    },
    {
      "entropy": 5.7018474578857425,
      "epoch": 3.326220722508932,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003919861424096707,
      "loss": 5.0374,
      "mean_token_accuracy": 0.18502648919820786,
      "num_tokens": 73424909.0,
      "step": 33515
    },
    {
      "entropy": 5.765803098678589,
      "epoch": 3.3267169511710994,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003919558752424098,
      "loss": 5.0354,
      "mean_token_accuracy": 0.18085672408342363,
      "num_tokens": 73435106.0,
      "step": 33520
    },
    {
      "entropy": 5.71702938079834,
      "epoch": 3.327213179833267,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003919256051748952,
      "loss": 5.0554,
      "mean_token_accuracy": 0.18190431147813796,
      "num_tokens": 73445606.0,
      "step": 33525
    },
    {
      "entropy": 5.61006727218628,
      "epoch": 3.327709408495435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039189533220787765,
      "loss": 4.9932,
      "mean_token_accuracy": 0.1852023795247078,
      "num_tokens": 73456309.0,
      "step": 33530
    },
    {
      "entropy": 5.625271129608154,
      "epoch": 3.3282056371576023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003918650563421076,
      "loss": 4.9711,
      "mean_token_accuracy": 0.19044086039066316,
      "num_tokens": 73467817.0,
      "step": 33535
    },
    {
      "entropy": 5.616325998306275,
      "epoch": 3.32870186581977,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000391834777578336,
      "loss": 4.9285,
      "mean_token_accuracy": 0.19254331290721893,
      "num_tokens": 73477312.0,
      "step": 33540
    },
    {
      "entropy": 5.6972699642181395,
      "epoch": 3.3291980944819373,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003918044959173138,
      "loss": 5.0984,
      "mean_token_accuracy": 0.1836540475487709,
      "num_tokens": 73488772.0,
      "step": 33545
    },
    {
      "entropy": 5.642685842514038,
      "epoch": 3.3296943231441047,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003917742113597918,
      "loss": 4.9696,
      "mean_token_accuracy": 0.189340041577816,
      "num_tokens": 73500570.0,
      "step": 33550
    },
    {
      "entropy": 5.606554174423218,
      "epoch": 3.330190551806272,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003917439239065208,
      "loss": 4.9239,
      "mean_token_accuracy": 0.1946498081088066,
      "num_tokens": 73511679.0,
      "step": 33555
    },
    {
      "entropy": 5.639152097702026,
      "epoch": 3.3306867804684397,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039171363355825216,
      "loss": 4.9601,
      "mean_token_accuracy": 0.19668059796094894,
      "num_tokens": 73523312.0,
      "step": 33560
    },
    {
      "entropy": 5.697023344039917,
      "epoch": 3.3311830091306076,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039168334031573695,
      "loss": 5.0455,
      "mean_token_accuracy": 0.18710092902183534,
      "num_tokens": 73534789.0,
      "step": 33565
    },
    {
      "entropy": 5.730168581008911,
      "epoch": 3.331679237792775,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003916530441797262,
      "loss": 5.0,
      "mean_token_accuracy": 0.18439022600650787,
      "num_tokens": 73545331.0,
      "step": 33570
    },
    {
      "entropy": 5.6203468322753904,
      "epoch": 3.3321754664549426,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039162274515097147,
      "loss": 4.9364,
      "mean_token_accuracy": 0.1871286600828171,
      "num_tokens": 73555347.0,
      "step": 33575
    },
    {
      "entropy": 5.603216457366943,
      "epoch": 3.33267169511711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003915924432302239,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18329586684703827,
      "num_tokens": 73567180.0,
      "step": 33580
    },
    {
      "entropy": 5.668341636657715,
      "epoch": 3.3331679237792775,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003915621384182349,
      "loss": 5.0612,
      "mean_token_accuracy": 0.18515645563602448,
      "num_tokens": 73578589.0,
      "step": 33585
    },
    {
      "entropy": 5.658218479156494,
      "epoch": 3.333664152441445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003915318307157561,
      "loss": 4.9593,
      "mean_token_accuracy": 0.18917322009801865,
      "num_tokens": 73588939.0,
      "step": 33590
    },
    {
      "entropy": 5.694159269332886,
      "epoch": 3.3341603811036125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039150152012353896,
      "loss": 5.0214,
      "mean_token_accuracy": 0.18520694971084595,
      "num_tokens": 73600476.0,
      "step": 33595
    },
    {
      "entropy": 5.722176647186279,
      "epoch": 3.33465660976578,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039147120664233525,
      "loss": 5.0485,
      "mean_token_accuracy": 0.19593417793512344,
      "num_tokens": 73610826.0,
      "step": 33600
    },
    {
      "entropy": 5.660782527923584,
      "epoch": 3.3351528384279474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039144089027289653,
      "loss": 4.9763,
      "mean_token_accuracy": 0.1865175485610962,
      "num_tokens": 73623029.0,
      "step": 33605
    },
    {
      "entropy": 5.6054737091064455,
      "epoch": 3.335649067090115,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039141057101597465,
      "loss": 4.9608,
      "mean_token_accuracy": 0.18714158236980438,
      "num_tokens": 73634092.0,
      "step": 33610
    },
    {
      "entropy": 5.640215921401977,
      "epoch": 3.336145295752283,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003913802488723215,
      "loss": 4.9425,
      "mean_token_accuracy": 0.20204665064811705,
      "num_tokens": 73643297.0,
      "step": 33615
    },
    {
      "entropy": 5.629556322097779,
      "epoch": 3.3366415244144503,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003913499238426889,
      "loss": 4.9855,
      "mean_token_accuracy": 0.1913979709148407,
      "num_tokens": 73654001.0,
      "step": 33620
    },
    {
      "entropy": 5.652475166320801,
      "epoch": 3.3371377530766178,
      "grad_norm": 1.109375,
      "learning_rate": 0.00039131959592782884,
      "loss": 4.9708,
      "mean_token_accuracy": 0.19035686403512955,
      "num_tokens": 73664509.0,
      "step": 33625
    },
    {
      "entropy": 5.585935783386231,
      "epoch": 3.3376339817387852,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039128926512849356,
      "loss": 4.8743,
      "mean_token_accuracy": 0.1981704980134964,
      "num_tokens": 73674585.0,
      "step": 33630
    },
    {
      "entropy": 5.5772867679595945,
      "epoch": 3.3381302104009527,
      "grad_norm": 1.125,
      "learning_rate": 0.000391258931445435,
      "loss": 4.8608,
      "mean_token_accuracy": 0.19290991872549057,
      "num_tokens": 73684550.0,
      "step": 33635
    },
    {
      "entropy": 5.669147491455078,
      "epoch": 3.33862643906312,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003912285948794055,
      "loss": 5.0174,
      "mean_token_accuracy": 0.18562725633382798,
      "num_tokens": 73694767.0,
      "step": 33640
    },
    {
      "entropy": 5.786861419677734,
      "epoch": 3.3391226677252877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003911982554311573,
      "loss": 5.064,
      "mean_token_accuracy": 0.19083971679210662,
      "num_tokens": 73707900.0,
      "step": 33645
    },
    {
      "entropy": 5.6858320236206055,
      "epoch": 3.3396188963874556,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039116791310144275,
      "loss": 5.0402,
      "mean_token_accuracy": 0.184928235411644,
      "num_tokens": 73719014.0,
      "step": 33650
    },
    {
      "entropy": 5.70911865234375,
      "epoch": 3.340115125049623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003911375678910142,
      "loss": 5.0997,
      "mean_token_accuracy": 0.18363464027643203,
      "num_tokens": 73730107.0,
      "step": 33655
    },
    {
      "entropy": 5.716792011260987,
      "epoch": 3.3406113537117905,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039110721980062434,
      "loss": 5.1172,
      "mean_token_accuracy": 0.1801718771457672,
      "num_tokens": 73742355.0,
      "step": 33660
    },
    {
      "entropy": 5.737859582901001,
      "epoch": 3.341107582373958,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039107686883102554,
      "loss": 5.0517,
      "mean_token_accuracy": 0.18479863852262496,
      "num_tokens": 73753372.0,
      "step": 33665
    },
    {
      "entropy": 5.681610012054444,
      "epoch": 3.3416038110361255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039104651498297053,
      "loss": 4.9768,
      "mean_token_accuracy": 0.1913033127784729,
      "num_tokens": 73764547.0,
      "step": 33670
    },
    {
      "entropy": 5.7104650974273685,
      "epoch": 3.342100039698293,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000391016158257212,
      "loss": 5.065,
      "mean_token_accuracy": 0.18229050785303116,
      "num_tokens": 73775505.0,
      "step": 33675
    },
    {
      "entropy": 5.757601499557495,
      "epoch": 3.3425962683604604,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003909857986545028,
      "loss": 5.0808,
      "mean_token_accuracy": 0.18112506568431855,
      "num_tokens": 73785963.0,
      "step": 33680
    },
    {
      "entropy": 5.6944005489349365,
      "epoch": 3.343092497022628,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00039095543617559566,
      "loss": 4.9351,
      "mean_token_accuracy": 0.1938188463449478,
      "num_tokens": 73796457.0,
      "step": 33685
    },
    {
      "entropy": 5.705489540100098,
      "epoch": 3.3435887256847954,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003909250708212436,
      "loss": 5.0143,
      "mean_token_accuracy": 0.18873758167028426,
      "num_tokens": 73807662.0,
      "step": 33690
    },
    {
      "entropy": 5.653219842910767,
      "epoch": 3.344084954346963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039089470259219956,
      "loss": 5.0192,
      "mean_token_accuracy": 0.1845548689365387,
      "num_tokens": 73818688.0,
      "step": 33695
    },
    {
      "entropy": 5.731749820709228,
      "epoch": 3.3445811830091308,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039086433148921665,
      "loss": 5.0469,
      "mean_token_accuracy": 0.18607240468263625,
      "num_tokens": 73831234.0,
      "step": 33700
    },
    {
      "entropy": 5.653010988235474,
      "epoch": 3.3450774116712982,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003908339575130479,
      "loss": 5.0091,
      "mean_token_accuracy": 0.19199507385492326,
      "num_tokens": 73841505.0,
      "step": 33705
    },
    {
      "entropy": 5.6795525550842285,
      "epoch": 3.3455736403334657,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003908035806644467,
      "loss": 5.0491,
      "mean_token_accuracy": 0.18339713960886,
      "num_tokens": 73851900.0,
      "step": 33710
    },
    {
      "entropy": 5.716470861434937,
      "epoch": 3.346069868995633,
      "grad_norm": 1.3125,
      "learning_rate": 0.00039077320094416617,
      "loss": 4.9784,
      "mean_token_accuracy": 0.1928652420639992,
      "num_tokens": 73862317.0,
      "step": 33715
    },
    {
      "entropy": 5.713816738128662,
      "epoch": 3.3465660976578007,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003907428183529598,
      "loss": 5.0219,
      "mean_token_accuracy": 0.17927050292491914,
      "num_tokens": 73872166.0,
      "step": 33720
    },
    {
      "entropy": 5.696255874633789,
      "epoch": 3.347062326319968,
      "grad_norm": 1.4375,
      "learning_rate": 0.00039071243289158085,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18570499122142792,
      "num_tokens": 73883064.0,
      "step": 33725
    },
    {
      "entropy": 5.693522214889526,
      "epoch": 3.3475585549821356,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039068204456078296,
      "loss": 5.0044,
      "mean_token_accuracy": 0.18708179146051407,
      "num_tokens": 73895032.0,
      "step": 33730
    },
    {
      "entropy": 5.680994033813477,
      "epoch": 3.3480547836443035,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039065165336131963,
      "loss": 4.9732,
      "mean_token_accuracy": 0.19002187997102737,
      "num_tokens": 73906309.0,
      "step": 33735
    },
    {
      "entropy": 5.650827646255493,
      "epoch": 3.348551012306471,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003906212592939445,
      "loss": 4.9873,
      "mean_token_accuracy": 0.1873575672507286,
      "num_tokens": 73917240.0,
      "step": 33740
    },
    {
      "entropy": 5.623562002182007,
      "epoch": 3.3490472409686385,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039059086235941127,
      "loss": 4.9596,
      "mean_token_accuracy": 0.1897199869155884,
      "num_tokens": 73928216.0,
      "step": 33745
    },
    {
      "entropy": 5.634237241744995,
      "epoch": 3.349543469630806,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003905604625584737,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18724946230649947,
      "num_tokens": 73940115.0,
      "step": 33750
    },
    {
      "entropy": 5.703455400466919,
      "epoch": 3.3500396982929734,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003905300598918857,
      "loss": 5.0845,
      "mean_token_accuracy": 0.18400780111551285,
      "num_tokens": 73950756.0,
      "step": 33755
    },
    {
      "entropy": 5.651713943481445,
      "epoch": 3.350535926955141,
      "grad_norm": 1.1875,
      "learning_rate": 0.00039049965436040125,
      "loss": 4.9822,
      "mean_token_accuracy": 0.1922023430466652,
      "num_tokens": 73962269.0,
      "step": 33760
    },
    {
      "entropy": 5.680441093444824,
      "epoch": 3.3510321556173084,
      "grad_norm": 1.25,
      "learning_rate": 0.0003904692459647741,
      "loss": 4.9635,
      "mean_token_accuracy": 0.18371140956878662,
      "num_tokens": 73971883.0,
      "step": 33765
    },
    {
      "entropy": 5.738015365600586,
      "epoch": 3.351528384279476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039043883470575856,
      "loss": 5.0817,
      "mean_token_accuracy": 0.18666813522577286,
      "num_tokens": 73983801.0,
      "step": 33770
    },
    {
      "entropy": 5.612909364700317,
      "epoch": 3.3520246129416433,
      "grad_norm": 1.203125,
      "learning_rate": 0.00039040842058410863,
      "loss": 4.9763,
      "mean_token_accuracy": 0.19230974614620208,
      "num_tokens": 73994198.0,
      "step": 33775
    },
    {
      "entropy": 5.695518970489502,
      "epoch": 3.352520841603811,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039037800360057864,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18092825561761855,
      "num_tokens": 74005521.0,
      "step": 33780
    },
    {
      "entropy": 5.731161308288574,
      "epoch": 3.3530170702659787,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003903475837559226,
      "loss": 5.0564,
      "mean_token_accuracy": 0.19308481514453887,
      "num_tokens": 74016606.0,
      "step": 33785
    },
    {
      "entropy": 5.695236158370972,
      "epoch": 3.353513298928146,
      "grad_norm": 1.234375,
      "learning_rate": 0.00039031716105089516,
      "loss": 5.0008,
      "mean_token_accuracy": 0.19350726753473282,
      "num_tokens": 74027111.0,
      "step": 33790
    },
    {
      "entropy": 5.642651462554932,
      "epoch": 3.3540095275903137,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003902867354862506,
      "loss": 4.9268,
      "mean_token_accuracy": 0.18914511650800706,
      "num_tokens": 74037425.0,
      "step": 33795
    },
    {
      "entropy": 5.634895658493042,
      "epoch": 3.354505756252481,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003902563070627434,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18747079819440843,
      "num_tokens": 74048622.0,
      "step": 33800
    },
    {
      "entropy": 5.638277912139893,
      "epoch": 3.3550019849146486,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039022587578112813,
      "loss": 4.9737,
      "mean_token_accuracy": 0.1920724019408226,
      "num_tokens": 74058430.0,
      "step": 33805
    },
    {
      "entropy": 5.708781576156616,
      "epoch": 3.355498213576816,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00039019544164215956,
      "loss": 4.9916,
      "mean_token_accuracy": 0.18916950970888138,
      "num_tokens": 74069488.0,
      "step": 33810
    },
    {
      "entropy": 5.714410400390625,
      "epoch": 3.3559944422389836,
      "grad_norm": 1.21875,
      "learning_rate": 0.00039016500464659213,
      "loss": 4.9897,
      "mean_token_accuracy": 0.19073303192853927,
      "num_tokens": 74079142.0,
      "step": 33815
    },
    {
      "entropy": 5.687404584884644,
      "epoch": 3.3564906709011515,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003901345647951807,
      "loss": 5.036,
      "mean_token_accuracy": 0.18090758323669434,
      "num_tokens": 74089984.0,
      "step": 33820
    },
    {
      "entropy": 5.653545141220093,
      "epoch": 3.356986899563319,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00039010412208868026,
      "loss": 5.011,
      "mean_token_accuracy": 0.1842905357480049,
      "num_tokens": 74101225.0,
      "step": 33825
    },
    {
      "entropy": 5.640694427490234,
      "epoch": 3.3574831282254864,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003900736765278455,
      "loss": 4.9165,
      "mean_token_accuracy": 0.19844931215047837,
      "num_tokens": 74111198.0,
      "step": 33830
    },
    {
      "entropy": 5.731590414047242,
      "epoch": 3.357979356887654,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039004322811343165,
      "loss": 5.0834,
      "mean_token_accuracy": 0.1785370260477066,
      "num_tokens": 74121458.0,
      "step": 33835
    },
    {
      "entropy": 5.669651222229004,
      "epoch": 3.3584755855498214,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039001277684619357,
      "loss": 5.0078,
      "mean_token_accuracy": 0.19264528453350066,
      "num_tokens": 74133338.0,
      "step": 33840
    },
    {
      "entropy": 5.638198614120483,
      "epoch": 3.358971814211989,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003899823227268865,
      "loss": 5.0395,
      "mean_token_accuracy": 0.18176195472478868,
      "num_tokens": 74144062.0,
      "step": 33845
    },
    {
      "entropy": 5.638459491729736,
      "epoch": 3.3594680428741563,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038995186575626554,
      "loss": 4.9718,
      "mean_token_accuracy": 0.18808696866035463,
      "num_tokens": 74154001.0,
      "step": 33850
    },
    {
      "entropy": 5.641145324707031,
      "epoch": 3.359964271536324,
      "grad_norm": 1.140625,
      "learning_rate": 0.000389921405935086,
      "loss": 4.9668,
      "mean_token_accuracy": 0.19124677926301956,
      "num_tokens": 74164254.0,
      "step": 33855
    },
    {
      "entropy": 5.610278558731079,
      "epoch": 3.3604605001984913,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038989094326410327,
      "loss": 4.933,
      "mean_token_accuracy": 0.192996484041214,
      "num_tokens": 74173915.0,
      "step": 33860
    },
    {
      "entropy": 5.779168796539307,
      "epoch": 3.3609567288606588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003898604777440727,
      "loss": 5.1135,
      "mean_token_accuracy": 0.17460862845182418,
      "num_tokens": 74185903.0,
      "step": 33865
    },
    {
      "entropy": 5.741213417053222,
      "epoch": 3.3614529575228267,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003898300093757499,
      "loss": 5.0253,
      "mean_token_accuracy": 0.18167859613895415,
      "num_tokens": 74198361.0,
      "step": 33870
    },
    {
      "entropy": 5.626187896728515,
      "epoch": 3.361949186184994,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003897995381598901,
      "loss": 4.9748,
      "mean_token_accuracy": 0.18296902626752853,
      "num_tokens": 74210467.0,
      "step": 33875
    },
    {
      "entropy": 5.581474351882934,
      "epoch": 3.3624454148471616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003897690640972492,
      "loss": 4.9272,
      "mean_token_accuracy": 0.19539858847856523,
      "num_tokens": 74221717.0,
      "step": 33880
    },
    {
      "entropy": 5.624989557266235,
      "epoch": 3.362941643509329,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038973858718858295,
      "loss": 4.9952,
      "mean_token_accuracy": 0.18455081582069396,
      "num_tokens": 74233164.0,
      "step": 33885
    },
    {
      "entropy": 5.749930047988892,
      "epoch": 3.3634378721714966,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003897081074346468,
      "loss": 5.0486,
      "mean_token_accuracy": 0.18696786165237428,
      "num_tokens": 74245693.0,
      "step": 33890
    },
    {
      "entropy": 5.650203800201416,
      "epoch": 3.363934100833664,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038967762483619684,
      "loss": 4.9289,
      "mean_token_accuracy": 0.1973839983344078,
      "num_tokens": 74256499.0,
      "step": 33895
    },
    {
      "entropy": 5.666788148880005,
      "epoch": 3.3644303294958315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003896471393939889,
      "loss": 5.0412,
      "mean_token_accuracy": 0.18949628472328187,
      "num_tokens": 74267713.0,
      "step": 33900
    },
    {
      "entropy": 5.677918529510498,
      "epoch": 3.364926558157999,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003896166511087789,
      "loss": 5.0718,
      "mean_token_accuracy": 0.18289435505867005,
      "num_tokens": 74279149.0,
      "step": 33905
    },
    {
      "entropy": 5.669685506820679,
      "epoch": 3.365422786820167,
      "grad_norm": 1.1875,
      "learning_rate": 0.000389586159981323,
      "loss": 4.9925,
      "mean_token_accuracy": 0.1902776151895523,
      "num_tokens": 74289520.0,
      "step": 33910
    },
    {
      "entropy": 5.7012704372406,
      "epoch": 3.3659190154823344,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003895556660123771,
      "loss": 4.9982,
      "mean_token_accuracy": 0.19324909448623656,
      "num_tokens": 74300546.0,
      "step": 33915
    },
    {
      "entropy": 5.619011497497558,
      "epoch": 3.366415244144502,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038952516920269764,
      "loss": 4.9453,
      "mean_token_accuracy": 0.18342457562685013,
      "num_tokens": 74310687.0,
      "step": 33920
    },
    {
      "entropy": 5.646815156936645,
      "epoch": 3.3669114728066694,
      "grad_norm": 1.25,
      "learning_rate": 0.0003894946695530408,
      "loss": 4.9916,
      "mean_token_accuracy": 0.18411651849746705,
      "num_tokens": 74322251.0,
      "step": 33925
    },
    {
      "entropy": 5.717867279052735,
      "epoch": 3.367407701468837,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038946416706416275,
      "loss": 5.0379,
      "mean_token_accuracy": 0.18231473863124847,
      "num_tokens": 74333845.0,
      "step": 33930
    },
    {
      "entropy": 5.688938188552856,
      "epoch": 3.3679039301310043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003894336617368201,
      "loss": 5.0392,
      "mean_token_accuracy": 0.18734661042690276,
      "num_tokens": 74344188.0,
      "step": 33935
    },
    {
      "entropy": 5.690020561218262,
      "epoch": 3.368400158793172,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038940315357176926,
      "loss": 5.0583,
      "mean_token_accuracy": 0.1840840682387352,
      "num_tokens": 74355174.0,
      "step": 33940
    },
    {
      "entropy": 5.667742109298706,
      "epoch": 3.3688963874553393,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038937264256976657,
      "loss": 4.9361,
      "mean_token_accuracy": 0.1912268504500389,
      "num_tokens": 74365396.0,
      "step": 33945
    },
    {
      "entropy": 5.683155107498169,
      "epoch": 3.3693926161175067,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003893421287315689,
      "loss": 4.9952,
      "mean_token_accuracy": 0.18311749994754792,
      "num_tokens": 74377303.0,
      "step": 33950
    },
    {
      "entropy": 5.627586078643799,
      "epoch": 3.3698888447796747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003893116120579328,
      "loss": 4.9993,
      "mean_token_accuracy": 0.1894754499197006,
      "num_tokens": 74389967.0,
      "step": 33955
    },
    {
      "entropy": 5.6795391082763675,
      "epoch": 3.370385073441842,
      "grad_norm": 1.125,
      "learning_rate": 0.0003892810925496151,
      "loss": 4.98,
      "mean_token_accuracy": 0.1902260586619377,
      "num_tokens": 74400252.0,
      "step": 33960
    },
    {
      "entropy": 5.643511867523193,
      "epoch": 3.3708813021040096,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003892505702073725,
      "loss": 4.9705,
      "mean_token_accuracy": 0.18851204365491867,
      "num_tokens": 74412165.0,
      "step": 33965
    },
    {
      "entropy": 5.610087108612061,
      "epoch": 3.371377530766177,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038922004503196195,
      "loss": 4.9255,
      "mean_token_accuracy": 0.1961517572402954,
      "num_tokens": 74423078.0,
      "step": 33970
    },
    {
      "entropy": 5.591834831237793,
      "epoch": 3.3718737594283446,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003891895170241405,
      "loss": 4.9156,
      "mean_token_accuracy": 0.19147985577583312,
      "num_tokens": 74434552.0,
      "step": 33975
    },
    {
      "entropy": 5.691589832305908,
      "epoch": 3.372369988090512,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038915898618466487,
      "loss": 5.0316,
      "mean_token_accuracy": 0.18602710515260695,
      "num_tokens": 74446188.0,
      "step": 33980
    },
    {
      "entropy": 5.622373008728028,
      "epoch": 3.3728662167526795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038912845251429256,
      "loss": 4.9226,
      "mean_token_accuracy": 0.19175295084714888,
      "num_tokens": 74456358.0,
      "step": 33985
    },
    {
      "entropy": 5.585901975631714,
      "epoch": 3.373362445414847,
      "grad_norm": 1.125,
      "learning_rate": 0.00038909791601378054,
      "loss": 4.8936,
      "mean_token_accuracy": 0.19483769685029984,
      "num_tokens": 74466155.0,
      "step": 33990
    },
    {
      "entropy": 5.717492341995239,
      "epoch": 3.373858674077015,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038906737668388595,
      "loss": 5.08,
      "mean_token_accuracy": 0.17776141613721846,
      "num_tokens": 74477282.0,
      "step": 33995
    },
    {
      "entropy": 5.599173879623413,
      "epoch": 3.3743549027391824,
      "grad_norm": 1.125,
      "learning_rate": 0.00038903683452536634,
      "loss": 4.957,
      "mean_token_accuracy": 0.18897999227046966,
      "num_tokens": 74489736.0,
      "step": 34000
    },
    {
      "entropy": 5.584864473342895,
      "epoch": 3.37485113140135,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003890062895389789,
      "loss": 4.9573,
      "mean_token_accuracy": 0.18596386164426804,
      "num_tokens": 74500797.0,
      "step": 34005
    },
    {
      "entropy": 5.622355890274048,
      "epoch": 3.3753473600635173,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003889757417254811,
      "loss": 4.982,
      "mean_token_accuracy": 0.19108611345291138,
      "num_tokens": 74510811.0,
      "step": 34010
    },
    {
      "entropy": 5.662883234024048,
      "epoch": 3.375843588725685,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038894519108563047,
      "loss": 4.9667,
      "mean_token_accuracy": 0.19344783574342728,
      "num_tokens": 74521726.0,
      "step": 34015
    },
    {
      "entropy": 5.638061285018921,
      "epoch": 3.3763398173878523,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038891463762018476,
      "loss": 4.9784,
      "mean_token_accuracy": 0.19426548779010772,
      "num_tokens": 74532260.0,
      "step": 34020
    },
    {
      "entropy": 5.719440269470215,
      "epoch": 3.3768360460500197,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038888408132990146,
      "loss": 5.0641,
      "mean_token_accuracy": 0.17768387347459794,
      "num_tokens": 74544691.0,
      "step": 34025
    },
    {
      "entropy": 5.723568248748779,
      "epoch": 3.377332274712187,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038885352221553827,
      "loss": 5.0296,
      "mean_token_accuracy": 0.18223040103912352,
      "num_tokens": 74555250.0,
      "step": 34030
    },
    {
      "entropy": 5.644944334030152,
      "epoch": 3.3778285033743547,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003888229602778532,
      "loss": 4.9442,
      "mean_token_accuracy": 0.19133111238479614,
      "num_tokens": 74566079.0,
      "step": 34035
    },
    {
      "entropy": 5.676452541351319,
      "epoch": 3.3783247320365226,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038879239551760386,
      "loss": 5.013,
      "mean_token_accuracy": 0.18377969563007354,
      "num_tokens": 74577741.0,
      "step": 34040
    },
    {
      "entropy": 5.629088401794434,
      "epoch": 3.37882096069869,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003887618279355483,
      "loss": 4.9468,
      "mean_token_accuracy": 0.19178782552480697,
      "num_tokens": 74588146.0,
      "step": 34045
    },
    {
      "entropy": 5.666579484939575,
      "epoch": 3.3793171893608576,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038873125753244466,
      "loss": 5.088,
      "mean_token_accuracy": 0.1867673620581627,
      "num_tokens": 74599745.0,
      "step": 34050
    },
    {
      "entropy": 5.6874805927276615,
      "epoch": 3.379813418023025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003887006843090508,
      "loss": 4.9438,
      "mean_token_accuracy": 0.1840873882174492,
      "num_tokens": 74611215.0,
      "step": 34055
    },
    {
      "entropy": 5.73456449508667,
      "epoch": 3.3803096466851925,
      "grad_norm": 1.203125,
      "learning_rate": 0.000388670108266125,
      "loss": 5.0498,
      "mean_token_accuracy": 0.19403781294822692,
      "num_tokens": 74620676.0,
      "step": 34060
    },
    {
      "entropy": 5.72516131401062,
      "epoch": 3.38080587534736,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003886395294044255,
      "loss": 5.0809,
      "mean_token_accuracy": 0.18085369616746902,
      "num_tokens": 74633129.0,
      "step": 34065
    },
    {
      "entropy": 5.677772760391235,
      "epoch": 3.3813021040095275,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003886089477247105,
      "loss": 4.9504,
      "mean_token_accuracy": 0.1881747379899025,
      "num_tokens": 74643125.0,
      "step": 34070
    },
    {
      "entropy": 5.756685829162597,
      "epoch": 3.381798332671695,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003885783632277384,
      "loss": 5.092,
      "mean_token_accuracy": 0.18524766117334365,
      "num_tokens": 74653573.0,
      "step": 34075
    },
    {
      "entropy": 5.6284827709198,
      "epoch": 3.382294561333863,
      "grad_norm": 1.25,
      "learning_rate": 0.0003885477759142676,
      "loss": 4.9227,
      "mean_token_accuracy": 0.1922967776656151,
      "num_tokens": 74663451.0,
      "step": 34080
    },
    {
      "entropy": 5.601763343811035,
      "epoch": 3.3827907899960303,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003885171857850567,
      "loss": 4.9444,
      "mean_token_accuracy": 0.19517677277326584,
      "num_tokens": 74673698.0,
      "step": 34085
    },
    {
      "entropy": 5.624125242233276,
      "epoch": 3.383287018658198,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003884865928408642,
      "loss": 5.0259,
      "mean_token_accuracy": 0.1851004257798195,
      "num_tokens": 74685952.0,
      "step": 34090
    },
    {
      "entropy": 5.640368032455444,
      "epoch": 3.3837832473203653,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003884559970824486,
      "loss": 4.9818,
      "mean_token_accuracy": 0.1905374765396118,
      "num_tokens": 74698159.0,
      "step": 34095
    },
    {
      "entropy": 5.734451341629028,
      "epoch": 3.3842794759825328,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003884253985105689,
      "loss": 5.0565,
      "mean_token_accuracy": 0.18477711081504822,
      "num_tokens": 74708718.0,
      "step": 34100
    },
    {
      "entropy": 5.646965551376343,
      "epoch": 3.3847757046447002,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003883947971259837,
      "loss": 4.9222,
      "mean_token_accuracy": 0.19178639948368073,
      "num_tokens": 74718249.0,
      "step": 34105
    },
    {
      "entropy": 5.681203842163086,
      "epoch": 3.3852719333068677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003883641929294519,
      "loss": 4.9084,
      "mean_token_accuracy": 0.20431592762470246,
      "num_tokens": 74728856.0,
      "step": 34110
    },
    {
      "entropy": 5.637758827209472,
      "epoch": 3.385768161969035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003883335859217323,
      "loss": 4.9833,
      "mean_token_accuracy": 0.19112443625926973,
      "num_tokens": 74740126.0,
      "step": 34115
    },
    {
      "entropy": 5.631464004516602,
      "epoch": 3.3862643906312027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000388302976103584,
      "loss": 4.9872,
      "mean_token_accuracy": 0.1898806482553482,
      "num_tokens": 74751277.0,
      "step": 34120
    },
    {
      "entropy": 5.627474927902222,
      "epoch": 3.3867606192933706,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003882723634757661,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18442867696285248,
      "num_tokens": 74761498.0,
      "step": 34125
    },
    {
      "entropy": 5.637354469299316,
      "epoch": 3.387256847955538,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003882417480390376,
      "loss": 4.9628,
      "mean_token_accuracy": 0.18227755576372145,
      "num_tokens": 74771482.0,
      "step": 34130
    },
    {
      "entropy": 5.7021321773529055,
      "epoch": 3.3877530766177055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038821112979415786,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18583563566207886,
      "num_tokens": 74783649.0,
      "step": 34135
    },
    {
      "entropy": 5.6576722145080565,
      "epoch": 3.388249305279873,
      "grad_norm": 1.34375,
      "learning_rate": 0.000388180508741886,
      "loss": 4.9726,
      "mean_token_accuracy": 0.193347504734993,
      "num_tokens": 74794142.0,
      "step": 34140
    },
    {
      "entropy": 5.62070631980896,
      "epoch": 3.3887455339420405,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038814988488298136,
      "loss": 4.9752,
      "mean_token_accuracy": 0.18797930628061293,
      "num_tokens": 74805369.0,
      "step": 34145
    },
    {
      "entropy": 5.651659393310547,
      "epoch": 3.389241762604208,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003881192582182034,
      "loss": 5.0009,
      "mean_token_accuracy": 0.18141744434833526,
      "num_tokens": 74816558.0,
      "step": 34150
    },
    {
      "entropy": 5.68314847946167,
      "epoch": 3.3897379912663754,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038808862874831164,
      "loss": 4.9634,
      "mean_token_accuracy": 0.19343261420726776,
      "num_tokens": 74826838.0,
      "step": 34155
    },
    {
      "entropy": 5.590038013458252,
      "epoch": 3.390234219928543,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003880579964740655,
      "loss": 4.9528,
      "mean_token_accuracy": 0.181502865254879,
      "num_tokens": 74837182.0,
      "step": 34160
    },
    {
      "entropy": 5.64497447013855,
      "epoch": 3.390730448590711,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038802736139622473,
      "loss": 4.9152,
      "mean_token_accuracy": 0.19948959201574326,
      "num_tokens": 74847973.0,
      "step": 34165
    },
    {
      "entropy": 5.664077091217041,
      "epoch": 3.3912266772528783,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038799672351554895,
      "loss": 4.9916,
      "mean_token_accuracy": 0.19287513792514802,
      "num_tokens": 74858952.0,
      "step": 34170
    },
    {
      "entropy": 5.751750135421753,
      "epoch": 3.3917229059150458,
      "grad_norm": 1.125,
      "learning_rate": 0.00038796608283279786,
      "loss": 5.0259,
      "mean_token_accuracy": 0.18157842457294465,
      "num_tokens": 74870427.0,
      "step": 34175
    },
    {
      "entropy": 5.635307312011719,
      "epoch": 3.3922191345772132,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003879354393487314,
      "loss": 4.9794,
      "mean_token_accuracy": 0.19191219359636308,
      "num_tokens": 74881502.0,
      "step": 34180
    },
    {
      "entropy": 5.673541641235351,
      "epoch": 3.3927153632393807,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038790479306410934,
      "loss": 5.0452,
      "mean_token_accuracy": 0.18721715956926346,
      "num_tokens": 74893065.0,
      "step": 34185
    },
    {
      "entropy": 5.592374324798584,
      "epoch": 3.393211591901548,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003878741439796918,
      "loss": 4.9736,
      "mean_token_accuracy": 0.19070639312267304,
      "num_tokens": 74903806.0,
      "step": 34190
    },
    {
      "entropy": 5.756089973449707,
      "epoch": 3.3937078205637157,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003878434920962386,
      "loss": 5.1045,
      "mean_token_accuracy": 0.18428581804037095,
      "num_tokens": 74915025.0,
      "step": 34195
    },
    {
      "entropy": 5.705061149597168,
      "epoch": 3.394204049225883,
      "grad_norm": 1.0,
      "learning_rate": 0.00038781283741451,
      "loss": 5.0305,
      "mean_token_accuracy": 0.1814444914460182,
      "num_tokens": 74926438.0,
      "step": 34200
    },
    {
      "entropy": 5.701881837844849,
      "epoch": 3.3947002778880506,
      "grad_norm": 1.296875,
      "learning_rate": 0.00038778217993526616,
      "loss": 4.9909,
      "mean_token_accuracy": 0.18721581548452376,
      "num_tokens": 74937707.0,
      "step": 34205
    },
    {
      "entropy": 5.687070322036743,
      "epoch": 3.395196506550218,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003877515196592672,
      "loss": 4.9649,
      "mean_token_accuracy": 0.1872158870100975,
      "num_tokens": 74947436.0,
      "step": 34210
    },
    {
      "entropy": 5.6985969066619875,
      "epoch": 3.395692735212386,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038772085658727353,
      "loss": 5.0487,
      "mean_token_accuracy": 0.18515178859233855,
      "num_tokens": 74958707.0,
      "step": 34215
    },
    {
      "entropy": 5.663392496109009,
      "epoch": 3.3961889638745535,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038769019072004555,
      "loss": 5.0007,
      "mean_token_accuracy": 0.18968628644943236,
      "num_tokens": 74970145.0,
      "step": 34220
    },
    {
      "entropy": 5.6560650825500485,
      "epoch": 3.396685192536721,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003876595220583436,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18608374893665314,
      "num_tokens": 74982095.0,
      "step": 34225
    },
    {
      "entropy": 5.644152641296387,
      "epoch": 3.3971814211988884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003876288506029283,
      "loss": 4.9091,
      "mean_token_accuracy": 0.19121867418289185,
      "num_tokens": 74993648.0,
      "step": 34230
    },
    {
      "entropy": 5.654671096801758,
      "epoch": 3.397677649861056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003875981763545602,
      "loss": 4.9505,
      "mean_token_accuracy": 0.18830231428146363,
      "num_tokens": 75003707.0,
      "step": 34235
    },
    {
      "entropy": 5.614401865005493,
      "epoch": 3.3981738785232234,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038756749931399993,
      "loss": 4.9444,
      "mean_token_accuracy": 0.19071283489465712,
      "num_tokens": 75013029.0,
      "step": 34240
    },
    {
      "entropy": 5.646882247924805,
      "epoch": 3.398670107185391,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003875368194820083,
      "loss": 4.9854,
      "mean_token_accuracy": 0.18914397209882736,
      "num_tokens": 75024555.0,
      "step": 34245
    },
    {
      "entropy": 5.690110588073731,
      "epoch": 3.399166335847559,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000387506136859346,
      "loss": 5.0452,
      "mean_token_accuracy": 0.18313146978616715,
      "num_tokens": 75034529.0,
      "step": 34250
    },
    {
      "entropy": 5.734067440032959,
      "epoch": 3.3996625645097263,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003874754514467739,
      "loss": 5.0895,
      "mean_token_accuracy": 0.18292243033647537,
      "num_tokens": 75047733.0,
      "step": 34255
    },
    {
      "entropy": 5.710949373245239,
      "epoch": 3.4001587931718937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000387444763245053,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18897807002067565,
      "num_tokens": 75058576.0,
      "step": 34260
    },
    {
      "entropy": 5.675225877761841,
      "epoch": 3.400655021834061,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003874140722549443,
      "loss": 5.0338,
      "mean_token_accuracy": 0.18276857286691667,
      "num_tokens": 75068470.0,
      "step": 34265
    },
    {
      "entropy": 5.627457904815674,
      "epoch": 3.4011512504962287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003873833784772088,
      "loss": 4.9137,
      "mean_token_accuracy": 0.1938214346766472,
      "num_tokens": 75079383.0,
      "step": 34270
    },
    {
      "entropy": 5.69977912902832,
      "epoch": 3.401647479158396,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003873526819126077,
      "loss": 5.0358,
      "mean_token_accuracy": 0.1796089753508568,
      "num_tokens": 75089474.0,
      "step": 34275
    },
    {
      "entropy": 5.648632621765136,
      "epoch": 3.4021437078205636,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003873219825619022,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18893240839242936,
      "num_tokens": 75100286.0,
      "step": 34280
    },
    {
      "entropy": 5.727016973495483,
      "epoch": 3.402639936482731,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003872912804258536,
      "loss": 5.0589,
      "mean_token_accuracy": 0.1852736845612526,
      "num_tokens": 75111916.0,
      "step": 34285
    },
    {
      "entropy": 5.687526273727417,
      "epoch": 3.4031361651448986,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038726057550522307,
      "loss": 4.9412,
      "mean_token_accuracy": 0.1919090062379837,
      "num_tokens": 75122231.0,
      "step": 34290
    },
    {
      "entropy": 5.765628623962402,
      "epoch": 3.403632393807066,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038722986780077227,
      "loss": 5.0896,
      "mean_token_accuracy": 0.1803779974579811,
      "num_tokens": 75133289.0,
      "step": 34295
    },
    {
      "entropy": 5.64514389038086,
      "epoch": 3.404128622469234,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038719915731326264,
      "loss": 4.9548,
      "mean_token_accuracy": 0.19745618849992752,
      "num_tokens": 75143409.0,
      "step": 34300
    },
    {
      "entropy": 5.676765251159668,
      "epoch": 3.4046248511314015,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038716844404345556,
      "loss": 5.057,
      "mean_token_accuracy": 0.19041226655244828,
      "num_tokens": 75154848.0,
      "step": 34305
    },
    {
      "entropy": 5.682065677642822,
      "epoch": 3.405121079793569,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038713772799211284,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18254928886890412,
      "num_tokens": 75166861.0,
      "step": 34310
    },
    {
      "entropy": 5.700848865509033,
      "epoch": 3.4056173084557364,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003871070091599961,
      "loss": 4.9696,
      "mean_token_accuracy": 0.19415232986211778,
      "num_tokens": 75176934.0,
      "step": 34315
    },
    {
      "entropy": 5.613965272903442,
      "epoch": 3.406113537117904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038707628754786706,
      "loss": 4.9239,
      "mean_token_accuracy": 0.19031272381544112,
      "num_tokens": 75187604.0,
      "step": 34320
    },
    {
      "entropy": 5.620761489868164,
      "epoch": 3.4066097657800714,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038704556315648757,
      "loss": 4.9083,
      "mean_token_accuracy": 0.19223257452249526,
      "num_tokens": 75198524.0,
      "step": 34325
    },
    {
      "entropy": 5.691309690475464,
      "epoch": 3.407105994442239,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003870148359866196,
      "loss": 5.0102,
      "mean_token_accuracy": 0.19071080833673476,
      "num_tokens": 75209857.0,
      "step": 34330
    },
    {
      "entropy": 5.6728208541870115,
      "epoch": 3.4076022231044067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000386984106039025,
      "loss": 5.0558,
      "mean_token_accuracy": 0.18862020671367646,
      "num_tokens": 75220912.0,
      "step": 34335
    },
    {
      "entropy": 5.6204001903533936,
      "epoch": 3.408098451766574,
      "grad_norm": 1.1875,
      "learning_rate": 0.000386953373314466,
      "loss": 4.9471,
      "mean_token_accuracy": 0.1970468908548355,
      "num_tokens": 75232257.0,
      "step": 34340
    },
    {
      "entropy": 5.683148765563965,
      "epoch": 3.4085946804287417,
      "grad_norm": 1.125,
      "learning_rate": 0.00038692263781370446,
      "loss": 5.0376,
      "mean_token_accuracy": 0.18430572897195815,
      "num_tokens": 75244147.0,
      "step": 34345
    },
    {
      "entropy": 5.723097038269043,
      "epoch": 3.409090909090909,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038689189953750275,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18676792234182357,
      "num_tokens": 75255910.0,
      "step": 34350
    },
    {
      "entropy": 5.671143579483032,
      "epoch": 3.4095871377530766,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038686115848662303,
      "loss": 5.0177,
      "mean_token_accuracy": 0.187236787378788,
      "num_tokens": 75267611.0,
      "step": 34355
    },
    {
      "entropy": 5.628881883621216,
      "epoch": 3.410083366415244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003868304146618276,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18118802309036255,
      "num_tokens": 75278835.0,
      "step": 34360
    },
    {
      "entropy": 5.729888391494751,
      "epoch": 3.4105795950774116,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038679966806387886,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1875419571995735,
      "num_tokens": 75289317.0,
      "step": 34365
    },
    {
      "entropy": 5.668822240829468,
      "epoch": 3.411075823739579,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003867689186935393,
      "loss": 5.0631,
      "mean_token_accuracy": 0.19112553596496581,
      "num_tokens": 75302304.0,
      "step": 34370
    },
    {
      "entropy": 5.719882822036743,
      "epoch": 3.4115720524017465,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003867381665515712,
      "loss": 5.0043,
      "mean_token_accuracy": 0.1905020996928215,
      "num_tokens": 75312081.0,
      "step": 34375
    },
    {
      "entropy": 5.615127182006836,
      "epoch": 3.412068281063914,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003867074116387376,
      "loss": 4.9286,
      "mean_token_accuracy": 0.19096726477146148,
      "num_tokens": 75322948.0,
      "step": 34380
    },
    {
      "entropy": 5.661934089660645,
      "epoch": 3.412564509726082,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038667665395580083,
      "loss": 5.0003,
      "mean_token_accuracy": 0.18785174936056137,
      "num_tokens": 75333817.0,
      "step": 34385
    },
    {
      "entropy": 5.650663137435913,
      "epoch": 3.4130607383882494,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003866458935035236,
      "loss": 5.0041,
      "mean_token_accuracy": 0.18518042117357253,
      "num_tokens": 75344014.0,
      "step": 34390
    },
    {
      "entropy": 5.702540254592895,
      "epoch": 3.413556967050417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038661513028266886,
      "loss": 4.9927,
      "mean_token_accuracy": 0.18984459489583969,
      "num_tokens": 75355350.0,
      "step": 34395
    },
    {
      "entropy": 5.695373153686523,
      "epoch": 3.4140531957125844,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003865843642939994,
      "loss": 5.0195,
      "mean_token_accuracy": 0.18346207737922668,
      "num_tokens": 75364633.0,
      "step": 34400
    },
    {
      "entropy": 5.644761610031128,
      "epoch": 3.414549424374752,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038655359553827807,
      "loss": 4.935,
      "mean_token_accuracy": 0.19111986756324767,
      "num_tokens": 75375103.0,
      "step": 34405
    },
    {
      "entropy": 5.665784788131714,
      "epoch": 3.4150456530369193,
      "grad_norm": 1.234375,
      "learning_rate": 0.000386522824016268,
      "loss": 5.0916,
      "mean_token_accuracy": 0.18418586403131484,
      "num_tokens": 75386112.0,
      "step": 34410
    },
    {
      "entropy": 5.596890592575074,
      "epoch": 3.415541881699087,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038649204972873225,
      "loss": 4.8841,
      "mean_token_accuracy": 0.19034783095121383,
      "num_tokens": 75396744.0,
      "step": 34415
    },
    {
      "entropy": 5.713328266143799,
      "epoch": 3.4160381103612547,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003864612726764339,
      "loss": 5.0482,
      "mean_token_accuracy": 0.18508583009243013,
      "num_tokens": 75408706.0,
      "step": 34420
    },
    {
      "entropy": 5.728656482696533,
      "epoch": 3.416534339023422,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003864304928601361,
      "loss": 4.978,
      "mean_token_accuracy": 0.186972214281559,
      "num_tokens": 75419036.0,
      "step": 34425
    },
    {
      "entropy": 5.63287878036499,
      "epoch": 3.4170305676855897,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003863997102806023,
      "loss": 4.9689,
      "mean_token_accuracy": 0.19351246803998948,
      "num_tokens": 75429415.0,
      "step": 34430
    },
    {
      "entropy": 5.602952718734741,
      "epoch": 3.417526796347757,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003863689249385956,
      "loss": 4.8975,
      "mean_token_accuracy": 0.2000804290175438,
      "num_tokens": 75441077.0,
      "step": 34435
    },
    {
      "entropy": 5.709703207015991,
      "epoch": 3.4180230250099246,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038633813683487954,
      "loss": 5.0905,
      "mean_token_accuracy": 0.18384968787431716,
      "num_tokens": 75452516.0,
      "step": 34440
    },
    {
      "entropy": 5.667143440246582,
      "epoch": 3.418519253672092,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003863073459702177,
      "loss": 5.0169,
      "mean_token_accuracy": 0.18707888275384904,
      "num_tokens": 75463205.0,
      "step": 34445
    },
    {
      "entropy": 5.705711841583252,
      "epoch": 3.4190154823342596,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038627655234537345,
      "loss": 5.0245,
      "mean_token_accuracy": 0.18821119964122773,
      "num_tokens": 75474201.0,
      "step": 34450
    },
    {
      "entropy": 5.633301830291748,
      "epoch": 3.419511710996427,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038624575596111056,
      "loss": 4.964,
      "mean_token_accuracy": 0.19614984542131425,
      "num_tokens": 75484310.0,
      "step": 34455
    },
    {
      "entropy": 5.620262670516968,
      "epoch": 3.4200079396585945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038621495681819256,
      "loss": 4.9279,
      "mean_token_accuracy": 0.1864564299583435,
      "num_tokens": 75494583.0,
      "step": 34460
    },
    {
      "entropy": 5.68745002746582,
      "epoch": 3.420504168320762,
      "grad_norm": 1.125,
      "learning_rate": 0.0003861841549173833,
      "loss": 5.0667,
      "mean_token_accuracy": 0.18766067177057266,
      "num_tokens": 75505176.0,
      "step": 34465
    },
    {
      "entropy": 5.737483930587769,
      "epoch": 3.42100039698293,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003861533502594466,
      "loss": 5.0182,
      "mean_token_accuracy": 0.18775905966758727,
      "num_tokens": 75514795.0,
      "step": 34470
    },
    {
      "entropy": 5.713310956954956,
      "epoch": 3.4214966256450974,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003861225428451463,
      "loss": 5.009,
      "mean_token_accuracy": 0.1855470985174179,
      "num_tokens": 75525975.0,
      "step": 34475
    },
    {
      "entropy": 5.670758819580078,
      "epoch": 3.421992854307265,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003860917326752464,
      "loss": 4.9765,
      "mean_token_accuracy": 0.18959172815084457,
      "num_tokens": 75536648.0,
      "step": 34480
    },
    {
      "entropy": 5.712639999389649,
      "epoch": 3.4224890829694323,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038606091975051097,
      "loss": 5.1529,
      "mean_token_accuracy": 0.17965207397937774,
      "num_tokens": 75547724.0,
      "step": 34485
    },
    {
      "entropy": 5.675191879272461,
      "epoch": 3.4229853116316,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00038603010407170393,
      "loss": 5.0506,
      "mean_token_accuracy": 0.18283447176218032,
      "num_tokens": 75559061.0,
      "step": 34490
    },
    {
      "entropy": 5.666395044326782,
      "epoch": 3.4234815402937673,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003859992856395896,
      "loss": 4.9994,
      "mean_token_accuracy": 0.1794769674539566,
      "num_tokens": 75569744.0,
      "step": 34495
    },
    {
      "entropy": 5.74343318939209,
      "epoch": 3.4239777689559348,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003859684644549322,
      "loss": 5.1033,
      "mean_token_accuracy": 0.1857685223221779,
      "num_tokens": 75579780.0,
      "step": 34500
    },
    {
      "entropy": 5.673283910751342,
      "epoch": 3.4244739976181022,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003859376405184959,
      "loss": 4.9854,
      "mean_token_accuracy": 0.1857822135090828,
      "num_tokens": 75591526.0,
      "step": 34505
    },
    {
      "entropy": 5.713583469390869,
      "epoch": 3.42497022628027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003859068138310453,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1866588920354843,
      "num_tokens": 75602573.0,
      "step": 34510
    },
    {
      "entropy": 5.657305765151977,
      "epoch": 3.4254664549424376,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038587598439334466,
      "loss": 5.037,
      "mean_token_accuracy": 0.18470027595758437,
      "num_tokens": 75613774.0,
      "step": 34515
    },
    {
      "entropy": 5.6225423336029055,
      "epoch": 3.425962683604605,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038584515220615843,
      "loss": 4.9573,
      "mean_token_accuracy": 0.19574080109596254,
      "num_tokens": 75624435.0,
      "step": 34520
    },
    {
      "entropy": 5.634544420242309,
      "epoch": 3.4264589122667726,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038581431727025136,
      "loss": 4.9547,
      "mean_token_accuracy": 0.19107698202133178,
      "num_tokens": 75634429.0,
      "step": 34525
    },
    {
      "entropy": 5.6354655742645265,
      "epoch": 3.42695514092894,
      "grad_norm": 1.0625,
      "learning_rate": 0.000385783479586388,
      "loss": 4.9335,
      "mean_token_accuracy": 0.19038204550743104,
      "num_tokens": 75646066.0,
      "step": 34530
    },
    {
      "entropy": 5.651575088500977,
      "epoch": 3.4274513695911075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003857526391553331,
      "loss": 5.0052,
      "mean_token_accuracy": 0.19446648508310319,
      "num_tokens": 75655718.0,
      "step": 34535
    },
    {
      "entropy": 5.646937513351441,
      "epoch": 3.427947598253275,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038572179597785134,
      "loss": 4.9372,
      "mean_token_accuracy": 0.19541468918323518,
      "num_tokens": 75666581.0,
      "step": 34540
    },
    {
      "entropy": 5.593952560424805,
      "epoch": 3.4284438269154425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038569095005470765,
      "loss": 4.9356,
      "mean_token_accuracy": 0.19473477602005004,
      "num_tokens": 75676405.0,
      "step": 34545
    },
    {
      "entropy": 5.561600542068481,
      "epoch": 3.42894005557761,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038566010138666693,
      "loss": 4.9043,
      "mean_token_accuracy": 0.19537552148103715,
      "num_tokens": 75687723.0,
      "step": 34550
    },
    {
      "entropy": 5.649035453796387,
      "epoch": 3.429436284239778,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038562924997449414,
      "loss": 5.0157,
      "mean_token_accuracy": 0.1887913390994072,
      "num_tokens": 75699461.0,
      "step": 34555
    },
    {
      "entropy": 5.7165600776672365,
      "epoch": 3.4299325129019453,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003855983958189543,
      "loss": 5.0514,
      "mean_token_accuracy": 0.18115678578615188,
      "num_tokens": 75711679.0,
      "step": 34560
    },
    {
      "entropy": 5.735452795028687,
      "epoch": 3.430428741564113,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003855675389208126,
      "loss": 5.094,
      "mean_token_accuracy": 0.1796677902340889,
      "num_tokens": 75721027.0,
      "step": 34565
    },
    {
      "entropy": 5.641846084594727,
      "epoch": 3.4309249702262803,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003855366792808341,
      "loss": 4.98,
      "mean_token_accuracy": 0.18609365671873093,
      "num_tokens": 75731194.0,
      "step": 34570
    },
    {
      "entropy": 5.745919895172119,
      "epoch": 3.4314211988884478,
      "grad_norm": 1.328125,
      "learning_rate": 0.00038550581689978427,
      "loss": 5.0361,
      "mean_token_accuracy": 0.1834784507751465,
      "num_tokens": 75741730.0,
      "step": 34575
    },
    {
      "entropy": 5.695012140274048,
      "epoch": 3.4319174275506152,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038547495177842817,
      "loss": 5.0296,
      "mean_token_accuracy": 0.18602571487426758,
      "num_tokens": 75752980.0,
      "step": 34580
    },
    {
      "entropy": 5.675933742523194,
      "epoch": 3.4324136562127827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003854440839175314,
      "loss": 5.0692,
      "mean_token_accuracy": 0.1817755714058876,
      "num_tokens": 75764015.0,
      "step": 34585
    },
    {
      "entropy": 5.688815689086914,
      "epoch": 3.43290988487495,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038541321331785934,
      "loss": 5.0664,
      "mean_token_accuracy": 0.1789198637008667,
      "num_tokens": 75775267.0,
      "step": 34590
    },
    {
      "entropy": 5.7171492099761965,
      "epoch": 3.433406113537118,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038538233998017745,
      "loss": 4.9979,
      "mean_token_accuracy": 0.19189073890447617,
      "num_tokens": 75785456.0,
      "step": 34595
    },
    {
      "entropy": 5.61717586517334,
      "epoch": 3.4339023421992856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00038535146390525143,
      "loss": 4.9763,
      "mean_token_accuracy": 0.18910711854696274,
      "num_tokens": 75795563.0,
      "step": 34600
    },
    {
      "entropy": 5.64457893371582,
      "epoch": 3.434398570861453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038532058509384694,
      "loss": 4.9993,
      "mean_token_accuracy": 0.18637934029102327,
      "num_tokens": 75807068.0,
      "step": 34605
    },
    {
      "entropy": 5.648428583145142,
      "epoch": 3.4348947995236205,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003852897035467296,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18599336296319963,
      "num_tokens": 75818354.0,
      "step": 34610
    },
    {
      "entropy": 5.677057266235352,
      "epoch": 3.435391028185788,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038525881926466525,
      "loss": 5.0044,
      "mean_token_accuracy": 0.1907135397195816,
      "num_tokens": 75829382.0,
      "step": 34615
    },
    {
      "entropy": 5.627370691299438,
      "epoch": 3.4358872568479555,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038522793224841977,
      "loss": 4.9335,
      "mean_token_accuracy": 0.18767326176166535,
      "num_tokens": 75839278.0,
      "step": 34620
    },
    {
      "entropy": 5.64017505645752,
      "epoch": 3.436383485510123,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00038519704249875917,
      "loss": 4.9729,
      "mean_token_accuracy": 0.18601458966732026,
      "num_tokens": 75851074.0,
      "step": 34625
    },
    {
      "entropy": 5.686547422409058,
      "epoch": 3.4368797141722904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003851661500164493,
      "loss": 5.0257,
      "mean_token_accuracy": 0.18708205074071885,
      "num_tokens": 75862747.0,
      "step": 34630
    },
    {
      "entropy": 5.614958000183106,
      "epoch": 3.437375942834458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003851352548022563,
      "loss": 4.981,
      "mean_token_accuracy": 0.18855116814374923,
      "num_tokens": 75872807.0,
      "step": 34635
    },
    {
      "entropy": 5.63934440612793,
      "epoch": 3.437872171496626,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003851043568569463,
      "loss": 5.0023,
      "mean_token_accuracy": 0.18804893046617507,
      "num_tokens": 75883630.0,
      "step": 34640
    },
    {
      "entropy": 5.7207029342651365,
      "epoch": 3.4383684001587933,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003850734561812855,
      "loss": 5.0391,
      "mean_token_accuracy": 0.1856728196144104,
      "num_tokens": 75895147.0,
      "step": 34645
    },
    {
      "entropy": 5.6733808517456055,
      "epoch": 3.438864628820961,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038504255277604016,
      "loss": 5.0188,
      "mean_token_accuracy": 0.18825910240411758,
      "num_tokens": 75906685.0,
      "step": 34650
    },
    {
      "entropy": 5.571512126922608,
      "epoch": 3.4393608574831283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003850116466419766,
      "loss": 4.8954,
      "mean_token_accuracy": 0.19887349307537078,
      "num_tokens": 75916859.0,
      "step": 34655
    },
    {
      "entropy": 5.674856090545655,
      "epoch": 3.4398570861452957,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003849807377798614,
      "loss": 5.1079,
      "mean_token_accuracy": 0.1801234319806099,
      "num_tokens": 75927138.0,
      "step": 34660
    },
    {
      "entropy": 5.6671459674835205,
      "epoch": 3.440353314807463,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038494982619046076,
      "loss": 4.9895,
      "mean_token_accuracy": 0.1858809620141983,
      "num_tokens": 75939542.0,
      "step": 34665
    },
    {
      "entropy": 5.673594045639038,
      "epoch": 3.4408495434696307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003849189118745414,
      "loss": 5.059,
      "mean_token_accuracy": 0.18838145434856415,
      "num_tokens": 75951252.0,
      "step": 34670
    },
    {
      "entropy": 5.668855094909668,
      "epoch": 3.441345772131798,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003848879948328699,
      "loss": 4.9972,
      "mean_token_accuracy": 0.18264940977096558,
      "num_tokens": 75962440.0,
      "step": 34675
    },
    {
      "entropy": 5.697276639938354,
      "epoch": 3.441842000793966,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003848570750662129,
      "loss": 4.9413,
      "mean_token_accuracy": 0.19302009046077728,
      "num_tokens": 75971177.0,
      "step": 34680
    },
    {
      "entropy": 5.632142543792725,
      "epoch": 3.4423382294561335,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003848261525753371,
      "loss": 5.062,
      "mean_token_accuracy": 0.18365033715963364,
      "num_tokens": 75983112.0,
      "step": 34685
    },
    {
      "entropy": 5.651433420181275,
      "epoch": 3.442834458118301,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003847952273610094,
      "loss": 5.0153,
      "mean_token_accuracy": 0.18824541866779326,
      "num_tokens": 75994347.0,
      "step": 34690
    },
    {
      "entropy": 5.686566162109375,
      "epoch": 3.4433306867804685,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003847642994239968,
      "loss": 4.9941,
      "mean_token_accuracy": 0.19162528365850448,
      "num_tokens": 76004712.0,
      "step": 34695
    },
    {
      "entropy": 5.709883165359497,
      "epoch": 3.443826915442636,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038473336876506585,
      "loss": 5.0593,
      "mean_token_accuracy": 0.1827565461397171,
      "num_tokens": 76014586.0,
      "step": 34700
    },
    {
      "entropy": 5.664228916168213,
      "epoch": 3.4443231441048034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038470243538498393,
      "loss": 4.9665,
      "mean_token_accuracy": 0.19279976934194565,
      "num_tokens": 76024132.0,
      "step": 34705
    },
    {
      "entropy": 5.7010386943817135,
      "epoch": 3.444819372766971,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038467149928451803,
      "loss": 5.0471,
      "mean_token_accuracy": 0.1802176207304001,
      "num_tokens": 76035122.0,
      "step": 34710
    },
    {
      "entropy": 5.696985530853271,
      "epoch": 3.4453156014291384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038464056046443523,
      "loss": 5.0833,
      "mean_token_accuracy": 0.18344300240278244,
      "num_tokens": 76047027.0,
      "step": 34715
    },
    {
      "entropy": 5.618285989761352,
      "epoch": 3.445811830091306,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003846096189255027,
      "loss": 4.9808,
      "mean_token_accuracy": 0.18835756331682205,
      "num_tokens": 76057743.0,
      "step": 34720
    },
    {
      "entropy": 5.567685937881469,
      "epoch": 3.446308058753474,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038457867466848796,
      "loss": 4.8934,
      "mean_token_accuracy": 0.194198577105999,
      "num_tokens": 76068433.0,
      "step": 34725
    },
    {
      "entropy": 5.715467739105224,
      "epoch": 3.4468042874156413,
      "grad_norm": 1.125,
      "learning_rate": 0.0003845477276941582,
      "loss": 5.0387,
      "mean_token_accuracy": 0.1872687190771103,
      "num_tokens": 76080681.0,
      "step": 34730
    },
    {
      "entropy": 5.664191246032715,
      "epoch": 3.4473005160778087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003845167780032807,
      "loss": 5.0727,
      "mean_token_accuracy": 0.17885841727256774,
      "num_tokens": 76092584.0,
      "step": 34735
    },
    {
      "entropy": 5.619916343688965,
      "epoch": 3.447796744739976,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00038448582559662324,
      "loss": 4.9284,
      "mean_token_accuracy": 0.19324991106987,
      "num_tokens": 76103629.0,
      "step": 34740
    },
    {
      "entropy": 5.671837663650512,
      "epoch": 3.4482929734021437,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038445487047495313,
      "loss": 4.9817,
      "mean_token_accuracy": 0.19219615906476975,
      "num_tokens": 76114069.0,
      "step": 34745
    },
    {
      "entropy": 5.6643640995025635,
      "epoch": 3.448789202064311,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003844239126390382,
      "loss": 5.0381,
      "mean_token_accuracy": 0.19170139580965043,
      "num_tokens": 76125896.0,
      "step": 34750
    },
    {
      "entropy": 5.675234031677246,
      "epoch": 3.4492854307264786,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038439295208964595,
      "loss": 4.9967,
      "mean_token_accuracy": 0.18869018107652663,
      "num_tokens": 76137066.0,
      "step": 34755
    },
    {
      "entropy": 5.639168310165405,
      "epoch": 3.449781659388646,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003843619888275442,
      "loss": 4.9303,
      "mean_token_accuracy": 0.19569581151008605,
      "num_tokens": 76148778.0,
      "step": 34760
    },
    {
      "entropy": 5.679354858398438,
      "epoch": 3.450277888050814,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038433102285350074,
      "loss": 5.0187,
      "mean_token_accuracy": 0.18268633633852005,
      "num_tokens": 76158760.0,
      "step": 34765
    },
    {
      "entropy": 5.680347204208374,
      "epoch": 3.4507741167129815,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038430005416828354,
      "loss": 4.9549,
      "mean_token_accuracy": 0.194074085354805,
      "num_tokens": 76169459.0,
      "step": 34770
    },
    {
      "entropy": 5.651906442642212,
      "epoch": 3.451270345375149,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038426908277266054,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18764864057302474,
      "num_tokens": 76179074.0,
      "step": 34775
    },
    {
      "entropy": 5.6680456638336185,
      "epoch": 3.4517665740373165,
      "grad_norm": 1.25,
      "learning_rate": 0.00038423810866739975,
      "loss": 5.0122,
      "mean_token_accuracy": 0.19188939929008483,
      "num_tokens": 76189470.0,
      "step": 34780
    },
    {
      "entropy": 5.67020058631897,
      "epoch": 3.452262802699484,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038420713185326926,
      "loss": 4.9709,
      "mean_token_accuracy": 0.1892136663198471,
      "num_tokens": 76200427.0,
      "step": 34785
    },
    {
      "entropy": 5.66608419418335,
      "epoch": 3.4527590313616514,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003841761523310372,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18594156205654144,
      "num_tokens": 76211132.0,
      "step": 34790
    },
    {
      "entropy": 5.653756332397461,
      "epoch": 3.453255260023819,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003841451701014718,
      "loss": 5.0226,
      "mean_token_accuracy": 0.1891536980867386,
      "num_tokens": 76222661.0,
      "step": 34795
    },
    {
      "entropy": 5.717365360260009,
      "epoch": 3.4537514886859864,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038411418516534135,
      "loss": 4.9953,
      "mean_token_accuracy": 0.19274254888296127,
      "num_tokens": 76232727.0,
      "step": 34800
    },
    {
      "entropy": 5.730730390548706,
      "epoch": 3.454247717348154,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038408319752341425,
      "loss": 5.0213,
      "mean_token_accuracy": 0.19189198315143585,
      "num_tokens": 76242783.0,
      "step": 34805
    },
    {
      "entropy": 5.6209276676177975,
      "epoch": 3.4547439460103213,
      "grad_norm": 1.15625,
      "learning_rate": 0.000384052207176459,
      "loss": 4.9012,
      "mean_token_accuracy": 0.1823547288775444,
      "num_tokens": 76254223.0,
      "step": 34810
    },
    {
      "entropy": 5.716167640686035,
      "epoch": 3.4552401746724892,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003840212141252438,
      "loss": 5.0203,
      "mean_token_accuracy": 0.18807147592306137,
      "num_tokens": 76265094.0,
      "step": 34815
    },
    {
      "entropy": 5.671154737472534,
      "epoch": 3.4557364033346567,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038399021837053754,
      "loss": 4.9987,
      "mean_token_accuracy": 0.18382764905691146,
      "num_tokens": 76276722.0,
      "step": 34820
    },
    {
      "entropy": 5.635523509979248,
      "epoch": 3.456232631996824,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003839592199131088,
      "loss": 4.9633,
      "mean_token_accuracy": 0.18831268548965455,
      "num_tokens": 76288027.0,
      "step": 34825
    },
    {
      "entropy": 5.672514629364014,
      "epoch": 3.4567288606589917,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038392821875372603,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2040116384625435,
      "num_tokens": 76299147.0,
      "step": 34830
    },
    {
      "entropy": 5.693408727645874,
      "epoch": 3.457225089321159,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003838972148931582,
      "loss": 5.1155,
      "mean_token_accuracy": 0.18161285668611526,
      "num_tokens": 76309375.0,
      "step": 34835
    },
    {
      "entropy": 5.676216268539429,
      "epoch": 3.4577213179833266,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003838662083321741,
      "loss": 4.9417,
      "mean_token_accuracy": 0.18866847157478334,
      "num_tokens": 76319953.0,
      "step": 34840
    },
    {
      "entropy": 5.71015944480896,
      "epoch": 3.458217546645494,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038383519907154263,
      "loss": 5.0245,
      "mean_token_accuracy": 0.1906597286462784,
      "num_tokens": 76331525.0,
      "step": 34845
    },
    {
      "entropy": 5.631751775741577,
      "epoch": 3.458713775307662,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038380418711203275,
      "loss": 4.918,
      "mean_token_accuracy": 0.20764069259166718,
      "num_tokens": 76343290.0,
      "step": 34850
    },
    {
      "entropy": 5.700077533721924,
      "epoch": 3.4592100039698295,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038377317245441345,
      "loss": 5.1364,
      "mean_token_accuracy": 0.18033249378204347,
      "num_tokens": 76354993.0,
      "step": 34855
    },
    {
      "entropy": 5.695354461669922,
      "epoch": 3.459706232631997,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038374215509945386,
      "loss": 5.0111,
      "mean_token_accuracy": 0.19027825742959975,
      "num_tokens": 76365206.0,
      "step": 34860
    },
    {
      "entropy": 5.679875373840332,
      "epoch": 3.4602024612941644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003837111350479232,
      "loss": 5.0343,
      "mean_token_accuracy": 0.18465604931116103,
      "num_tokens": 76376549.0,
      "step": 34865
    },
    {
      "entropy": 5.652709054946899,
      "epoch": 3.460698689956332,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038368011230059055,
      "loss": 4.9703,
      "mean_token_accuracy": 0.1869020238518715,
      "num_tokens": 76387654.0,
      "step": 34870
    },
    {
      "entropy": 5.784362268447876,
      "epoch": 3.4611949186184994,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038364908685822543,
      "loss": 5.0365,
      "mean_token_accuracy": 0.18833450525999068,
      "num_tokens": 76399462.0,
      "step": 34875
    },
    {
      "entropy": 5.73277006149292,
      "epoch": 3.461691147280667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038361805872159693,
      "loss": 5.0029,
      "mean_token_accuracy": 0.19368977546691896,
      "num_tokens": 76411394.0,
      "step": 34880
    },
    {
      "entropy": 5.718883657455445,
      "epoch": 3.4621873759428343,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003835870278914747,
      "loss": 5.06,
      "mean_token_accuracy": 0.1847855642437935,
      "num_tokens": 76423251.0,
      "step": 34885
    },
    {
      "entropy": 5.6885058879852295,
      "epoch": 3.462683604605002,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003835559943686282,
      "loss": 4.9492,
      "mean_token_accuracy": 0.1895807221531868,
      "num_tokens": 76433828.0,
      "step": 34890
    },
    {
      "entropy": 5.642913675308227,
      "epoch": 3.4631798332671693,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003835249581538269,
      "loss": 4.9269,
      "mean_token_accuracy": 0.19748566448688507,
      "num_tokens": 76445006.0,
      "step": 34895
    },
    {
      "entropy": 5.60023512840271,
      "epoch": 3.463676061929337,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00038349391924784055,
      "loss": 4.9822,
      "mean_token_accuracy": 0.1906496539711952,
      "num_tokens": 76456296.0,
      "step": 34900
    },
    {
      "entropy": 5.586465883255005,
      "epoch": 3.4641722905915047,
      "grad_norm": 1.125,
      "learning_rate": 0.00038346287765143866,
      "loss": 4.9312,
      "mean_token_accuracy": 0.18935026079416276,
      "num_tokens": 76467330.0,
      "step": 34905
    },
    {
      "entropy": 5.7397197723388675,
      "epoch": 3.464668519253672,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038343183336539123,
      "loss": 5.0681,
      "mean_token_accuracy": 0.18291239887475969,
      "num_tokens": 76478393.0,
      "step": 34910
    },
    {
      "entropy": 5.696448278427124,
      "epoch": 3.4651647479158396,
      "grad_norm": 1.078125,
      "learning_rate": 0.000383400786390468,
      "loss": 4.9728,
      "mean_token_accuracy": 0.19425248801708223,
      "num_tokens": 76489695.0,
      "step": 34915
    },
    {
      "entropy": 5.636416578292847,
      "epoch": 3.465660976578007,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038336973672743876,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18518905341625214,
      "num_tokens": 76500831.0,
      "step": 34920
    },
    {
      "entropy": 5.629849863052368,
      "epoch": 3.4661572052401746,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038333868437707367,
      "loss": 4.9581,
      "mean_token_accuracy": 0.18982370048761368,
      "num_tokens": 76511054.0,
      "step": 34925
    },
    {
      "entropy": 5.687385559082031,
      "epoch": 3.466653433902342,
      "grad_norm": 1.125,
      "learning_rate": 0.0003833076293401425,
      "loss": 5.0008,
      "mean_token_accuracy": 0.18888331651687623,
      "num_tokens": 76522948.0,
      "step": 34930
    },
    {
      "entropy": 5.619164657592774,
      "epoch": 3.46714966256451,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003832765716174156,
      "loss": 4.9763,
      "mean_token_accuracy": 0.1900976538658142,
      "num_tokens": 76533827.0,
      "step": 34935
    },
    {
      "entropy": 5.7599669933319095,
      "epoch": 3.4676458912266774,
      "grad_norm": 1.125,
      "learning_rate": 0.0003832455112096631,
      "loss": 5.1061,
      "mean_token_accuracy": 0.18630134612321853,
      "num_tokens": 76545252.0,
      "step": 34940
    },
    {
      "entropy": 5.670439147949219,
      "epoch": 3.468142119888845,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003832144481176551,
      "loss": 4.9504,
      "mean_token_accuracy": 0.1919375792145729,
      "num_tokens": 76555774.0,
      "step": 34945
    },
    {
      "entropy": 5.65475549697876,
      "epoch": 3.4686383485510124,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038318338234216194,
      "loss": 4.9901,
      "mean_token_accuracy": 0.18696626722812654,
      "num_tokens": 76566403.0,
      "step": 34950
    },
    {
      "entropy": 5.663796663284302,
      "epoch": 3.46913457721318,
      "grad_norm": 1.5,
      "learning_rate": 0.00038315231388395403,
      "loss": 5.0066,
      "mean_token_accuracy": 0.19106864482164382,
      "num_tokens": 76578938.0,
      "step": 34955
    },
    {
      "entropy": 5.69391508102417,
      "epoch": 3.4696308058753473,
      "grad_norm": 1.25,
      "learning_rate": 0.0003831212427438018,
      "loss": 5.0576,
      "mean_token_accuracy": 0.18583184331655503,
      "num_tokens": 76589355.0,
      "step": 34960
    },
    {
      "entropy": 5.654910373687744,
      "epoch": 3.470127034537515,
      "grad_norm": 1.125,
      "learning_rate": 0.0003830901689224757,
      "loss": 4.955,
      "mean_token_accuracy": 0.19581982642412185,
      "num_tokens": 76601674.0,
      "step": 34965
    },
    {
      "entropy": 5.735566186904907,
      "epoch": 3.4706232631996823,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038305909242074645,
      "loss": 4.9988,
      "mean_token_accuracy": 0.19464583098888397,
      "num_tokens": 76612814.0,
      "step": 34970
    },
    {
      "entropy": 5.6498082160949705,
      "epoch": 3.4711194918618498,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003830280132393844,
      "loss": 4.9812,
      "mean_token_accuracy": 0.18443281650543214,
      "num_tokens": 76623177.0,
      "step": 34975
    },
    {
      "entropy": 5.658525514602661,
      "epoch": 3.4716157205240172,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038299693137916046,
      "loss": 5.0085,
      "mean_token_accuracy": 0.18906294405460358,
      "num_tokens": 76633286.0,
      "step": 34980
    },
    {
      "entropy": 5.717557907104492,
      "epoch": 3.472111949186185,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038296584684084536,
      "loss": 5.021,
      "mean_token_accuracy": 0.19463179707527162,
      "num_tokens": 76644851.0,
      "step": 34985
    },
    {
      "entropy": 5.668530511856079,
      "epoch": 3.4726081778483526,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003829347596252099,
      "loss": 4.9621,
      "mean_token_accuracy": 0.1867661714553833,
      "num_tokens": 76655165.0,
      "step": 34990
    },
    {
      "entropy": 5.642166328430176,
      "epoch": 3.47310440651052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038290366973302493,
      "loss": 4.9651,
      "mean_token_accuracy": 0.19083913266658784,
      "num_tokens": 76666282.0,
      "step": 34995
    },
    {
      "entropy": 5.6832178115844725,
      "epoch": 3.4736006351726876,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038287257716506154,
      "loss": 5.0079,
      "mean_token_accuracy": 0.18529495000839233,
      "num_tokens": 76675600.0,
      "step": 35000
    },
    {
      "entropy": 5.6849809169769285,
      "epoch": 3.474096863834855,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038284148192209064,
      "loss": 4.9884,
      "mean_token_accuracy": 0.1914227083325386,
      "num_tokens": 76687412.0,
      "step": 35005
    },
    {
      "entropy": 5.727017116546631,
      "epoch": 3.4745930924970225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00038281038400488326,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18463520258665084,
      "num_tokens": 76698799.0,
      "step": 35010
    },
    {
      "entropy": 5.713246488571167,
      "epoch": 3.47508932115919,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038277928341421086,
      "loss": 5.0225,
      "mean_token_accuracy": 0.18188643008470534,
      "num_tokens": 76709672.0,
      "step": 35015
    },
    {
      "entropy": 5.635591745376587,
      "epoch": 3.475585549821358,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003827481801508444,
      "loss": 4.9776,
      "mean_token_accuracy": 0.18985352665185928,
      "num_tokens": 76720641.0,
      "step": 35020
    },
    {
      "entropy": 5.619448184967041,
      "epoch": 3.4760817784835254,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038271707421555517,
      "loss": 4.9297,
      "mean_token_accuracy": 0.1926017001271248,
      "num_tokens": 76731173.0,
      "step": 35025
    },
    {
      "entropy": 5.723470878601074,
      "epoch": 3.476578007145693,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003826859656091147,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18056439459323884,
      "num_tokens": 76742801.0,
      "step": 35030
    },
    {
      "entropy": 5.742950582504273,
      "epoch": 3.4770742358078603,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003826548543322943,
      "loss": 4.979,
      "mean_token_accuracy": 0.19050878435373306,
      "num_tokens": 76753088.0,
      "step": 35035
    },
    {
      "entropy": 5.683184337615967,
      "epoch": 3.477570464470028,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003826237403858654,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18915691822767258,
      "num_tokens": 76765715.0,
      "step": 35040
    },
    {
      "entropy": 5.712738370895385,
      "epoch": 3.4780666931321953,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003825926237705998,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18618490248918534,
      "num_tokens": 76775626.0,
      "step": 35045
    },
    {
      "entropy": 5.632486724853516,
      "epoch": 3.4785629217943628,
      "grad_norm": 1.125,
      "learning_rate": 0.000382561504487269,
      "loss": 4.9041,
      "mean_token_accuracy": 0.19434472620487214,
      "num_tokens": 76785357.0,
      "step": 35050
    },
    {
      "entropy": 5.66010799407959,
      "epoch": 3.4790591504565302,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038253038253664446,
      "loss": 4.9892,
      "mean_token_accuracy": 0.18974587321281433,
      "num_tokens": 76797553.0,
      "step": 35055
    },
    {
      "entropy": 5.628153657913208,
      "epoch": 3.4795553791186977,
      "grad_norm": 1.1875,
      "learning_rate": 0.00038249925791949827,
      "loss": 4.9735,
      "mean_token_accuracy": 0.19426663219928741,
      "num_tokens": 76807564.0,
      "step": 35060
    },
    {
      "entropy": 5.672848892211914,
      "epoch": 3.480051607780865,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003824681306366021,
      "loss": 5.0236,
      "mean_token_accuracy": 0.18691761940717697,
      "num_tokens": 76818012.0,
      "step": 35065
    },
    {
      "entropy": 5.682764959335327,
      "epoch": 3.480547836443033,
      "grad_norm": 1.25,
      "learning_rate": 0.00038243700068872784,
      "loss": 5.0059,
      "mean_token_accuracy": 0.18987439274787904,
      "num_tokens": 76828552.0,
      "step": 35070
    },
    {
      "entropy": 5.748184633255005,
      "epoch": 3.4810440651052006,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038240586807664745,
      "loss": 5.0067,
      "mean_token_accuracy": 0.1908783197402954,
      "num_tokens": 76839652.0,
      "step": 35075
    },
    {
      "entropy": 5.671213865280151,
      "epoch": 3.481540293767368,
      "grad_norm": 1.15625,
      "learning_rate": 0.000382374732801133,
      "loss": 4.9766,
      "mean_token_accuracy": 0.19101715832948685,
      "num_tokens": 76851563.0,
      "step": 35080
    },
    {
      "entropy": 5.635352087020874,
      "epoch": 3.4820365224295355,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038234359486295663,
      "loss": 4.952,
      "mean_token_accuracy": 0.18262072801589965,
      "num_tokens": 76862769.0,
      "step": 35085
    },
    {
      "entropy": 5.704314470291138,
      "epoch": 3.482532751091703,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003823124542628904,
      "loss": 5.0295,
      "mean_token_accuracy": 0.18871040195226668,
      "num_tokens": 76873447.0,
      "step": 35090
    },
    {
      "entropy": 5.612252855300904,
      "epoch": 3.4830289797538705,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038228131100170637,
      "loss": 4.9722,
      "mean_token_accuracy": 0.1900129184126854,
      "num_tokens": 76884187.0,
      "step": 35095
    },
    {
      "entropy": 5.69522614479065,
      "epoch": 3.483525208416038,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038225016508017717,
      "loss": 4.9966,
      "mean_token_accuracy": 0.1898225337266922,
      "num_tokens": 76895592.0,
      "step": 35100
    },
    {
      "entropy": 5.75954155921936,
      "epoch": 3.484021437078206,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000382219016499075,
      "loss": 5.0556,
      "mean_token_accuracy": 0.18709314316511155,
      "num_tokens": 76907257.0,
      "step": 35105
    },
    {
      "entropy": 5.716058254241943,
      "epoch": 3.4845176657403734,
      "grad_norm": 1.25,
      "learning_rate": 0.00038218786525917207,
      "loss": 5.0499,
      "mean_token_accuracy": 0.18579675555229186,
      "num_tokens": 76917926.0,
      "step": 35110
    },
    {
      "entropy": 5.660886287689209,
      "epoch": 3.485013894402541,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003821567113612412,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18921630531549455,
      "num_tokens": 76929060.0,
      "step": 35115
    },
    {
      "entropy": 5.6213562965393065,
      "epoch": 3.4855101230647083,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038212555480605475,
      "loss": 4.9387,
      "mean_token_accuracy": 0.1945761263370514,
      "num_tokens": 76939694.0,
      "step": 35120
    },
    {
      "entropy": 5.666436672210693,
      "epoch": 3.486006351726876,
      "grad_norm": 1.25,
      "learning_rate": 0.00038209439559438535,
      "loss": 4.9673,
      "mean_token_accuracy": 0.1965893864631653,
      "num_tokens": 76951015.0,
      "step": 35125
    },
    {
      "entropy": 5.668056774139404,
      "epoch": 3.4865025803890433,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003820632337270057,
      "loss": 5.0309,
      "mean_token_accuracy": 0.1866053447127342,
      "num_tokens": 76962606.0,
      "step": 35130
    },
    {
      "entropy": 5.671468782424927,
      "epoch": 3.4869988090512107,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038203206920468867,
      "loss": 4.9987,
      "mean_token_accuracy": 0.1882817938923836,
      "num_tokens": 76972957.0,
      "step": 35135
    },
    {
      "entropy": 5.608737802505493,
      "epoch": 3.487495037713378,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038200090202820686,
      "loss": 4.9664,
      "mean_token_accuracy": 0.1941975921392441,
      "num_tokens": 76983396.0,
      "step": 35140
    },
    {
      "entropy": 5.619685888290405,
      "epoch": 3.4879912663755457,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038196973219833323,
      "loss": 4.934,
      "mean_token_accuracy": 0.19014874398708342,
      "num_tokens": 76994340.0,
      "step": 35145
    },
    {
      "entropy": 5.691436624526977,
      "epoch": 3.488487495037713,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038193855971584086,
      "loss": 4.9861,
      "mean_token_accuracy": 0.188553187251091,
      "num_tokens": 77004723.0,
      "step": 35150
    },
    {
      "entropy": 5.6587779998779295,
      "epoch": 3.488983723699881,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038190738458150256,
      "loss": 5.0726,
      "mean_token_accuracy": 0.1808892384171486,
      "num_tokens": 77017031.0,
      "step": 35155
    },
    {
      "entropy": 5.665141296386719,
      "epoch": 3.4894799523620486,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038187620679609144,
      "loss": 4.9477,
      "mean_token_accuracy": 0.19395169168710708,
      "num_tokens": 77028629.0,
      "step": 35160
    },
    {
      "entropy": 5.6731117248535154,
      "epoch": 3.489976181024216,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038184502636038074,
      "loss": 5.0247,
      "mean_token_accuracy": 0.19018638432025908,
      "num_tokens": 77037811.0,
      "step": 35165
    },
    {
      "entropy": 5.753475379943848,
      "epoch": 3.4904724096863835,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003818138432751436,
      "loss": 5.1314,
      "mean_token_accuracy": 0.18372822403907776,
      "num_tokens": 77049213.0,
      "step": 35170
    },
    {
      "entropy": 5.710118532180786,
      "epoch": 3.490968638348551,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003817826575411533,
      "loss": 5.0199,
      "mean_token_accuracy": 0.18329748809337615,
      "num_tokens": 77059090.0,
      "step": 35175
    },
    {
      "entropy": 5.689910793304444,
      "epoch": 3.4914648670107185,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038175146915918323,
      "loss": 4.9826,
      "mean_token_accuracy": 0.19015703350305557,
      "num_tokens": 77070092.0,
      "step": 35180
    },
    {
      "entropy": 5.680328893661499,
      "epoch": 3.491961095672886,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038172027813000676,
      "loss": 5.0001,
      "mean_token_accuracy": 0.18494639545679092,
      "num_tokens": 77079808.0,
      "step": 35185
    },
    {
      "entropy": 5.644695806503296,
      "epoch": 3.4924573243350534,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00038168908445439725,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18689099699258804,
      "num_tokens": 77090772.0,
      "step": 35190
    },
    {
      "entropy": 5.650036811828613,
      "epoch": 3.4929535529972213,
      "grad_norm": 1.3125,
      "learning_rate": 0.00038165788813312844,
      "loss": 4.9939,
      "mean_token_accuracy": 0.1855382189154625,
      "num_tokens": 77101358.0,
      "step": 35195
    },
    {
      "entropy": 5.64893536567688,
      "epoch": 3.493449781659389,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00038162668916697385,
      "loss": 4.9277,
      "mean_token_accuracy": 0.19504402577877045,
      "num_tokens": 77111198.0,
      "step": 35200
    },
    {
      "entropy": 5.714456939697266,
      "epoch": 3.4939460103215563,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00038159548755670707,
      "loss": 5.0752,
      "mean_token_accuracy": 0.18895391523838043,
      "num_tokens": 77121990.0,
      "step": 35205
    },
    {
      "entropy": 5.72253384590149,
      "epoch": 3.4944422389837237,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038156428330310193,
      "loss": 5.0386,
      "mean_token_accuracy": 0.1875072419643402,
      "num_tokens": 77132952.0,
      "step": 35210
    },
    {
      "entropy": 5.657019758224488,
      "epoch": 3.4949384676458912,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003815330764069322,
      "loss": 4.9979,
      "mean_token_accuracy": 0.18853669911623,
      "num_tokens": 77142870.0,
      "step": 35215
    },
    {
      "entropy": 5.702692174911499,
      "epoch": 3.4954346963080587,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003815018668689717,
      "loss": 5.0791,
      "mean_token_accuracy": 0.1864641025662422,
      "num_tokens": 77154180.0,
      "step": 35220
    },
    {
      "entropy": 5.7228950500488285,
      "epoch": 3.495930924970226,
      "grad_norm": 1.21875,
      "learning_rate": 0.00038147065468999435,
      "loss": 5.0661,
      "mean_token_accuracy": 0.1870198130607605,
      "num_tokens": 77165083.0,
      "step": 35225
    },
    {
      "entropy": 5.734848260879517,
      "epoch": 3.4964271536323936,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038143943987077426,
      "loss": 5.0184,
      "mean_token_accuracy": 0.18701690137386323,
      "num_tokens": 77176809.0,
      "step": 35230
    },
    {
      "entropy": 5.640147876739502,
      "epoch": 3.496923382294561,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003814082224120854,
      "loss": 4.9699,
      "mean_token_accuracy": 0.18375977128744125,
      "num_tokens": 77187325.0,
      "step": 35235
    },
    {
      "entropy": 5.753223037719726,
      "epoch": 3.497419610956729,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003813770023147019,
      "loss": 5.0437,
      "mean_token_accuracy": 0.18522380143404008,
      "num_tokens": 77197403.0,
      "step": 35240
    },
    {
      "entropy": 5.657914686203003,
      "epoch": 3.4979158396188965,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038134577957939795,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18947443068027497,
      "num_tokens": 77208262.0,
      "step": 35245
    },
    {
      "entropy": 5.6841973781585695,
      "epoch": 3.498412068281064,
      "grad_norm": 1.125,
      "learning_rate": 0.00038131455420694786,
      "loss": 4.9927,
      "mean_token_accuracy": 0.19786932468414306,
      "num_tokens": 77219327.0,
      "step": 35250
    },
    {
      "entropy": 5.7087812423706055,
      "epoch": 3.4989082969432315,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003812833261981259,
      "loss": 5.1009,
      "mean_token_accuracy": 0.1876046061515808,
      "num_tokens": 77230380.0,
      "step": 35255
    },
    {
      "entropy": 5.686140537261963,
      "epoch": 3.499404525605399,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038125209555370644,
      "loss": 4.9449,
      "mean_token_accuracy": 0.19445386081933974,
      "num_tokens": 77241173.0,
      "step": 35260
    },
    {
      "entropy": 5.642883205413819,
      "epoch": 3.4999007542675664,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000381220862274464,
      "loss": 4.9922,
      "mean_token_accuracy": 0.18269055783748628,
      "num_tokens": 77252674.0,
      "step": 35265
    },
    {
      "entropy": 5.654024600982666,
      "epoch": 3.500396982929734,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038118962636117304,
      "loss": 5.0231,
      "mean_token_accuracy": 0.18527212888002395,
      "num_tokens": 77264826.0,
      "step": 35270
    },
    {
      "entropy": 5.708418512344361,
      "epoch": 3.500893211591902,
      "grad_norm": 1.125,
      "learning_rate": 0.00038115838781460816,
      "loss": 5.0047,
      "mean_token_accuracy": 0.18939741849899291,
      "num_tokens": 77276068.0,
      "step": 35275
    },
    {
      "entropy": 5.717672872543335,
      "epoch": 3.5013894402540693,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000381127146635544,
      "loss": 5.042,
      "mean_token_accuracy": 0.18449034988880159,
      "num_tokens": 77285394.0,
      "step": 35280
    },
    {
      "entropy": 5.663644742965698,
      "epoch": 3.5018856689162368,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003810959028247554,
      "loss": 5.0069,
      "mean_token_accuracy": 0.1832285761833191,
      "num_tokens": 77295843.0,
      "step": 35285
    },
    {
      "entropy": 5.690907287597656,
      "epoch": 3.5023818975784042,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003810646563830169,
      "loss": 5.0592,
      "mean_token_accuracy": 0.18256790935993195,
      "num_tokens": 77306064.0,
      "step": 35290
    },
    {
      "entropy": 5.675547456741333,
      "epoch": 3.5028781262405717,
      "grad_norm": 1.28125,
      "learning_rate": 0.00038103340731110354,
      "loss": 4.9454,
      "mean_token_accuracy": 0.18828621059656142,
      "num_tokens": 77316425.0,
      "step": 35295
    },
    {
      "entropy": 5.796253299713134,
      "epoch": 3.503374354902739,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038100215560979016,
      "loss": 5.1559,
      "mean_token_accuracy": 0.1768753483891487,
      "num_tokens": 77327185.0,
      "step": 35300
    },
    {
      "entropy": 5.722209358215332,
      "epoch": 3.5038705835649067,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038097090127985165,
      "loss": 5.0541,
      "mean_token_accuracy": 0.18210097402334213,
      "num_tokens": 77338791.0,
      "step": 35305
    },
    {
      "entropy": 5.696476459503174,
      "epoch": 3.504366812227074,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038093964432206316,
      "loss": 4.9869,
      "mean_token_accuracy": 0.18689615577459334,
      "num_tokens": 77350035.0,
      "step": 35310
    },
    {
      "entropy": 5.685007047653198,
      "epoch": 3.5048630408892416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003809083847371999,
      "loss": 5.0122,
      "mean_token_accuracy": 0.18907151669263839,
      "num_tokens": 77362422.0,
      "step": 35315
    },
    {
      "entropy": 5.644372367858887,
      "epoch": 3.505359269551409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003808771225260368,
      "loss": 4.9175,
      "mean_token_accuracy": 0.18935412615537645,
      "num_tokens": 77373383.0,
      "step": 35320
    },
    {
      "entropy": 5.643435478210449,
      "epoch": 3.5058554982135766,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038084585768934917,
      "loss": 5.0255,
      "mean_token_accuracy": 0.18910157829523086,
      "num_tokens": 77383782.0,
      "step": 35325
    },
    {
      "entropy": 5.627861738204956,
      "epoch": 3.5063517268757445,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038081459022791243,
      "loss": 5.0005,
      "mean_token_accuracy": 0.18550433665513993,
      "num_tokens": 77394266.0,
      "step": 35330
    },
    {
      "entropy": 5.665834236145019,
      "epoch": 3.506847955537912,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038078332014250185,
      "loss": 4.9468,
      "mean_token_accuracy": 0.19056940376758574,
      "num_tokens": 77405042.0,
      "step": 35335
    },
    {
      "entropy": 5.688625860214233,
      "epoch": 3.5073441842000794,
      "grad_norm": 1.125,
      "learning_rate": 0.0003807520474338928,
      "loss": 4.9874,
      "mean_token_accuracy": 0.19326189160346985,
      "num_tokens": 77415411.0,
      "step": 35340
    },
    {
      "entropy": 5.640138578414917,
      "epoch": 3.507840412862247,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003807207721028609,
      "loss": 4.9404,
      "mean_token_accuracy": 0.19440511167049407,
      "num_tokens": 77423991.0,
      "step": 35345
    },
    {
      "entropy": 5.640832424163818,
      "epoch": 3.5083366415244144,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003806894941501817,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18528542816638946,
      "num_tokens": 77435701.0,
      "step": 35350
    },
    {
      "entropy": 5.632437515258789,
      "epoch": 3.508832870186582,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003806582135766307,
      "loss": 4.9109,
      "mean_token_accuracy": 0.19355177134275436,
      "num_tokens": 77445618.0,
      "step": 35355
    },
    {
      "entropy": 5.656196975708008,
      "epoch": 3.5093290988487498,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038062693038298367,
      "loss": 4.929,
      "mean_token_accuracy": 0.19054544419050218,
      "num_tokens": 77456674.0,
      "step": 35360
    },
    {
      "entropy": 5.593820571899414,
      "epoch": 3.5098253275109172,
      "grad_norm": 1.125,
      "learning_rate": 0.0003805956445700165,
      "loss": 4.9886,
      "mean_token_accuracy": 0.18825012743473052,
      "num_tokens": 77467841.0,
      "step": 35365
    },
    {
      "entropy": 5.661830615997315,
      "epoch": 3.5103215561730847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003805643561385048,
      "loss": 4.8733,
      "mean_token_accuracy": 0.19167218953371049,
      "num_tokens": 77477883.0,
      "step": 35370
    },
    {
      "entropy": 5.59886999130249,
      "epoch": 3.510817784835252,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038053306508922455,
      "loss": 4.9201,
      "mean_token_accuracy": 0.19138913005590438,
      "num_tokens": 77488714.0,
      "step": 35375
    },
    {
      "entropy": 5.600363636016846,
      "epoch": 3.5113140134974197,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00038050177142295166,
      "loss": 4.9694,
      "mean_token_accuracy": 0.1932400107383728,
      "num_tokens": 77499325.0,
      "step": 35380
    },
    {
      "entropy": 5.641973638534546,
      "epoch": 3.511810242159587,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038047047514046223,
      "loss": 4.984,
      "mean_token_accuracy": 0.18250783830881118,
      "num_tokens": 77510682.0,
      "step": 35385
    },
    {
      "entropy": 5.675427532196045,
      "epoch": 3.5123064708217546,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003804391762425322,
      "loss": 4.9916,
      "mean_token_accuracy": 0.185961776971817,
      "num_tokens": 77520604.0,
      "step": 35390
    },
    {
      "entropy": 5.647856426239014,
      "epoch": 3.512802699483922,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038040787472993793,
      "loss": 5.0285,
      "mean_token_accuracy": 0.189648899435997,
      "num_tokens": 77532481.0,
      "step": 35395
    },
    {
      "entropy": 5.726211166381836,
      "epoch": 3.5132989281460896,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003803765706034554,
      "loss": 5.0795,
      "mean_token_accuracy": 0.17476728558540344,
      "num_tokens": 77543940.0,
      "step": 35400
    },
    {
      "entropy": 5.709249019622803,
      "epoch": 3.513795156808257,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038034526386386097,
      "loss": 4.8987,
      "mean_token_accuracy": 0.19719187766313553,
      "num_tokens": 77553765.0,
      "step": 35405
    },
    {
      "entropy": 5.640388250350952,
      "epoch": 3.5142913854704245,
      "grad_norm": 1.203125,
      "learning_rate": 0.00038031395451193105,
      "loss": 5.0095,
      "mean_token_accuracy": 0.18604450970888137,
      "num_tokens": 77566132.0,
      "step": 35410
    },
    {
      "entropy": 5.639721250534057,
      "epoch": 3.5147876141325924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038028264254844204,
      "loss": 4.9644,
      "mean_token_accuracy": 0.19255032539367675,
      "num_tokens": 77577706.0,
      "step": 35415
    },
    {
      "entropy": 5.670883750915527,
      "epoch": 3.51528384279476,
      "grad_norm": 1.234375,
      "learning_rate": 0.00038025132797417024,
      "loss": 4.9643,
      "mean_token_accuracy": 0.19531885981559755,
      "num_tokens": 77589113.0,
      "step": 35420
    },
    {
      "entropy": 5.622539901733399,
      "epoch": 3.5157800714569274,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003802200107898924,
      "loss": 4.9277,
      "mean_token_accuracy": 0.1895652011036873,
      "num_tokens": 77599012.0,
      "step": 35425
    },
    {
      "entropy": 5.646908617019653,
      "epoch": 3.516276300119095,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038018869099638506,
      "loss": 5.0297,
      "mean_token_accuracy": 0.18398013561964036,
      "num_tokens": 77610389.0,
      "step": 35430
    },
    {
      "entropy": 5.665364742279053,
      "epoch": 3.5167725287812623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003801573685944247,
      "loss": 4.9264,
      "mean_token_accuracy": 0.19129791408777236,
      "num_tokens": 77621245.0,
      "step": 35435
    },
    {
      "entropy": 5.679275178909302,
      "epoch": 3.51726875744343,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00038012604358478823,
      "loss": 4.9449,
      "mean_token_accuracy": 0.19369776099920272,
      "num_tokens": 77631003.0,
      "step": 35440
    },
    {
      "entropy": 5.607549095153809,
      "epoch": 3.5177649861055977,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00038009471596825244,
      "loss": 4.8971,
      "mean_token_accuracy": 0.193467815220356,
      "num_tokens": 77641222.0,
      "step": 35445
    },
    {
      "entropy": 5.713943147659302,
      "epoch": 3.518261214767765,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003800633857455942,
      "loss": 5.0583,
      "mean_token_accuracy": 0.18684111982584,
      "num_tokens": 77652705.0,
      "step": 35450
    },
    {
      "entropy": 5.692865991592408,
      "epoch": 3.5187574434299327,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003800320529175903,
      "loss": 5.0537,
      "mean_token_accuracy": 0.18326083123683928,
      "num_tokens": 77662883.0,
      "step": 35455
    },
    {
      "entropy": 5.650560188293457,
      "epoch": 3.5192536720921,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038000071748501787,
      "loss": 4.983,
      "mean_token_accuracy": 0.18985844999551774,
      "num_tokens": 77673979.0,
      "step": 35460
    },
    {
      "entropy": 5.690317296981812,
      "epoch": 3.5197499007542676,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037996937944865383,
      "loss": 4.9587,
      "mean_token_accuracy": 0.19126099199056626,
      "num_tokens": 77684387.0,
      "step": 35465
    },
    {
      "entropy": 5.7223350524902346,
      "epoch": 3.520246129416435,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003799380388092754,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18633685111999512,
      "num_tokens": 77695218.0,
      "step": 35470
    },
    {
      "entropy": 5.675091552734375,
      "epoch": 3.5207423580786026,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003799066955676598,
      "loss": 5.0103,
      "mean_token_accuracy": 0.1866128921508789,
      "num_tokens": 77705118.0,
      "step": 35475
    },
    {
      "entropy": 5.560525226593017,
      "epoch": 3.52123858674077,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003798753497245841,
      "loss": 4.9188,
      "mean_token_accuracy": 0.19423647671937944,
      "num_tokens": 77715440.0,
      "step": 35480
    },
    {
      "entropy": 5.6833850860595705,
      "epoch": 3.5217348154029375,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037984400128082576,
      "loss": 5.1051,
      "mean_token_accuracy": 0.18079975992441177,
      "num_tokens": 77725146.0,
      "step": 35485
    },
    {
      "entropy": 5.788061904907226,
      "epoch": 3.522231044065105,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037981265023716217,
      "loss": 5.0261,
      "mean_token_accuracy": 0.18843149095773698,
      "num_tokens": 77735987.0,
      "step": 35490
    },
    {
      "entropy": 5.703178739547729,
      "epoch": 3.5227272727272725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037978129659437063,
      "loss": 4.9521,
      "mean_token_accuracy": 0.1963578164577484,
      "num_tokens": 77747003.0,
      "step": 35495
    },
    {
      "entropy": 5.599821424484253,
      "epoch": 3.5232235013894404,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003797499403532287,
      "loss": 4.9527,
      "mean_token_accuracy": 0.19055369943380357,
      "num_tokens": 77757866.0,
      "step": 35500
    },
    {
      "entropy": 5.633730554580689,
      "epoch": 3.523719730051608,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037971858151451403,
      "loss": 4.9796,
      "mean_token_accuracy": 0.19212547391653062,
      "num_tokens": 77769448.0,
      "step": 35505
    },
    {
      "entropy": 5.6843994140625,
      "epoch": 3.5242159587137754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037968722007900415,
      "loss": 4.9397,
      "mean_token_accuracy": 0.19315038174390792,
      "num_tokens": 77781207.0,
      "step": 35510
    },
    {
      "entropy": 5.731891298294068,
      "epoch": 3.524712187375943,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037965585604747684,
      "loss": 5.0202,
      "mean_token_accuracy": 0.18125320971012115,
      "num_tokens": 77791050.0,
      "step": 35515
    },
    {
      "entropy": 5.674694442749024,
      "epoch": 3.5252084160381103,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003796244894207097,
      "loss": 4.9611,
      "mean_token_accuracy": 0.1939166083931923,
      "num_tokens": 77802486.0,
      "step": 35520
    },
    {
      "entropy": 5.637510585784912,
      "epoch": 3.525704644700278,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037959312019948075,
      "loss": 4.9698,
      "mean_token_accuracy": 0.1879166692495346,
      "num_tokens": 77815581.0,
      "step": 35525
    },
    {
      "entropy": 5.661537599563599,
      "epoch": 3.5262008733624453,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003795617483845678,
      "loss": 4.9766,
      "mean_token_accuracy": 0.19822381287813187,
      "num_tokens": 77826139.0,
      "step": 35530
    },
    {
      "entropy": 5.731377410888672,
      "epoch": 3.526697102024613,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037953037397674876,
      "loss": 5.0716,
      "mean_token_accuracy": 0.1834945410490036,
      "num_tokens": 77837929.0,
      "step": 35535
    },
    {
      "entropy": 5.695299434661865,
      "epoch": 3.5271933306867806,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003794989969768017,
      "loss": 4.9462,
      "mean_token_accuracy": 0.1947939470410347,
      "num_tokens": 77848118.0,
      "step": 35540
    },
    {
      "entropy": 5.679597091674805,
      "epoch": 3.527689559348948,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003794676173855048,
      "loss": 4.9817,
      "mean_token_accuracy": 0.18719785511493683,
      "num_tokens": 77859378.0,
      "step": 35545
    },
    {
      "entropy": 5.625251197814942,
      "epoch": 3.5281857880111156,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037943623520363595,
      "loss": 4.9661,
      "mean_token_accuracy": 0.1881416767835617,
      "num_tokens": 77872524.0,
      "step": 35550
    },
    {
      "entropy": 5.619844961166382,
      "epoch": 3.528682016673283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003794048504319736,
      "loss": 4.9768,
      "mean_token_accuracy": 0.18918735533952713,
      "num_tokens": 77883053.0,
      "step": 35555
    },
    {
      "entropy": 5.696208572387695,
      "epoch": 3.5291782453354505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037937346307129584,
      "loss": 5.0752,
      "mean_token_accuracy": 0.18888791799545288,
      "num_tokens": 77894709.0,
      "step": 35560
    },
    {
      "entropy": 5.732572317123413,
      "epoch": 3.529674473997618,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037934207312238116,
      "loss": 5.0361,
      "mean_token_accuracy": 0.18347702026367188,
      "num_tokens": 77906213.0,
      "step": 35565
    },
    {
      "entropy": 5.701269960403442,
      "epoch": 3.5301707026597855,
      "grad_norm": 1.125,
      "learning_rate": 0.00037931068058600785,
      "loss": 4.9869,
      "mean_token_accuracy": 0.18258585184812545,
      "num_tokens": 77917015.0,
      "step": 35570
    },
    {
      "entropy": 5.737707710266113,
      "epoch": 3.530666931321953,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037927928546295454,
      "loss": 5.0706,
      "mean_token_accuracy": 0.18819415122270583,
      "num_tokens": 77927504.0,
      "step": 35575
    },
    {
      "entropy": 5.744711112976074,
      "epoch": 3.5311631599841204,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003792478877539995,
      "loss": 5.0006,
      "mean_token_accuracy": 0.19327492117881775,
      "num_tokens": 77937656.0,
      "step": 35580
    },
    {
      "entropy": 5.786378717422485,
      "epoch": 3.5316593886462884,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003792164874599216,
      "loss": 5.1151,
      "mean_token_accuracy": 0.1744204118847847,
      "num_tokens": 77948626.0,
      "step": 35585
    },
    {
      "entropy": 5.71426739692688,
      "epoch": 3.532155617308456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003791850845814994,
      "loss": 5.0127,
      "mean_token_accuracy": 0.186875618994236,
      "num_tokens": 77960055.0,
      "step": 35590
    },
    {
      "entropy": 5.646316766738892,
      "epoch": 3.5326518459706233,
      "grad_norm": 1.125,
      "learning_rate": 0.00037915367911951146,
      "loss": 4.9942,
      "mean_token_accuracy": 0.18229075968265535,
      "num_tokens": 77971379.0,
      "step": 35595
    },
    {
      "entropy": 5.640619134902954,
      "epoch": 3.533148074632791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037912227107473684,
      "loss": 4.9575,
      "mean_token_accuracy": 0.1927997887134552,
      "num_tokens": 77982746.0,
      "step": 35600
    },
    {
      "entropy": 5.646386623382568,
      "epoch": 3.5336443032949583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037909086044795427,
      "loss": 5.0094,
      "mean_token_accuracy": 0.1855776995420456,
      "num_tokens": 77994561.0,
      "step": 35605
    },
    {
      "entropy": 5.7119792938232425,
      "epoch": 3.5341405319571257,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037905944723994257,
      "loss": 5.0753,
      "mean_token_accuracy": 0.19208500534296036,
      "num_tokens": 78004835.0,
      "step": 35610
    },
    {
      "entropy": 5.616388416290283,
      "epoch": 3.534636760619293,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037902803145148085,
      "loss": 4.9013,
      "mean_token_accuracy": 0.19474005252122878,
      "num_tokens": 78015669.0,
      "step": 35615
    },
    {
      "entropy": 5.6968296527862545,
      "epoch": 3.535132989281461,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003789966130833481,
      "loss": 5.0418,
      "mean_token_accuracy": 0.17706861943006516,
      "num_tokens": 78025340.0,
      "step": 35620
    },
    {
      "entropy": 5.691270399093628,
      "epoch": 3.5356292179436286,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003789651921363234,
      "loss": 5.0063,
      "mean_token_accuracy": 0.19316482543945312,
      "num_tokens": 78034790.0,
      "step": 35625
    },
    {
      "entropy": 5.699869298934937,
      "epoch": 3.536125446605796,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037893376861118606,
      "loss": 5.0888,
      "mean_token_accuracy": 0.1843210443854332,
      "num_tokens": 78045803.0,
      "step": 35630
    },
    {
      "entropy": 5.685393190383911,
      "epoch": 3.5366216752679636,
      "grad_norm": 1.125,
      "learning_rate": 0.00037890234250871515,
      "loss": 4.9669,
      "mean_token_accuracy": 0.19170989394187926,
      "num_tokens": 78056690.0,
      "step": 35635
    },
    {
      "entropy": 5.705612659454346,
      "epoch": 3.537117903930131,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037887091382969,
      "loss": 4.9711,
      "mean_token_accuracy": 0.19968766272068023,
      "num_tokens": 78067888.0,
      "step": 35640
    },
    {
      "entropy": 5.684579086303711,
      "epoch": 3.5376141325922985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003788394825748901,
      "loss": 4.9818,
      "mean_token_accuracy": 0.19644466340541838,
      "num_tokens": 78077877.0,
      "step": 35645
    },
    {
      "entropy": 5.6204887390136715,
      "epoch": 3.538110361254466,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037880804874509467,
      "loss": 4.9818,
      "mean_token_accuracy": 0.18991949111223222,
      "num_tokens": 78090114.0,
      "step": 35650
    },
    {
      "entropy": 5.776602792739868,
      "epoch": 3.5386065899166335,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003787766123410834,
      "loss": 5.1416,
      "mean_token_accuracy": 0.1817834734916687,
      "num_tokens": 78099737.0,
      "step": 35655
    },
    {
      "entropy": 5.669385385513306,
      "epoch": 3.539102818578801,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003787451733636357,
      "loss": 4.9209,
      "mean_token_accuracy": 0.19326305538415908,
      "num_tokens": 78110979.0,
      "step": 35660
    },
    {
      "entropy": 5.702703666687012,
      "epoch": 3.5395990472409684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037871373181353135,
      "loss": 4.9667,
      "mean_token_accuracy": 0.19102310091257096,
      "num_tokens": 78123107.0,
      "step": 35665
    },
    {
      "entropy": 5.691550064086914,
      "epoch": 3.5400952759031363,
      "grad_norm": 1.25,
      "learning_rate": 0.00037868228769154987,
      "loss": 5.0265,
      "mean_token_accuracy": 0.1880363330245018,
      "num_tokens": 78134128.0,
      "step": 35670
    },
    {
      "entropy": 5.71596508026123,
      "epoch": 3.540591504565304,
      "grad_norm": 1.25,
      "learning_rate": 0.0003786508409984711,
      "loss": 5.0451,
      "mean_token_accuracy": 0.17885391861200334,
      "num_tokens": 78144561.0,
      "step": 35675
    },
    {
      "entropy": 5.64503698348999,
      "epoch": 3.5410877332274713,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037861939173507484,
      "loss": 4.9114,
      "mean_token_accuracy": 0.19113813191652299,
      "num_tokens": 78155700.0,
      "step": 35680
    },
    {
      "entropy": 5.731560850143433,
      "epoch": 3.5415839618896388,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037858793990214087,
      "loss": 5.0863,
      "mean_token_accuracy": 0.17945425659418107,
      "num_tokens": 78166904.0,
      "step": 35685
    },
    {
      "entropy": 5.618485736846924,
      "epoch": 3.5420801905518062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037855648550044925,
      "loss": 4.9578,
      "mean_token_accuracy": 0.1929793879389763,
      "num_tokens": 78178409.0,
      "step": 35690
    },
    {
      "entropy": 5.760079574584961,
      "epoch": 3.5425764192139737,
      "grad_norm": 1.125,
      "learning_rate": 0.0003785250285307799,
      "loss": 5.0264,
      "mean_token_accuracy": 0.18198666125535964,
      "num_tokens": 78190639.0,
      "step": 35695
    },
    {
      "entropy": 5.7165422439575195,
      "epoch": 3.543072647876141,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037849356899391295,
      "loss": 5.0041,
      "mean_token_accuracy": 0.19032979756593704,
      "num_tokens": 78202345.0,
      "step": 35700
    },
    {
      "entropy": 5.655549001693726,
      "epoch": 3.543568876538309,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003784621068906285,
      "loss": 4.9193,
      "mean_token_accuracy": 0.18742664456367492,
      "num_tokens": 78212767.0,
      "step": 35705
    },
    {
      "entropy": 5.626258277893067,
      "epoch": 3.5440651052004766,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037843064222170677,
      "loss": 4.936,
      "mean_token_accuracy": 0.18709788620471954,
      "num_tokens": 78223338.0,
      "step": 35710
    },
    {
      "entropy": 5.613433313369751,
      "epoch": 3.544561333862644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037839917498792787,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1912049949169159,
      "num_tokens": 78234903.0,
      "step": 35715
    },
    {
      "entropy": 5.655250072479248,
      "epoch": 3.5450575625248115,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037836770519007236,
      "loss": 4.9994,
      "mean_token_accuracy": 0.18900089710950851,
      "num_tokens": 78245936.0,
      "step": 35720
    },
    {
      "entropy": 5.6610438346862795,
      "epoch": 3.545553791186979,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003783362328289205,
      "loss": 5.0014,
      "mean_token_accuracy": 0.18994010537862777,
      "num_tokens": 78256294.0,
      "step": 35725
    },
    {
      "entropy": 5.7119818210601805,
      "epoch": 3.5460500198491465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003783047579052526,
      "loss": 5.0661,
      "mean_token_accuracy": 0.18666032552719117,
      "num_tokens": 78266385.0,
      "step": 35730
    },
    {
      "entropy": 5.686742973327637,
      "epoch": 3.546546248511314,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037827328041984944,
      "loss": 4.9659,
      "mean_token_accuracy": 0.19874135553836822,
      "num_tokens": 78276044.0,
      "step": 35735
    },
    {
      "entropy": 5.672509717941284,
      "epoch": 3.5470424771734814,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003782418003734914,
      "loss": 5.0129,
      "mean_token_accuracy": 0.18693076819181442,
      "num_tokens": 78286943.0,
      "step": 35740
    },
    {
      "entropy": 5.6289825439453125,
      "epoch": 3.547538705835649,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003782103177669592,
      "loss": 4.9492,
      "mean_token_accuracy": 0.1940180018544197,
      "num_tokens": 78298313.0,
      "step": 35745
    },
    {
      "entropy": 5.664329099655151,
      "epoch": 3.5480349344978164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037817883260103354,
      "loss": 5.0395,
      "mean_token_accuracy": 0.17851733565330505,
      "num_tokens": 78309750.0,
      "step": 35750
    },
    {
      "entropy": 5.655488109588623,
      "epoch": 3.548531163159984,
      "grad_norm": 1.125,
      "learning_rate": 0.00037814734487649516,
      "loss": 4.9291,
      "mean_token_accuracy": 0.19507017880678176,
      "num_tokens": 78321998.0,
      "step": 35755
    },
    {
      "entropy": 5.761549615859986,
      "epoch": 3.5490273918221518,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003781158545941248,
      "loss": 5.0969,
      "mean_token_accuracy": 0.18280594944953918,
      "num_tokens": 78333973.0,
      "step": 35760
    },
    {
      "entropy": 5.755862903594971,
      "epoch": 3.5495236204843192,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037808436175470344,
      "loss": 5.1113,
      "mean_token_accuracy": 0.17861372828483582,
      "num_tokens": 78345281.0,
      "step": 35765
    },
    {
      "entropy": 5.731892871856689,
      "epoch": 3.5500198491464867,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003780528663590122,
      "loss": 5.0702,
      "mean_token_accuracy": 0.18003044426441192,
      "num_tokens": 78356995.0,
      "step": 35770
    },
    {
      "entropy": 5.69048376083374,
      "epoch": 3.550516077808654,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003780213684078318,
      "loss": 5.003,
      "mean_token_accuracy": 0.1857777163386345,
      "num_tokens": 78367910.0,
      "step": 35775
    },
    {
      "entropy": 5.652951860427857,
      "epoch": 3.5510123064708217,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037798986790194347,
      "loss": 4.9345,
      "mean_token_accuracy": 0.19258201718330384,
      "num_tokens": 78378995.0,
      "step": 35780
    },
    {
      "entropy": 5.735172176361084,
      "epoch": 3.551508535132989,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037795836484212837,
      "loss": 5.006,
      "mean_token_accuracy": 0.19083615094423295,
      "num_tokens": 78389881.0,
      "step": 35785
    },
    {
      "entropy": 5.614833354949951,
      "epoch": 3.552004763795157,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00037792685922916763,
      "loss": 4.8952,
      "mean_token_accuracy": 0.19067423939704894,
      "num_tokens": 78400242.0,
      "step": 35790
    },
    {
      "entropy": 5.623944711685181,
      "epoch": 3.5525009924573245,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037789535106384266,
      "loss": 4.9166,
      "mean_token_accuracy": 0.19015021175146102,
      "num_tokens": 78410057.0,
      "step": 35795
    },
    {
      "entropy": 5.644933748245239,
      "epoch": 3.552997221119492,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003778638403469346,
      "loss": 5.0161,
      "mean_token_accuracy": 0.18992754817008972,
      "num_tokens": 78419684.0,
      "step": 35800
    },
    {
      "entropy": 5.660487508773803,
      "epoch": 3.5534934497816595,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037783232707922507,
      "loss": 4.9395,
      "mean_token_accuracy": 0.19001346677541733,
      "num_tokens": 78430569.0,
      "step": 35805
    },
    {
      "entropy": 5.735655117034912,
      "epoch": 3.553989678443827,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037780081126149523,
      "loss": 5.0018,
      "mean_token_accuracy": 0.19197714179754258,
      "num_tokens": 78441290.0,
      "step": 35810
    },
    {
      "entropy": 5.685708522796631,
      "epoch": 3.5544859071059944,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037776929289452696,
      "loss": 4.9809,
      "mean_token_accuracy": 0.19895345121622085,
      "num_tokens": 78451638.0,
      "step": 35815
    },
    {
      "entropy": 5.669237804412842,
      "epoch": 3.554982135768162,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003777377719791016,
      "loss": 4.9963,
      "mean_token_accuracy": 0.19092458337545395,
      "num_tokens": 78462759.0,
      "step": 35820
    },
    {
      "entropy": 5.639545583724976,
      "epoch": 3.5554783644303294,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037770624851600086,
      "loss": 4.9746,
      "mean_token_accuracy": 0.1924720212817192,
      "num_tokens": 78475465.0,
      "step": 35825
    },
    {
      "entropy": 5.587998533248902,
      "epoch": 3.555974593092497,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003776747225060065,
      "loss": 4.8545,
      "mean_token_accuracy": 0.19091791659593582,
      "num_tokens": 78487900.0,
      "step": 35830
    },
    {
      "entropy": 5.619225883483887,
      "epoch": 3.5564708217546643,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003776431939499003,
      "loss": 4.932,
      "mean_token_accuracy": 0.19006531238555907,
      "num_tokens": 78497538.0,
      "step": 35835
    },
    {
      "entropy": 5.652000999450683,
      "epoch": 3.556967050416832,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000377611662848464,
      "loss": 5.018,
      "mean_token_accuracy": 0.18540110290050507,
      "num_tokens": 78508299.0,
      "step": 35840
    },
    {
      "entropy": 5.684651279449463,
      "epoch": 3.5574632790789997,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003775801292024796,
      "loss": 4.9119,
      "mean_token_accuracy": 0.19404597133398055,
      "num_tokens": 78518188.0,
      "step": 35845
    },
    {
      "entropy": 5.639324951171875,
      "epoch": 3.557959507741167,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037754859301272905,
      "loss": 4.9769,
      "mean_token_accuracy": 0.19475458711385726,
      "num_tokens": 78528390.0,
      "step": 35850
    },
    {
      "entropy": 5.584597063064575,
      "epoch": 3.5584557364033347,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00037751705427999436,
      "loss": 5.0523,
      "mean_token_accuracy": 0.18620749264955522,
      "num_tokens": 78540266.0,
      "step": 35855
    },
    {
      "entropy": 5.670819616317749,
      "epoch": 3.558951965065502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003774855130050576,
      "loss": 5.0288,
      "mean_token_accuracy": 0.18674387782812119,
      "num_tokens": 78550937.0,
      "step": 35860
    },
    {
      "entropy": 5.711783170700073,
      "epoch": 3.5594481937276696,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037745396918870095,
      "loss": 4.95,
      "mean_token_accuracy": 0.19311609566211702,
      "num_tokens": 78561619.0,
      "step": 35865
    },
    {
      "entropy": 5.656787538528443,
      "epoch": 3.559944422389837,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003774224228317067,
      "loss": 4.9512,
      "mean_token_accuracy": 0.1941488727927208,
      "num_tokens": 78573840.0,
      "step": 35870
    },
    {
      "entropy": 5.637988567352295,
      "epoch": 3.560440651052005,
      "grad_norm": 1.09375,
      "learning_rate": 0.000377390873934857,
      "loss": 4.9963,
      "mean_token_accuracy": 0.1847207263112068,
      "num_tokens": 78585477.0,
      "step": 35875
    },
    {
      "entropy": 5.720789909362793,
      "epoch": 3.5609368797141725,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003773593224989343,
      "loss": 5.0735,
      "mean_token_accuracy": 0.18477414250373841,
      "num_tokens": 78597281.0,
      "step": 35880
    },
    {
      "entropy": 5.680354166030884,
      "epoch": 3.56143310837634,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003773277685247209,
      "loss": 4.9765,
      "mean_token_accuracy": 0.18808562755584718,
      "num_tokens": 78609300.0,
      "step": 35885
    },
    {
      "entropy": 5.7109581470489506,
      "epoch": 3.5619293370385074,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037729621201299933,
      "loss": 5.0013,
      "mean_token_accuracy": 0.1923844873905182,
      "num_tokens": 78620443.0,
      "step": 35890
    },
    {
      "entropy": 5.703012132644654,
      "epoch": 3.562425565700675,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003772646529645522,
      "loss": 4.9988,
      "mean_token_accuracy": 0.18791887164115906,
      "num_tokens": 78630113.0,
      "step": 35895
    },
    {
      "entropy": 5.698500919342041,
      "epoch": 3.5629217943628424,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003772330913801621,
      "loss": 4.9344,
      "mean_token_accuracy": 0.1898600235581398,
      "num_tokens": 78640092.0,
      "step": 35900
    },
    {
      "entropy": 5.661121416091919,
      "epoch": 3.56341802302501,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003772015272606115,
      "loss": 4.9486,
      "mean_token_accuracy": 0.19012387543916703,
      "num_tokens": 78651092.0,
      "step": 35905
    },
    {
      "entropy": 5.652026033401489,
      "epoch": 3.5639142516871773,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003771699606066833,
      "loss": 4.9944,
      "mean_token_accuracy": 0.191082164645195,
      "num_tokens": 78662435.0,
      "step": 35910
    },
    {
      "entropy": 5.631215763092041,
      "epoch": 3.564410480349345,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003771383914191602,
      "loss": 4.9741,
      "mean_token_accuracy": 0.1900736168026924,
      "num_tokens": 78673299.0,
      "step": 35915
    },
    {
      "entropy": 5.690202093124389,
      "epoch": 3.5649067090115123,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003771068196988251,
      "loss": 4.9748,
      "mean_token_accuracy": 0.19355975538492204,
      "num_tokens": 78683286.0,
      "step": 35920
    },
    {
      "entropy": 5.671480178833008,
      "epoch": 3.5654029376736798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037707524544646094,
      "loss": 4.929,
      "mean_token_accuracy": 0.19175591319799423,
      "num_tokens": 78695282.0,
      "step": 35925
    },
    {
      "entropy": 5.645294141769409,
      "epoch": 3.5658991663358477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003770436686628506,
      "loss": 4.9548,
      "mean_token_accuracy": 0.1903252959251404,
      "num_tokens": 78706880.0,
      "step": 35930
    },
    {
      "entropy": 5.714405536651611,
      "epoch": 3.566395394998015,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037701208934877716,
      "loss": 5.0111,
      "mean_token_accuracy": 0.18341223150491714,
      "num_tokens": 78717515.0,
      "step": 35935
    },
    {
      "entropy": 5.653937911987304,
      "epoch": 3.5668916236601826,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037698050750502375,
      "loss": 4.9609,
      "mean_token_accuracy": 0.18856991529464723,
      "num_tokens": 78728959.0,
      "step": 35940
    },
    {
      "entropy": 5.595774364471436,
      "epoch": 3.56738785232235,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003769489231323735,
      "loss": 4.9378,
      "mean_token_accuracy": 0.19072185456752777,
      "num_tokens": 78740580.0,
      "step": 35945
    },
    {
      "entropy": 5.640825605392456,
      "epoch": 3.5678840809845176,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003769173362316096,
      "loss": 4.9524,
      "mean_token_accuracy": 0.19413156509399415,
      "num_tokens": 78751388.0,
      "step": 35950
    },
    {
      "entropy": 5.642160844802857,
      "epoch": 3.568380309646685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037688574680351544,
      "loss": 5.0065,
      "mean_token_accuracy": 0.18479550182819365,
      "num_tokens": 78763474.0,
      "step": 35955
    },
    {
      "entropy": 5.635367584228516,
      "epoch": 3.568876538308853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037685415484887426,
      "loss": 4.9866,
      "mean_token_accuracy": 0.1860400527715683,
      "num_tokens": 78775570.0,
      "step": 35960
    },
    {
      "entropy": 5.692687034606934,
      "epoch": 3.5693727669710205,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037682256036846947,
      "loss": 4.9739,
      "mean_token_accuracy": 0.18299733549356462,
      "num_tokens": 78786558.0,
      "step": 35965
    },
    {
      "entropy": 5.691313886642456,
      "epoch": 3.569868995633188,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037679096336308473,
      "loss": 4.9719,
      "mean_token_accuracy": 0.19433978348970413,
      "num_tokens": 78796872.0,
      "step": 35970
    },
    {
      "entropy": 5.695421028137207,
      "epoch": 3.5703652242953554,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037675936383350326,
      "loss": 4.9713,
      "mean_token_accuracy": 0.18338507711887359,
      "num_tokens": 78807971.0,
      "step": 35975
    },
    {
      "entropy": 5.716197299957275,
      "epoch": 3.570861452957523,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003767277617805089,
      "loss": 5.0292,
      "mean_token_accuracy": 0.1841510459780693,
      "num_tokens": 78818593.0,
      "step": 35980
    },
    {
      "entropy": 5.674802827835083,
      "epoch": 3.5713576816196904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003766961572048853,
      "loss": 4.9813,
      "mean_token_accuracy": 0.1924201339483261,
      "num_tokens": 78830573.0,
      "step": 35985
    },
    {
      "entropy": 5.7079243659973145,
      "epoch": 3.571853910281858,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003766645501074161,
      "loss": 5.0504,
      "mean_token_accuracy": 0.18252751380205154,
      "num_tokens": 78842041.0,
      "step": 35990
    },
    {
      "entropy": 5.690223741531372,
      "epoch": 3.5723501389440253,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037663294048888513,
      "loss": 5.0075,
      "mean_token_accuracy": 0.1873215392231941,
      "num_tokens": 78852549.0,
      "step": 35995
    },
    {
      "entropy": 5.706714868545532,
      "epoch": 3.572846367606193,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037660132835007625,
      "loss": 5.0793,
      "mean_token_accuracy": 0.18995071351528167,
      "num_tokens": 78862706.0,
      "step": 36000
    },
    {
      "epoch": 3.572846367606193,
      "eval_entropy": 5.42946990319685,
      "eval_loss": 5.1162943840026855,
      "eval_mean_token_accuracy": 0.18835964696877897,
      "eval_num_tokens": 78862706.0,
      "eval_runtime": 20.5926,
      "eval_samples_per_second": 1635.006,
      "eval_steps_per_second": 204.394,
      "step": 36000
    },
    {
      "entropy": 5.689231443405151,
      "epoch": 3.5733425962683603,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003765697136917734,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18451897352933883,
      "num_tokens": 78874214.0,
      "step": 36005
    },
    {
      "entropy": 5.700612735748291,
      "epoch": 3.5738388249305277,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003765380965147604,
      "loss": 5.0687,
      "mean_token_accuracy": 0.1870029464364052,
      "num_tokens": 78886178.0,
      "step": 36010
    },
    {
      "entropy": 5.677816772460938,
      "epoch": 3.5743350535926957,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037650647681982145,
      "loss": 4.9739,
      "mean_token_accuracy": 0.1897470623254776,
      "num_tokens": 78896114.0,
      "step": 36015
    },
    {
      "entropy": 5.732187414169312,
      "epoch": 3.574831282254863,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037647485460774065,
      "loss": 4.9873,
      "mean_token_accuracy": 0.1899136483669281,
      "num_tokens": 78906447.0,
      "step": 36020
    },
    {
      "entropy": 5.597247123718262,
      "epoch": 3.5753275109170306,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003764432298793021,
      "loss": 4.8845,
      "mean_token_accuracy": 0.19207495599985122,
      "num_tokens": 78917737.0,
      "step": 36025
    },
    {
      "entropy": 5.612577676773071,
      "epoch": 3.575823739579198,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037641160263529003,
      "loss": 4.9989,
      "mean_token_accuracy": 0.19048761427402497,
      "num_tokens": 78929682.0,
      "step": 36030
    },
    {
      "entropy": 5.664119005203247,
      "epoch": 3.5763199682413656,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037637997287648877,
      "loss": 4.9592,
      "mean_token_accuracy": 0.1917135238647461,
      "num_tokens": 78941875.0,
      "step": 36035
    },
    {
      "entropy": 5.681607866287232,
      "epoch": 3.576816196903533,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037634834060368256,
      "loss": 4.9868,
      "mean_token_accuracy": 0.18890285193920137,
      "num_tokens": 78951891.0,
      "step": 36040
    },
    {
      "entropy": 5.726670408248902,
      "epoch": 3.577312425565701,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003763167058176559,
      "loss": 5.066,
      "mean_token_accuracy": 0.1869365394115448,
      "num_tokens": 78964141.0,
      "step": 36045
    },
    {
      "entropy": 5.694410800933838,
      "epoch": 3.5778086542278684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003762850685191933,
      "loss": 4.9894,
      "mean_token_accuracy": 0.19469261914491653,
      "num_tokens": 78973644.0,
      "step": 36050
    },
    {
      "entropy": 5.617409658432007,
      "epoch": 3.578304882890036,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003762534287090792,
      "loss": 4.9813,
      "mean_token_accuracy": 0.1850503623485565,
      "num_tokens": 78984666.0,
      "step": 36055
    },
    {
      "entropy": 5.628836441040039,
      "epoch": 3.5788011115522034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003762217863880983,
      "loss": 4.9473,
      "mean_token_accuracy": 0.1897743061184883,
      "num_tokens": 78996064.0,
      "step": 36060
    },
    {
      "entropy": 5.70395131111145,
      "epoch": 3.579297340214371,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003761901415570351,
      "loss": 5.0451,
      "mean_token_accuracy": 0.18385178297758104,
      "num_tokens": 79006733.0,
      "step": 36065
    },
    {
      "entropy": 5.699537324905395,
      "epoch": 3.5797935688765383,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037615849421667466,
      "loss": 5.001,
      "mean_token_accuracy": 0.19161095917224885,
      "num_tokens": 79018027.0,
      "step": 36070
    },
    {
      "entropy": 5.70254955291748,
      "epoch": 3.580289797538706,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003761268443678014,
      "loss": 5.0028,
      "mean_token_accuracy": 0.18604833334684373,
      "num_tokens": 79028281.0,
      "step": 36075
    },
    {
      "entropy": 5.70380973815918,
      "epoch": 3.5807860262008733,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003760951920112003,
      "loss": 4.9858,
      "mean_token_accuracy": 0.18746557533740998,
      "num_tokens": 79039110.0,
      "step": 36080
    },
    {
      "entropy": 5.7426252365112305,
      "epoch": 3.5812822548630407,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037606353714765626,
      "loss": 5.0608,
      "mean_token_accuracy": 0.19006237834692002,
      "num_tokens": 79051294.0,
      "step": 36085
    },
    {
      "entropy": 5.697900724411011,
      "epoch": 3.5817784835252082,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003760318797779543,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1773891866207123,
      "num_tokens": 79062790.0,
      "step": 36090
    },
    {
      "entropy": 5.679536533355713,
      "epoch": 3.5822747121873757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003760002199028794,
      "loss": 5.0421,
      "mean_token_accuracy": 0.1876528561115265,
      "num_tokens": 79073699.0,
      "step": 36095
    },
    {
      "entropy": 5.695623683929443,
      "epoch": 3.5827709408495436,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003759685575232168,
      "loss": 4.9162,
      "mean_token_accuracy": 0.1936618283390999,
      "num_tokens": 79084588.0,
      "step": 36100
    },
    {
      "entropy": 5.662362766265869,
      "epoch": 3.583267169511711,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003759368926397514,
      "loss": 5.0048,
      "mean_token_accuracy": 0.19094982594251633,
      "num_tokens": 79094767.0,
      "step": 36105
    },
    {
      "entropy": 5.66282091140747,
      "epoch": 3.5837633981738786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003759052252532686,
      "loss": 5.0166,
      "mean_token_accuracy": 0.18598652780056,
      "num_tokens": 79105186.0,
      "step": 36110
    },
    {
      "entropy": 5.677522277832031,
      "epoch": 3.584259626836046,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003758735553645537,
      "loss": 4.9922,
      "mean_token_accuracy": 0.18822258412837983,
      "num_tokens": 79116184.0,
      "step": 36115
    },
    {
      "entropy": 5.64170994758606,
      "epoch": 3.5847558554982135,
      "grad_norm": 1.25,
      "learning_rate": 0.000375841882974392,
      "loss": 4.939,
      "mean_token_accuracy": 0.18939086496829988,
      "num_tokens": 79126676.0,
      "step": 36120
    },
    {
      "entropy": 5.725362014770508,
      "epoch": 3.585252084160381,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003758102080835688,
      "loss": 5.1011,
      "mean_token_accuracy": 0.1822570949792862,
      "num_tokens": 79138090.0,
      "step": 36125
    },
    {
      "entropy": 5.699523115158081,
      "epoch": 3.5857483128225485,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003757785306928697,
      "loss": 4.9645,
      "mean_token_accuracy": 0.19610797613859177,
      "num_tokens": 79150087.0,
      "step": 36130
    },
    {
      "entropy": 5.680775547027588,
      "epoch": 3.5862445414847164,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003757468508030803,
      "loss": 5.032,
      "mean_token_accuracy": 0.19545863419771195,
      "num_tokens": 79162296.0,
      "step": 36135
    },
    {
      "entropy": 5.658356380462647,
      "epoch": 3.586740770146884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003757151684149859,
      "loss": 5.0299,
      "mean_token_accuracy": 0.18592896759510041,
      "num_tokens": 79173763.0,
      "step": 36140
    },
    {
      "entropy": 5.723645973205566,
      "epoch": 3.5872369988090513,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003756834835293725,
      "loss": 5.071,
      "mean_token_accuracy": 0.19095649421215058,
      "num_tokens": 79184323.0,
      "step": 36145
    },
    {
      "entropy": 5.700101232528686,
      "epoch": 3.587733227471219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003756517961470256,
      "loss": 5.0447,
      "mean_token_accuracy": 0.183742156624794,
      "num_tokens": 79195811.0,
      "step": 36150
    },
    {
      "entropy": 5.754171991348267,
      "epoch": 3.5882294561333863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003756201062687311,
      "loss": 5.0859,
      "mean_token_accuracy": 0.1879430204629898,
      "num_tokens": 79207243.0,
      "step": 36155
    },
    {
      "entropy": 5.7058258056640625,
      "epoch": 3.5887256847955538,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037558841389527466,
      "loss": 5.035,
      "mean_token_accuracy": 0.19018321931362153,
      "num_tokens": 79218909.0,
      "step": 36160
    },
    {
      "entropy": 5.690326738357544,
      "epoch": 3.5892219134577212,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037555671902744244,
      "loss": 4.9818,
      "mean_token_accuracy": 0.19155339300632476,
      "num_tokens": 79229170.0,
      "step": 36165
    },
    {
      "entropy": 5.67734661102295,
      "epoch": 3.5897181421198887,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037552502166602014,
      "loss": 5.0381,
      "mean_token_accuracy": 0.18901422917842864,
      "num_tokens": 79240001.0,
      "step": 36170
    },
    {
      "entropy": 5.693157529830932,
      "epoch": 3.590214370782056,
      "grad_norm": 1.109375,
      "learning_rate": 0.00037549332181179403,
      "loss": 5.0522,
      "mean_token_accuracy": 0.1861259937286377,
      "num_tokens": 79251525.0,
      "step": 36175
    },
    {
      "entropy": 5.6799290657043455,
      "epoch": 3.5907105994442237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037546161946555004,
      "loss": 5.0352,
      "mean_token_accuracy": 0.18421755135059356,
      "num_tokens": 79264125.0,
      "step": 36180
    },
    {
      "entropy": 5.673059749603271,
      "epoch": 3.5912068281063916,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037542991462807433,
      "loss": 4.9733,
      "mean_token_accuracy": 0.19114619940519334,
      "num_tokens": 79273880.0,
      "step": 36185
    },
    {
      "entropy": 5.683645296096802,
      "epoch": 3.591703056768559,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037539820730015316,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18789510279893876,
      "num_tokens": 79284542.0,
      "step": 36190
    },
    {
      "entropy": 5.6126069068908695,
      "epoch": 3.5921992854307265,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037536649748257285,
      "loss": 4.8823,
      "mean_token_accuracy": 0.19679659157991408,
      "num_tokens": 79295182.0,
      "step": 36195
    },
    {
      "entropy": 5.597224998474121,
      "epoch": 3.592695514092894,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037533478517611955,
      "loss": 4.9624,
      "mean_token_accuracy": 0.1951667696237564,
      "num_tokens": 79305613.0,
      "step": 36200
    },
    {
      "entropy": 5.648825788497925,
      "epoch": 3.5931917427550615,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003753030703815799,
      "loss": 4.9946,
      "mean_token_accuracy": 0.19632148444652558,
      "num_tokens": 79316566.0,
      "step": 36205
    },
    {
      "entropy": 5.651117992401123,
      "epoch": 3.593687971417229,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003752713530997402,
      "loss": 4.9694,
      "mean_token_accuracy": 0.19697742611169816,
      "num_tokens": 79327224.0,
      "step": 36210
    },
    {
      "entropy": 5.7126007080078125,
      "epoch": 3.5941842000793964,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037523963333138693,
      "loss": 5.0752,
      "mean_token_accuracy": 0.18034445345401764,
      "num_tokens": 79338069.0,
      "step": 36215
    },
    {
      "entropy": 5.779147481918335,
      "epoch": 3.5946804287415643,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003752079110773068,
      "loss": 5.0951,
      "mean_token_accuracy": 0.1850396528840065,
      "num_tokens": 79348677.0,
      "step": 36220
    },
    {
      "entropy": 5.621577978134155,
      "epoch": 3.595176657403732,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003751761863382865,
      "loss": 4.9718,
      "mean_token_accuracy": 0.18884248435497283,
      "num_tokens": 79359253.0,
      "step": 36225
    },
    {
      "entropy": 5.657811498641967,
      "epoch": 3.5956728860658993,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037514445911511257,
      "loss": 4.9493,
      "mean_token_accuracy": 0.1910020515322685,
      "num_tokens": 79370409.0,
      "step": 36230
    },
    {
      "entropy": 5.673265647888184,
      "epoch": 3.5961691147280668,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003751127294085719,
      "loss": 5.0156,
      "mean_token_accuracy": 0.18243845999240876,
      "num_tokens": 79381259.0,
      "step": 36235
    },
    {
      "entropy": 5.630610227584839,
      "epoch": 3.5966653433902342,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037508099721945113,
      "loss": 4.9504,
      "mean_token_accuracy": 0.1888488858938217,
      "num_tokens": 79391723.0,
      "step": 36240
    },
    {
      "entropy": 5.694388484954834,
      "epoch": 3.5971615720524017,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003750492625485374,
      "loss": 4.9799,
      "mean_token_accuracy": 0.1868538662791252,
      "num_tokens": 79401015.0,
      "step": 36245
    },
    {
      "entropy": 5.6195837497711185,
      "epoch": 3.597657800714569,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003750175253966175,
      "loss": 4.8883,
      "mean_token_accuracy": 0.20117703229188919,
      "num_tokens": 79411923.0,
      "step": 36250
    },
    {
      "entropy": 5.629544878005982,
      "epoch": 3.5981540293767367,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003749857857644785,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18272101283073425,
      "num_tokens": 79423934.0,
      "step": 36255
    },
    {
      "entropy": 5.610666418075562,
      "epoch": 3.598650258038904,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003749540436529074,
      "loss": 4.936,
      "mean_token_accuracy": 0.1960848182439804,
      "num_tokens": 79433718.0,
      "step": 36260
    },
    {
      "entropy": 5.679204845428467,
      "epoch": 3.5991464867010716,
      "grad_norm": 1.3125,
      "learning_rate": 0.00037492229906269147,
      "loss": 4.9602,
      "mean_token_accuracy": 0.18186224102973939,
      "num_tokens": 79443935.0,
      "step": 36265
    },
    {
      "entropy": 5.682668399810791,
      "epoch": 3.5996427153632395,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037489055199461783,
      "loss": 4.954,
      "mean_token_accuracy": 0.19887565970420837,
      "num_tokens": 79455442.0,
      "step": 36270
    },
    {
      "entropy": 5.641422176361084,
      "epoch": 3.600138944025407,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003748588024494737,
      "loss": 4.9506,
      "mean_token_accuracy": 0.18795233815908433,
      "num_tokens": 79466591.0,
      "step": 36275
    },
    {
      "entropy": 5.691841793060303,
      "epoch": 3.6006351726875745,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037482705042804655,
      "loss": 5.0842,
      "mean_token_accuracy": 0.18038203120231627,
      "num_tokens": 79477064.0,
      "step": 36280
    },
    {
      "entropy": 5.747005319595337,
      "epoch": 3.601131401349742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037479529593112355,
      "loss": 5.0756,
      "mean_token_accuracy": 0.18711491376161576,
      "num_tokens": 79488799.0,
      "step": 36285
    },
    {
      "entropy": 5.695085096359253,
      "epoch": 3.6016276300119094,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003747635389594923,
      "loss": 5.0104,
      "mean_token_accuracy": 0.19378212392330169,
      "num_tokens": 79499406.0,
      "step": 36290
    },
    {
      "entropy": 5.6215822219848635,
      "epoch": 3.602123858674077,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037473177951394027,
      "loss": 4.909,
      "mean_token_accuracy": 0.19826410412788392,
      "num_tokens": 79509347.0,
      "step": 36295
    },
    {
      "entropy": 5.717696237564087,
      "epoch": 3.6026200873362444,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037470001759525495,
      "loss": 5.0402,
      "mean_token_accuracy": 0.18726304918527603,
      "num_tokens": 79520515.0,
      "step": 36300
    },
    {
      "entropy": 5.591885232925415,
      "epoch": 3.6031163159984123,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037466825320422407,
      "loss": 4.9011,
      "mean_token_accuracy": 0.19650056213140488,
      "num_tokens": 79530563.0,
      "step": 36305
    },
    {
      "entropy": 5.6710833549499515,
      "epoch": 3.60361254466058,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037463648634163534,
      "loss": 4.9926,
      "mean_token_accuracy": 0.1844656988978386,
      "num_tokens": 79541578.0,
      "step": 36310
    },
    {
      "entropy": 5.627803993225098,
      "epoch": 3.6041087733227473,
      "grad_norm": 1.125,
      "learning_rate": 0.0003746047170082764,
      "loss": 4.9056,
      "mean_token_accuracy": 0.1997298315167427,
      "num_tokens": 79552170.0,
      "step": 36315
    },
    {
      "entropy": 5.691776371002197,
      "epoch": 3.6046050019849147,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037457294520493504,
      "loss": 5.0205,
      "mean_token_accuracy": 0.18946897983551025,
      "num_tokens": 79562205.0,
      "step": 36320
    },
    {
      "entropy": 5.696606016159057,
      "epoch": 3.605101230647082,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037454117093239925,
      "loss": 5.0294,
      "mean_token_accuracy": 0.18052190840244292,
      "num_tokens": 79572683.0,
      "step": 36325
    },
    {
      "entropy": 5.771498584747315,
      "epoch": 3.6055974593092497,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003745093941914569,
      "loss": 5.0716,
      "mean_token_accuracy": 0.17510385066270828,
      "num_tokens": 79583453.0,
      "step": 36330
    },
    {
      "entropy": 5.687484979629517,
      "epoch": 3.606093687971417,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037447761498289595,
      "loss": 4.939,
      "mean_token_accuracy": 0.19225471019744872,
      "num_tokens": 79593902.0,
      "step": 36335
    },
    {
      "entropy": 5.628468179702759,
      "epoch": 3.6065899166335846,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003744458333075046,
      "loss": 4.9525,
      "mean_token_accuracy": 0.18814547806978227,
      "num_tokens": 79606511.0,
      "step": 36340
    },
    {
      "entropy": 5.652497100830078,
      "epoch": 3.607086145295752,
      "grad_norm": 1.25,
      "learning_rate": 0.0003744140491660708,
      "loss": 4.9886,
      "mean_token_accuracy": 0.1870461568236351,
      "num_tokens": 79617082.0,
      "step": 36345
    },
    {
      "entropy": 5.730311632156372,
      "epoch": 3.6075823739579196,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003743822625593829,
      "loss": 5.0684,
      "mean_token_accuracy": 0.18119352161884308,
      "num_tokens": 79627721.0,
      "step": 36350
    },
    {
      "entropy": 5.716165494918823,
      "epoch": 3.608078602620087,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000374350473488229,
      "loss": 4.9873,
      "mean_token_accuracy": 0.18449195176362992,
      "num_tokens": 79637464.0,
      "step": 36355
    },
    {
      "entropy": 5.6306211948394775,
      "epoch": 3.608574831282255,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037431868195339735,
      "loss": 4.9983,
      "mean_token_accuracy": 0.18466707468032836,
      "num_tokens": 79647128.0,
      "step": 36360
    },
    {
      "entropy": 5.618447208404541,
      "epoch": 3.6090710599444225,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003742868879556764,
      "loss": 4.9389,
      "mean_token_accuracy": 0.19923633188009263,
      "num_tokens": 79658569.0,
      "step": 36365
    },
    {
      "entropy": 5.6781651973724365,
      "epoch": 3.60956728860659,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00037425509149585466,
      "loss": 4.9905,
      "mean_token_accuracy": 0.1892881378531456,
      "num_tokens": 79669660.0,
      "step": 36370
    },
    {
      "entropy": 5.723041820526123,
      "epoch": 3.6100635172687574,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003742232925747205,
      "loss": 5.013,
      "mean_token_accuracy": 0.18280348032712937,
      "num_tokens": 79680017.0,
      "step": 36375
    },
    {
      "entropy": 5.641502714157104,
      "epoch": 3.610559745930925,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003741914911930625,
      "loss": 4.933,
      "mean_token_accuracy": 0.2051854819059372,
      "num_tokens": 79691291.0,
      "step": 36380
    },
    {
      "entropy": 5.6952728748321535,
      "epoch": 3.6110559745930924,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003741596873516693,
      "loss": 5.0076,
      "mean_token_accuracy": 0.184572970867157,
      "num_tokens": 79701539.0,
      "step": 36385
    },
    {
      "entropy": 5.669416189193726,
      "epoch": 3.6115522032552603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003741278810513294,
      "loss": 4.9912,
      "mean_token_accuracy": 0.1954258605837822,
      "num_tokens": 79711869.0,
      "step": 36390
    },
    {
      "entropy": 5.657706642150879,
      "epoch": 3.6120484319174277,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037409607229283175,
      "loss": 5.0494,
      "mean_token_accuracy": 0.1869868740439415,
      "num_tokens": 79722171.0,
      "step": 36395
    },
    {
      "entropy": 5.691597270965576,
      "epoch": 3.6125446605795952,
      "grad_norm": 1.234375,
      "learning_rate": 0.000374064261076965,
      "loss": 4.9978,
      "mean_token_accuracy": 0.18399623334407805,
      "num_tokens": 79733877.0,
      "step": 36400
    },
    {
      "entropy": 5.675327968597412,
      "epoch": 3.6130408892417627,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000374032447404518,
      "loss": 4.9789,
      "mean_token_accuracy": 0.1882468655705452,
      "num_tokens": 79744494.0,
      "step": 36405
    },
    {
      "entropy": 5.740707683563232,
      "epoch": 3.61353711790393,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003740006312762799,
      "loss": 5.0508,
      "mean_token_accuracy": 0.1871474042534828,
      "num_tokens": 79755989.0,
      "step": 36410
    },
    {
      "entropy": 5.7116687297821045,
      "epoch": 3.6140333465660976,
      "grad_norm": 1.796875,
      "learning_rate": 0.00037396881269303924,
      "loss": 5.0614,
      "mean_token_accuracy": 0.19300121814012527,
      "num_tokens": 79767184.0,
      "step": 36415
    },
    {
      "entropy": 5.71405439376831,
      "epoch": 3.614529575228265,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037393699165558534,
      "loss": 5.0319,
      "mean_token_accuracy": 0.18833743780851364,
      "num_tokens": 79778493.0,
      "step": 36420
    },
    {
      "entropy": 5.711003589630127,
      "epoch": 3.6150258038904326,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003739051681647073,
      "loss": 5.0276,
      "mean_token_accuracy": 0.1887346476316452,
      "num_tokens": 79787998.0,
      "step": 36425
    },
    {
      "entropy": 5.705341005325318,
      "epoch": 3.6155220325526,
      "grad_norm": 1.125,
      "learning_rate": 0.00037387334222119424,
      "loss": 5.0597,
      "mean_token_accuracy": 0.18867634683847428,
      "num_tokens": 79799469.0,
      "step": 36430
    },
    {
      "entropy": 5.714768743515014,
      "epoch": 3.6160182612147675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037384151382583526,
      "loss": 4.965,
      "mean_token_accuracy": 0.18614092767238616,
      "num_tokens": 79808656.0,
      "step": 36435
    },
    {
      "entropy": 5.709560918807983,
      "epoch": 3.616514489876935,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003738096829794197,
      "loss": 4.9795,
      "mean_token_accuracy": 0.1928809478878975,
      "num_tokens": 79818127.0,
      "step": 36440
    },
    {
      "entropy": 5.700338315963745,
      "epoch": 3.617010718539103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000373777849682737,
      "loss": 5.0045,
      "mean_token_accuracy": 0.19037398546934128,
      "num_tokens": 79830451.0,
      "step": 36445
    },
    {
      "entropy": 5.694755411148071,
      "epoch": 3.6175069472012704,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003737460139365764,
      "loss": 5.0375,
      "mean_token_accuracy": 0.18601388484239578,
      "num_tokens": 79842623.0,
      "step": 36450
    },
    {
      "entropy": 5.701075696945191,
      "epoch": 3.618003175863438,
      "grad_norm": 1.34375,
      "learning_rate": 0.0003737141757417275,
      "loss": 4.9813,
      "mean_token_accuracy": 0.1867730662226677,
      "num_tokens": 79852208.0,
      "step": 36455
    },
    {
      "entropy": 5.719927644729614,
      "epoch": 3.6184994045256054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003736823350989797,
      "loss": 5.0297,
      "mean_token_accuracy": 0.18529783934354782,
      "num_tokens": 79862356.0,
      "step": 36460
    },
    {
      "entropy": 5.64711012840271,
      "epoch": 3.618995633187773,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037365049200912274,
      "loss": 5.0024,
      "mean_token_accuracy": 0.18645581901073455,
      "num_tokens": 79873045.0,
      "step": 36465
    },
    {
      "entropy": 5.735877561569214,
      "epoch": 3.6194918618499403,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00037361864647294596,
      "loss": 5.0308,
      "mean_token_accuracy": 0.1876615047454834,
      "num_tokens": 79882738.0,
      "step": 36470
    },
    {
      "entropy": 5.714392852783203,
      "epoch": 3.6199880905121082,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003735867984912394,
      "loss": 5.0334,
      "mean_token_accuracy": 0.18477843552827836,
      "num_tokens": 79894047.0,
      "step": 36475
    },
    {
      "entropy": 5.6580431938171385,
      "epoch": 3.6204843191742757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037355494806479264,
      "loss": 4.9344,
      "mean_token_accuracy": 0.195564104616642,
      "num_tokens": 79905607.0,
      "step": 36480
    },
    {
      "entropy": 5.674556684494019,
      "epoch": 3.620980547836443,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003735230951943955,
      "loss": 5.0638,
      "mean_token_accuracy": 0.18317975252866744,
      "num_tokens": 79917995.0,
      "step": 36485
    },
    {
      "entropy": 5.711395502090454,
      "epoch": 3.6214767764986107,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037349123988083795,
      "loss": 5.0074,
      "mean_token_accuracy": 0.18842198699712753,
      "num_tokens": 79929600.0,
      "step": 36490
    },
    {
      "entropy": 5.714243793487549,
      "epoch": 3.621973005160778,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037345938212490987,
      "loss": 5.0092,
      "mean_token_accuracy": 0.19222851395606994,
      "num_tokens": 79940597.0,
      "step": 36495
    },
    {
      "entropy": 5.770765161514282,
      "epoch": 3.6224692338229456,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037342752192740127,
      "loss": 5.0219,
      "mean_token_accuracy": 0.19198094606399535,
      "num_tokens": 79951531.0,
      "step": 36500
    },
    {
      "entropy": 5.693882942199707,
      "epoch": 3.622965462485113,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003733956592891022,
      "loss": 5.0339,
      "mean_token_accuracy": 0.18198160827159882,
      "num_tokens": 79962651.0,
      "step": 36505
    },
    {
      "entropy": 5.689920949935913,
      "epoch": 3.6234616911472806,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00037336379421080274,
      "loss": 5.0177,
      "mean_token_accuracy": 0.18871867656707764,
      "num_tokens": 79973219.0,
      "step": 36510
    },
    {
      "entropy": 5.668004941940308,
      "epoch": 3.623957919809448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003733319266932932,
      "loss": 4.9702,
      "mean_token_accuracy": 0.18972921520471572,
      "num_tokens": 79985037.0,
      "step": 36515
    },
    {
      "entropy": 5.702591276168823,
      "epoch": 3.6244541484716155,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00037330005673736385,
      "loss": 5.0588,
      "mean_token_accuracy": 0.1775040701031685,
      "num_tokens": 79994596.0,
      "step": 36520
    },
    {
      "entropy": 5.694736337661743,
      "epoch": 3.624950377133783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003732681843438048,
      "loss": 4.9986,
      "mean_token_accuracy": 0.18588554710149766,
      "num_tokens": 80006045.0,
      "step": 36525
    },
    {
      "entropy": 5.6835493564605715,
      "epoch": 3.625446605795951,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037323630951340654,
      "loss": 4.9882,
      "mean_token_accuracy": 0.1841891348361969,
      "num_tokens": 80016423.0,
      "step": 36530
    },
    {
      "entropy": 5.684215974807739,
      "epoch": 3.6259428344581184,
      "grad_norm": 1.28125,
      "learning_rate": 0.00037320443224695955,
      "loss": 5.0478,
      "mean_token_accuracy": 0.1858665317296982,
      "num_tokens": 80026579.0,
      "step": 36535
    },
    {
      "entropy": 5.6325257301330565,
      "epoch": 3.626439063120286,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003731725525452542,
      "loss": 4.9518,
      "mean_token_accuracy": 0.19589333832263947,
      "num_tokens": 80037227.0,
      "step": 36540
    },
    {
      "entropy": 5.709292459487915,
      "epoch": 3.6269352917824533,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003731406704090811,
      "loss": 5.0197,
      "mean_token_accuracy": 0.1826389268040657,
      "num_tokens": 80049421.0,
      "step": 36545
    },
    {
      "entropy": 5.737694835662841,
      "epoch": 3.627431520444621,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037310878583923085,
      "loss": 5.0165,
      "mean_token_accuracy": 0.18629119992256166,
      "num_tokens": 80059710.0,
      "step": 36550
    },
    {
      "entropy": 5.676320171356201,
      "epoch": 3.6279277491067883,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003730768988364941,
      "loss": 5.0002,
      "mean_token_accuracy": 0.19064594507217408,
      "num_tokens": 80071403.0,
      "step": 36555
    },
    {
      "entropy": 5.633984661102295,
      "epoch": 3.628423977768956,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037304500940166165,
      "loss": 4.8798,
      "mean_token_accuracy": 0.20385363548994065,
      "num_tokens": 80082978.0,
      "step": 36560
    },
    {
      "entropy": 5.694615507125855,
      "epoch": 3.6289202064311237,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00037301311753552407,
      "loss": 5.0401,
      "mean_token_accuracy": 0.19660512655973433,
      "num_tokens": 80093002.0,
      "step": 36565
    },
    {
      "entropy": 5.692898273468018,
      "epoch": 3.629416435093291,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003729812232388726,
      "loss": 5.0157,
      "mean_token_accuracy": 0.18681116849184037,
      "num_tokens": 80104738.0,
      "step": 36570
    },
    {
      "entropy": 5.662651395797729,
      "epoch": 3.6299126637554586,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003729493265124978,
      "loss": 4.9603,
      "mean_token_accuracy": 0.187140554189682,
      "num_tokens": 80113746.0,
      "step": 36575
    },
    {
      "entropy": 5.732413864135742,
      "epoch": 3.630408892417626,
      "grad_norm": 1.234375,
      "learning_rate": 0.00037291742735719077,
      "loss": 5.0535,
      "mean_token_accuracy": 0.18228576928377152,
      "num_tokens": 80126558.0,
      "step": 36580
    },
    {
      "entropy": 5.718997669219971,
      "epoch": 3.6309051210797936,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003728855257737425,
      "loss": 5.0658,
      "mean_token_accuracy": 0.18390565663576125,
      "num_tokens": 80138174.0,
      "step": 36585
    },
    {
      "entropy": 5.699109029769898,
      "epoch": 3.631401349741961,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003728536217629442,
      "loss": 4.9619,
      "mean_token_accuracy": 0.19313093572854995,
      "num_tokens": 80148044.0,
      "step": 36590
    },
    {
      "entropy": 5.757740926742554,
      "epoch": 3.6318975784041285,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037282171532558697,
      "loss": 5.0716,
      "mean_token_accuracy": 0.18041326403617858,
      "num_tokens": 80158725.0,
      "step": 36595
    },
    {
      "entropy": 5.790618515014648,
      "epoch": 3.632393807066296,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037278980646246186,
      "loss": 5.1218,
      "mean_token_accuracy": 0.1843035861849785,
      "num_tokens": 80170360.0,
      "step": 36600
    },
    {
      "entropy": 5.672171640396118,
      "epoch": 3.6328900357284635,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00037275789517436035,
      "loss": 4.9808,
      "mean_token_accuracy": 0.19156004935503007,
      "num_tokens": 80180167.0,
      "step": 36605
    },
    {
      "entropy": 5.593704414367676,
      "epoch": 3.633386264390631,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037272598146207363,
      "loss": 4.9258,
      "mean_token_accuracy": 0.18772467225790024,
      "num_tokens": 80190867.0,
      "step": 36610
    },
    {
      "entropy": 5.659312057495117,
      "epoch": 3.633882493052799,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00037269406532639324,
      "loss": 4.9604,
      "mean_token_accuracy": 0.19506973624229432,
      "num_tokens": 80201258.0,
      "step": 36615
    },
    {
      "entropy": 5.65741376876831,
      "epoch": 3.6343787217149663,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003726621467681105,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18810434490442277,
      "num_tokens": 80212499.0,
      "step": 36620
    },
    {
      "entropy": 5.679231739044189,
      "epoch": 3.634874950377134,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003726302257880169,
      "loss": 4.9734,
      "mean_token_accuracy": 0.194025719165802,
      "num_tokens": 80223611.0,
      "step": 36625
    },
    {
      "entropy": 5.643146181106568,
      "epoch": 3.6353711790393013,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003725983023869041,
      "loss": 4.9838,
      "mean_token_accuracy": 0.18607894480228424,
      "num_tokens": 80233678.0,
      "step": 36630
    },
    {
      "entropy": 5.719719982147216,
      "epoch": 3.6358674077014688,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003725663765655638,
      "loss": 5.0691,
      "mean_token_accuracy": 0.18255472630262376,
      "num_tokens": 80246307.0,
      "step": 36635
    },
    {
      "entropy": 5.6787597179412845,
      "epoch": 3.6363636363636362,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003725344483247875,
      "loss": 4.8684,
      "mean_token_accuracy": 0.19981939643621444,
      "num_tokens": 80256945.0,
      "step": 36640
    },
    {
      "entropy": 5.661499834060669,
      "epoch": 3.636859865025804,
      "grad_norm": 1.21875,
      "learning_rate": 0.00037250251766536715,
      "loss": 4.9906,
      "mean_token_accuracy": 0.1964288353919983,
      "num_tokens": 80266299.0,
      "step": 36645
    },
    {
      "entropy": 5.651803255081177,
      "epoch": 3.6373560936879716,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003724705845880944,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18971528559923173,
      "num_tokens": 80276004.0,
      "step": 36650
    },
    {
      "entropy": 5.696263456344605,
      "epoch": 3.637852322350139,
      "grad_norm": 1.40625,
      "learning_rate": 0.0003724386490937611,
      "loss": 5.0713,
      "mean_token_accuracy": 0.17707935720682144,
      "num_tokens": 80285962.0,
      "step": 36655
    },
    {
      "entropy": 5.746332883834839,
      "epoch": 3.6383485510123066,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003724067111831594,
      "loss": 5.0633,
      "mean_token_accuracy": 0.18857621848583223,
      "num_tokens": 80295457.0,
      "step": 36660
    },
    {
      "entropy": 5.6383991718292235,
      "epoch": 3.638844779674474,
      "grad_norm": 1.0,
      "learning_rate": 0.0003723747708570811,
      "loss": 4.9541,
      "mean_token_accuracy": 0.19641972333192825,
      "num_tokens": 80307125.0,
      "step": 36665
    },
    {
      "entropy": 5.658149671554566,
      "epoch": 3.6393410083366415,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00037234282811631826,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18511402904987334,
      "num_tokens": 80318629.0,
      "step": 36670
    },
    {
      "entropy": 5.731631135940551,
      "epoch": 3.639837236998809,
      "grad_norm": 1.25,
      "learning_rate": 0.0003723108829616631,
      "loss": 5.0742,
      "mean_token_accuracy": 0.18591103106737136,
      "num_tokens": 80330901.0,
      "step": 36675
    },
    {
      "entropy": 5.652381515502929,
      "epoch": 3.6403334656609765,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003722789353939077,
      "loss": 4.9425,
      "mean_token_accuracy": 0.19376811683177947,
      "num_tokens": 80341888.0,
      "step": 36680
    },
    {
      "entropy": 5.669185018539428,
      "epoch": 3.640829694323144,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037224698541384434,
      "loss": 4.9149,
      "mean_token_accuracy": 0.19740341901779174,
      "num_tokens": 80352903.0,
      "step": 36685
    },
    {
      "entropy": 5.6303267002105715,
      "epoch": 3.6413259229853114,
      "grad_norm": 1.125,
      "learning_rate": 0.00037221503302226533,
      "loss": 4.9371,
      "mean_token_accuracy": 0.19154756218194963,
      "num_tokens": 80364609.0,
      "step": 36690
    },
    {
      "entropy": 5.610120391845703,
      "epoch": 3.641822151647479,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003721830782199629,
      "loss": 5.0015,
      "mean_token_accuracy": 0.18371507823467254,
      "num_tokens": 80375431.0,
      "step": 36695
    },
    {
      "entropy": 5.6663583278656,
      "epoch": 3.642318380309647,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003721511210077296,
      "loss": 5.0032,
      "mean_token_accuracy": 0.18908468633890152,
      "num_tokens": 80385503.0,
      "step": 36700
    },
    {
      "entropy": 5.69906702041626,
      "epoch": 3.6428146089718143,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037211916138635775,
      "loss": 5.065,
      "mean_token_accuracy": 0.18563000708818436,
      "num_tokens": 80396752.0,
      "step": 36705
    },
    {
      "entropy": 5.679953622817993,
      "epoch": 3.643310837633982,
      "grad_norm": 1.25,
      "learning_rate": 0.00037208719935664003,
      "loss": 4.9736,
      "mean_token_accuracy": 0.1882636070251465,
      "num_tokens": 80407342.0,
      "step": 36710
    },
    {
      "entropy": 5.692767238616943,
      "epoch": 3.6438070662961493,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000372055234919369,
      "loss": 4.9549,
      "mean_token_accuracy": 0.18640069514513016,
      "num_tokens": 80418356.0,
      "step": 36715
    },
    {
      "entropy": 5.648395252227783,
      "epoch": 3.6443032949583167,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037202326807533724,
      "loss": 4.9603,
      "mean_token_accuracy": 0.19392492175102233,
      "num_tokens": 80428579.0,
      "step": 36720
    },
    {
      "entropy": 5.710886859893799,
      "epoch": 3.644799523620484,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003719912988253376,
      "loss": 5.0949,
      "mean_token_accuracy": 0.18192739188671112,
      "num_tokens": 80441494.0,
      "step": 36725
    },
    {
      "entropy": 5.693770265579223,
      "epoch": 3.6452957522826517,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037195932717016254,
      "loss": 5.0269,
      "mean_token_accuracy": 0.19205267429351808,
      "num_tokens": 80453766.0,
      "step": 36730
    },
    {
      "entropy": 5.731163120269775,
      "epoch": 3.6457919809448196,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003719273531106053,
      "loss": 4.93,
      "mean_token_accuracy": 0.1915822669863701,
      "num_tokens": 80464382.0,
      "step": 36735
    },
    {
      "entropy": 5.737446260452271,
      "epoch": 3.646288209606987,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037189537664745846,
      "loss": 5.0813,
      "mean_token_accuracy": 0.18499174267053603,
      "num_tokens": 80475605.0,
      "step": 36740
    },
    {
      "entropy": 5.6868585586547855,
      "epoch": 3.6467844382691545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003718633977815151,
      "loss": 4.9969,
      "mean_token_accuracy": 0.18907151371240616,
      "num_tokens": 80485612.0,
      "step": 36745
    },
    {
      "entropy": 5.68290376663208,
      "epoch": 3.647280666931322,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037183141651356814,
      "loss": 4.9827,
      "mean_token_accuracy": 0.19447409957647324,
      "num_tokens": 80497651.0,
      "step": 36750
    },
    {
      "entropy": 5.742915248870849,
      "epoch": 3.6477768955934895,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003717994328444108,
      "loss": 5.0771,
      "mean_token_accuracy": 0.18962849229574202,
      "num_tokens": 80509897.0,
      "step": 36755
    },
    {
      "entropy": 5.650699758529663,
      "epoch": 3.648273124255657,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037176744677483605,
      "loss": 5.0097,
      "mean_token_accuracy": 0.1910737320780754,
      "num_tokens": 80521883.0,
      "step": 36760
    },
    {
      "entropy": 5.640802049636841,
      "epoch": 3.6487693529178244,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037173545830563714,
      "loss": 4.9135,
      "mean_token_accuracy": 0.19302146583795549,
      "num_tokens": 80532580.0,
      "step": 36765
    },
    {
      "entropy": 5.729000568389893,
      "epoch": 3.649265581579992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003717034674376073,
      "loss": 5.0102,
      "mean_token_accuracy": 0.1823265254497528,
      "num_tokens": 80543163.0,
      "step": 36770
    },
    {
      "entropy": 5.679079914093018,
      "epoch": 3.6497618102421594,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00037167147417153983,
      "loss": 5.0496,
      "mean_token_accuracy": 0.1879713922739029,
      "num_tokens": 80553721.0,
      "step": 36775
    },
    {
      "entropy": 5.746005344390869,
      "epoch": 3.650258038904327,
      "grad_norm": 1.40625,
      "learning_rate": 0.00037163947850822816,
      "loss": 5.026,
      "mean_token_accuracy": 0.18376531451940536,
      "num_tokens": 80564340.0,
      "step": 36780
    },
    {
      "entropy": 5.7504003047943115,
      "epoch": 3.650754267566495,
      "grad_norm": 1.171875,
      "learning_rate": 0.00037160748044846566,
      "loss": 5.0939,
      "mean_token_accuracy": 0.18535516709089278,
      "num_tokens": 80575669.0,
      "step": 36785
    },
    {
      "entropy": 5.6861976146697994,
      "epoch": 3.6512504962286623,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003715754799930458,
      "loss": 4.9792,
      "mean_token_accuracy": 0.19012091010808946,
      "num_tokens": 80586946.0,
      "step": 36790
    },
    {
      "entropy": 5.633227777481079,
      "epoch": 3.6517467248908297,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003715434771427621,
      "loss": 4.967,
      "mean_token_accuracy": 0.19155206233263017,
      "num_tokens": 80596876.0,
      "step": 36795
    },
    {
      "entropy": 5.6450114250183105,
      "epoch": 3.652242953552997,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003715114718984083,
      "loss": 4.9875,
      "mean_token_accuracy": 0.18567190915346146,
      "num_tokens": 80608243.0,
      "step": 36800
    },
    {
      "entropy": 5.64986777305603,
      "epoch": 3.6527391822151647,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037147946426077785,
      "loss": 4.9903,
      "mean_token_accuracy": 0.1886451080441475,
      "num_tokens": 80618514.0,
      "step": 36805
    },
    {
      "entropy": 5.712722253799439,
      "epoch": 3.653235410877332,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003714474542306646,
      "loss": 5.0167,
      "mean_token_accuracy": 0.1813293218612671,
      "num_tokens": 80628686.0,
      "step": 36810
    },
    {
      "entropy": 5.698686122894287,
      "epoch": 3.6537316395394996,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037141544180886233,
      "loss": 4.9359,
      "mean_token_accuracy": 0.18671101480722427,
      "num_tokens": 80639414.0,
      "step": 36815
    },
    {
      "entropy": 5.571812629699707,
      "epoch": 3.6542278682016676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003713834269961649,
      "loss": 4.9081,
      "mean_token_accuracy": 0.19628764390945436,
      "num_tokens": 80652504.0,
      "step": 36820
    },
    {
      "entropy": 5.643714761734008,
      "epoch": 3.654724096863835,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00037135140979336614,
      "loss": 5.0057,
      "mean_token_accuracy": 0.18544722050428392,
      "num_tokens": 80663923.0,
      "step": 36825
    },
    {
      "entropy": 5.686873912811279,
      "epoch": 3.6552203255260025,
      "grad_norm": 1.390625,
      "learning_rate": 0.00037131939020126,
      "loss": 4.9944,
      "mean_token_accuracy": 0.18939546942710878,
      "num_tokens": 80674150.0,
      "step": 36830
    },
    {
      "entropy": 5.761466550827026,
      "epoch": 3.65571655418817,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003712873682206406,
      "loss": 5.1105,
      "mean_token_accuracy": 0.18332516700029372,
      "num_tokens": 80685963.0,
      "step": 36835
    },
    {
      "entropy": 5.675213432312011,
      "epoch": 3.6562127828503375,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003712553438523019,
      "loss": 4.939,
      "mean_token_accuracy": 0.19711731672286986,
      "num_tokens": 80696838.0,
      "step": 36840
    },
    {
      "entropy": 5.647917413711548,
      "epoch": 3.656709011512505,
      "grad_norm": 1.1875,
      "learning_rate": 0.00037122331709703803,
      "loss": 5.0571,
      "mean_token_accuracy": 0.1877598822116852,
      "num_tokens": 80708333.0,
      "step": 36845
    },
    {
      "entropy": 5.653958559036255,
      "epoch": 3.6572052401746724,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003711912879556433,
      "loss": 4.9769,
      "mean_token_accuracy": 0.19305110424757005,
      "num_tokens": 80718875.0,
      "step": 36850
    },
    {
      "entropy": 5.7747947692871096,
      "epoch": 3.65770146883684,
      "grad_norm": 1.34375,
      "learning_rate": 0.000371159256428912,
      "loss": 5.0272,
      "mean_token_accuracy": 0.18356576859951018,
      "num_tokens": 80728637.0,
      "step": 36855
    },
    {
      "entropy": 5.735069036483765,
      "epoch": 3.6581976974990074,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037112722251763817,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18337704241275787,
      "num_tokens": 80740083.0,
      "step": 36860
    },
    {
      "entropy": 5.696503925323486,
      "epoch": 3.658693926161175,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003710951862226165,
      "loss": 5.0296,
      "mean_token_accuracy": 0.18765013962984084,
      "num_tokens": 80752412.0,
      "step": 36865
    },
    {
      "entropy": 5.637906742095947,
      "epoch": 3.6591901548233428,
      "grad_norm": 1.140625,
      "learning_rate": 0.00037106314754464125,
      "loss": 4.9468,
      "mean_token_accuracy": 0.18853290975093842,
      "num_tokens": 80761653.0,
      "step": 36870
    },
    {
      "entropy": 5.636686182022094,
      "epoch": 3.6596863834855102,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003710311064845069,
      "loss": 4.9169,
      "mean_token_accuracy": 0.19410044848918914,
      "num_tokens": 80771135.0,
      "step": 36875
    },
    {
      "entropy": 5.682727098464966,
      "epoch": 3.6601826121476777,
      "grad_norm": 1.4375,
      "learning_rate": 0.000370999063043008,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18672192543745042,
      "num_tokens": 80782119.0,
      "step": 36880
    },
    {
      "entropy": 5.672715854644776,
      "epoch": 3.660678840809845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003709670172209393,
      "loss": 4.9893,
      "mean_token_accuracy": 0.18819754868745803,
      "num_tokens": 80793917.0,
      "step": 36885
    },
    {
      "entropy": 5.724987411499024,
      "epoch": 3.6611750694720127,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003709349690190952,
      "loss": 5.1094,
      "mean_token_accuracy": 0.17993849962949754,
      "num_tokens": 80807256.0,
      "step": 36890
    },
    {
      "entropy": 5.680075836181641,
      "epoch": 3.66167129813418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003709029184382708,
      "loss": 4.9645,
      "mean_token_accuracy": 0.1917562410235405,
      "num_tokens": 80817224.0,
      "step": 36895
    },
    {
      "entropy": 5.6518947124481205,
      "epoch": 3.6621675267963476,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003708708654792605,
      "loss": 4.9448,
      "mean_token_accuracy": 0.1897488683462143,
      "num_tokens": 80828692.0,
      "step": 36900
    },
    {
      "entropy": 5.683972358703613,
      "epoch": 3.6626637554585155,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037083881014285944,
      "loss": 5.0106,
      "mean_token_accuracy": 0.18764213919639589,
      "num_tokens": 80839201.0,
      "step": 36905
    },
    {
      "entropy": 5.750324058532715,
      "epoch": 3.663159984120683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00037080675242986234,
      "loss": 5.0754,
      "mean_token_accuracy": 0.1843709886074066,
      "num_tokens": 80853184.0,
      "step": 36910
    },
    {
      "entropy": 5.624285459518433,
      "epoch": 3.6636562127828505,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003707746923410642,
      "loss": 4.8988,
      "mean_token_accuracy": 0.1891788050532341,
      "num_tokens": 80862425.0,
      "step": 36915
    },
    {
      "entropy": 5.690069675445557,
      "epoch": 3.664152441445018,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037074262987726013,
      "loss": 4.9563,
      "mean_token_accuracy": 0.19377063661813737,
      "num_tokens": 80873652.0,
      "step": 36920
    },
    {
      "entropy": 5.653598499298096,
      "epoch": 3.6646486701071854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003707105650392451,
      "loss": 5.003,
      "mean_token_accuracy": 0.18579190075397492,
      "num_tokens": 80884505.0,
      "step": 36925
    },
    {
      "entropy": 5.636863660812378,
      "epoch": 3.665144898769353,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003706784978278144,
      "loss": 4.996,
      "mean_token_accuracy": 0.18806926757097245,
      "num_tokens": 80895728.0,
      "step": 36930
    },
    {
      "entropy": 5.735941982269287,
      "epoch": 3.6656411274315204,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003706464282437629,
      "loss": 4.9962,
      "mean_token_accuracy": 0.18926285207271576,
      "num_tokens": 80907054.0,
      "step": 36935
    },
    {
      "entropy": 5.708663845062256,
      "epoch": 3.666137356093688,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003706143562878863,
      "loss": 4.9642,
      "mean_token_accuracy": 0.19856008440256118,
      "num_tokens": 80917331.0,
      "step": 36940
    },
    {
      "entropy": 5.629128456115723,
      "epoch": 3.6666335847558553,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003705822819609796,
      "loss": 4.9555,
      "mean_token_accuracy": 0.18901679813861846,
      "num_tokens": 80929768.0,
      "step": 36945
    },
    {
      "entropy": 5.713042163848877,
      "epoch": 3.667129813418023,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003705502052638383,
      "loss": 5.0171,
      "mean_token_accuracy": 0.18168797791004182,
      "num_tokens": 80940226.0,
      "step": 36950
    },
    {
      "entropy": 5.647741842269897,
      "epoch": 3.6676260420801907,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00037051812619725785,
      "loss": 4.992,
      "mean_token_accuracy": 0.18432516604661942,
      "num_tokens": 80950622.0,
      "step": 36955
    },
    {
      "entropy": 5.6618493556976315,
      "epoch": 3.668122270742358,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003704860447620335,
      "loss": 4.9694,
      "mean_token_accuracy": 0.1869361862540245,
      "num_tokens": 80961439.0,
      "step": 36960
    },
    {
      "entropy": 5.702340507507325,
      "epoch": 3.6686184994045257,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003704539609589612,
      "loss": 5.0775,
      "mean_token_accuracy": 0.18091607093811035,
      "num_tokens": 80972898.0,
      "step": 36965
    },
    {
      "entropy": 5.606937837600708,
      "epoch": 3.669114728066693,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003704218747888363,
      "loss": 4.927,
      "mean_token_accuracy": 0.19921685308218,
      "num_tokens": 80984758.0,
      "step": 36970
    },
    {
      "entropy": 5.694287538528442,
      "epoch": 3.6696109567288606,
      "grad_norm": 1.265625,
      "learning_rate": 0.00037038978625245447,
      "loss": 4.9737,
      "mean_token_accuracy": 0.19402619898319245,
      "num_tokens": 80995987.0,
      "step": 36975
    },
    {
      "entropy": 5.73233380317688,
      "epoch": 3.670107185391028,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003703576953506115,
      "loss": 5.047,
      "mean_token_accuracy": 0.18662043958902358,
      "num_tokens": 81006834.0,
      "step": 36980
    },
    {
      "entropy": 5.709265327453613,
      "epoch": 3.6706034140531956,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00037032560208410313,
      "loss": 4.99,
      "mean_token_accuracy": 0.18901657909154893,
      "num_tokens": 81016273.0,
      "step": 36985
    },
    {
      "entropy": 5.656118154525757,
      "epoch": 3.6710996427153635,
      "grad_norm": 1.25,
      "learning_rate": 0.00037029350645372526,
      "loss": 4.9677,
      "mean_token_accuracy": 0.1922029122710228,
      "num_tokens": 81026292.0,
      "step": 36990
    },
    {
      "entropy": 5.633227062225342,
      "epoch": 3.671595871377531,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037026140846027383,
      "loss": 4.995,
      "mean_token_accuracy": 0.18553802222013474,
      "num_tokens": 81036824.0,
      "step": 36995
    },
    {
      "entropy": 5.6780040740966795,
      "epoch": 3.6720921000396984,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003702293081045448,
      "loss": 4.9788,
      "mean_token_accuracy": 0.19576750099658966,
      "num_tokens": 81047379.0,
      "step": 37000
    },
    {
      "entropy": 5.75179877281189,
      "epoch": 3.672588328701866,
      "grad_norm": 1.25,
      "learning_rate": 0.00037019720538733394,
      "loss": 5.1072,
      "mean_token_accuracy": 0.17822074741125107,
      "num_tokens": 81059097.0,
      "step": 37005
    },
    {
      "entropy": 5.725388765335083,
      "epoch": 3.6730845573640334,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003701651003094376,
      "loss": 5.0024,
      "mean_token_accuracy": 0.19039334803819657,
      "num_tokens": 81069747.0,
      "step": 37010
    },
    {
      "entropy": 5.709161567687988,
      "epoch": 3.673580786026201,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003701329928716519,
      "loss": 5.0328,
      "mean_token_accuracy": 0.188358810544014,
      "num_tokens": 81080947.0,
      "step": 37015
    },
    {
      "entropy": 5.727709579467773,
      "epoch": 3.6740770146883683,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037010088307477286,
      "loss": 5.002,
      "mean_token_accuracy": 0.19962458163499833,
      "num_tokens": 81091736.0,
      "step": 37020
    },
    {
      "entropy": 5.680815362930298,
      "epoch": 3.674573243350536,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003700687709195969,
      "loss": 4.9723,
      "mean_token_accuracy": 0.19976792633533477,
      "num_tokens": 81103289.0,
      "step": 37025
    },
    {
      "entropy": 5.665601110458374,
      "epoch": 3.6750694720127033,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003700366564069203,
      "loss": 4.9463,
      "mean_token_accuracy": 0.1981613129377365,
      "num_tokens": 81116738.0,
      "step": 37030
    },
    {
      "entropy": 5.665018892288208,
      "epoch": 3.6755657006748708,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00037000453953753943,
      "loss": 4.9526,
      "mean_token_accuracy": 0.19690635204315185,
      "num_tokens": 81126642.0,
      "step": 37035
    },
    {
      "entropy": 5.7530450344085695,
      "epoch": 3.6760619293370382,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003699724203122506,
      "loss": 5.0462,
      "mean_token_accuracy": 0.18811746537685395,
      "num_tokens": 81137744.0,
      "step": 37040
    },
    {
      "entropy": 5.594874143600464,
      "epoch": 3.676558157999206,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003699402987318504,
      "loss": 4.8662,
      "mean_token_accuracy": 0.19666044563055038,
      "num_tokens": 81149420.0,
      "step": 37045
    },
    {
      "entropy": 5.71316385269165,
      "epoch": 3.6770543866613736,
      "grad_norm": 1.25,
      "learning_rate": 0.00036990817479713543,
      "loss": 5.0398,
      "mean_token_accuracy": 0.18922048509120942,
      "num_tokens": 81158516.0,
      "step": 37050
    },
    {
      "entropy": 5.626112651824951,
      "epoch": 3.677550615323541,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036987604850890224,
      "loss": 4.9697,
      "mean_token_accuracy": 0.18887612521648406,
      "num_tokens": 81170230.0,
      "step": 37055
    },
    {
      "entropy": 5.659482526779175,
      "epoch": 3.6780468439857086,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036984391986794745,
      "loss": 4.9352,
      "mean_token_accuracy": 0.1933845102787018,
      "num_tokens": 81180841.0,
      "step": 37060
    },
    {
      "entropy": 5.690890312194824,
      "epoch": 3.678543072647876,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00036981178887506776,
      "loss": 5.0389,
      "mean_token_accuracy": 0.1882297307252884,
      "num_tokens": 81191808.0,
      "step": 37065
    },
    {
      "entropy": 5.698579931259156,
      "epoch": 3.6790393013100435,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036977965553106014,
      "loss": 4.9592,
      "mean_token_accuracy": 0.19031281471252443,
      "num_tokens": 81201326.0,
      "step": 37070
    },
    {
      "entropy": 5.693241882324219,
      "epoch": 3.6795355299722114,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003697475198367212,
      "loss": 4.9574,
      "mean_token_accuracy": 0.19190894663333893,
      "num_tokens": 81213458.0,
      "step": 37075
    },
    {
      "entropy": 5.694304752349853,
      "epoch": 3.680031758634379,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000369715381792848,
      "loss": 5.0653,
      "mean_token_accuracy": 0.1840776488184929,
      "num_tokens": 81225133.0,
      "step": 37080
    },
    {
      "entropy": 5.647092247009278,
      "epoch": 3.6805279872965464,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003696832414002374,
      "loss": 5.0619,
      "mean_token_accuracy": 0.18643631786108017,
      "num_tokens": 81237592.0,
      "step": 37085
    },
    {
      "entropy": 5.741287326812744,
      "epoch": 3.681024215958714,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003696510986596864,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19020403623580934,
      "num_tokens": 81247527.0,
      "step": 37090
    },
    {
      "entropy": 5.79331865310669,
      "epoch": 3.6815204446208813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036961895357199214,
      "loss": 5.1129,
      "mean_token_accuracy": 0.18105420470237732,
      "num_tokens": 81259109.0,
      "step": 37095
    },
    {
      "entropy": 5.683734035491943,
      "epoch": 3.682016673283049,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003695868061379518,
      "loss": 4.9962,
      "mean_token_accuracy": 0.19598653614521028,
      "num_tokens": 81269747.0,
      "step": 37100
    },
    {
      "entropy": 5.69773645401001,
      "epoch": 3.6825129019452163,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003695546563583623,
      "loss": 5.0531,
      "mean_token_accuracy": 0.18188595175743102,
      "num_tokens": 81279332.0,
      "step": 37105
    },
    {
      "entropy": 5.6331533908844,
      "epoch": 3.6830091306073838,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003695225042340212,
      "loss": 4.9144,
      "mean_token_accuracy": 0.20068598240613938,
      "num_tokens": 81290274.0,
      "step": 37110
    },
    {
      "entropy": 5.732427453994751,
      "epoch": 3.6835053592695512,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003694903497657256,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18859547078609468,
      "num_tokens": 81301665.0,
      "step": 37115
    },
    {
      "entropy": 5.718245506286621,
      "epoch": 3.6840015879317187,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000369458192954273,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1853310540318489,
      "num_tokens": 81311581.0,
      "step": 37120
    },
    {
      "entropy": 5.723367023468017,
      "epoch": 3.684497816593886,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003694260338004607,
      "loss": 5.069,
      "mean_token_accuracy": 0.1853489473462105,
      "num_tokens": 81322526.0,
      "step": 37125
    },
    {
      "entropy": 5.688404560089111,
      "epoch": 3.684994045256054,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036939387230508627,
      "loss": 4.9493,
      "mean_token_accuracy": 0.1899066150188446,
      "num_tokens": 81332132.0,
      "step": 37130
    },
    {
      "entropy": 5.703283357620239,
      "epoch": 3.6854902739182216,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036936170846894715,
      "loss": 4.9804,
      "mean_token_accuracy": 0.18831798136234285,
      "num_tokens": 81343425.0,
      "step": 37135
    },
    {
      "entropy": 5.667448663711548,
      "epoch": 3.685986502580389,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036932954229284106,
      "loss": 4.9576,
      "mean_token_accuracy": 0.19198692440986634,
      "num_tokens": 81354262.0,
      "step": 37140
    },
    {
      "entropy": 5.680183601379395,
      "epoch": 3.6864827312425565,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036929737377756563,
      "loss": 4.9502,
      "mean_token_accuracy": 0.18548403531312943,
      "num_tokens": 81365324.0,
      "step": 37145
    },
    {
      "entropy": 5.6281853199005125,
      "epoch": 3.686978959904724,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003692652029239184,
      "loss": 4.9408,
      "mean_token_accuracy": 0.19884274899959564,
      "num_tokens": 81376083.0,
      "step": 37150
    },
    {
      "entropy": 5.754215049743652,
      "epoch": 3.6874751885668915,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036923302973269727,
      "loss": 4.9806,
      "mean_token_accuracy": 0.18646748811006547,
      "num_tokens": 81386283.0,
      "step": 37155
    },
    {
      "entropy": 5.627856588363647,
      "epoch": 3.6879714172290594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003692008542047001,
      "loss": 4.9283,
      "mean_token_accuracy": 0.1908375069499016,
      "num_tokens": 81397878.0,
      "step": 37160
    },
    {
      "entropy": 5.690310525894165,
      "epoch": 3.688467645891227,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003691686763407248,
      "loss": 5.0538,
      "mean_token_accuracy": 0.19351317286491393,
      "num_tokens": 81409464.0,
      "step": 37165
    },
    {
      "entropy": 5.731747674942016,
      "epoch": 3.6889638745533944,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003691364961415691,
      "loss": 5.1163,
      "mean_token_accuracy": 0.1880739912390709,
      "num_tokens": 81421352.0,
      "step": 37170
    },
    {
      "entropy": 5.681969499588012,
      "epoch": 3.689460103215562,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036910431360803124,
      "loss": 4.8589,
      "mean_token_accuracy": 0.2005746990442276,
      "num_tokens": 81431045.0,
      "step": 37175
    },
    {
      "entropy": 5.6880388259887695,
      "epoch": 3.6899563318777293,
      "grad_norm": 1.25,
      "learning_rate": 0.00036907212874090923,
      "loss": 4.8953,
      "mean_token_accuracy": 0.20330401062965392,
      "num_tokens": 81441772.0,
      "step": 37180
    },
    {
      "entropy": 5.544881677627563,
      "epoch": 3.690452560539897,
      "grad_norm": 1.203125,
      "learning_rate": 0.000369039941541001,
      "loss": 4.9132,
      "mean_token_accuracy": 0.19341804683208466,
      "num_tokens": 81451964.0,
      "step": 37185
    },
    {
      "entropy": 5.6036036014556885,
      "epoch": 3.6909487892020643,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036900775200910497,
      "loss": 4.9268,
      "mean_token_accuracy": 0.1935271993279457,
      "num_tokens": 81462598.0,
      "step": 37190
    },
    {
      "entropy": 5.694213199615478,
      "epoch": 3.6914450178642317,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036897556014601926,
      "loss": 4.9935,
      "mean_token_accuracy": 0.1873597428202629,
      "num_tokens": 81474415.0,
      "step": 37195
    },
    {
      "entropy": 5.720984315872192,
      "epoch": 3.691941246526399,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036894336595254203,
      "loss": 5.0042,
      "mean_token_accuracy": 0.19004727452993392,
      "num_tokens": 81485510.0,
      "step": 37200
    },
    {
      "entropy": 5.699187898635865,
      "epoch": 3.6924374751885667,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036891116942947184,
      "loss": 5.038,
      "mean_token_accuracy": 0.18367286771535873,
      "num_tokens": 81496103.0,
      "step": 37205
    },
    {
      "entropy": 5.7078039169311525,
      "epoch": 3.692933703850734,
      "grad_norm": 1.078125,
      "learning_rate": 0.000368878970577607,
      "loss": 5.0813,
      "mean_token_accuracy": 0.18233431726694108,
      "num_tokens": 81507876.0,
      "step": 37210
    },
    {
      "entropy": 5.625004529953003,
      "epoch": 3.693429932512902,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036884676939774597,
      "loss": 4.9482,
      "mean_token_accuracy": 0.19128602892160415,
      "num_tokens": 81520615.0,
      "step": 37215
    },
    {
      "entropy": 5.619430875778198,
      "epoch": 3.6939261611750696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036881456589068725,
      "loss": 4.9644,
      "mean_token_accuracy": 0.18706940561532975,
      "num_tokens": 81531614.0,
      "step": 37220
    },
    {
      "entropy": 5.68661413192749,
      "epoch": 3.694422389837237,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036878236005722954,
      "loss": 4.9781,
      "mean_token_accuracy": 0.18777061402797698,
      "num_tokens": 81541027.0,
      "step": 37225
    },
    {
      "entropy": 5.695035457611084,
      "epoch": 3.6949186184994045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003687501518981713,
      "loss": 5.0037,
      "mean_token_accuracy": 0.18652433902025223,
      "num_tokens": 81551569.0,
      "step": 37230
    },
    {
      "entropy": 5.677881336212158,
      "epoch": 3.695414847161572,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003687179414143114,
      "loss": 4.9475,
      "mean_token_accuracy": 0.19200920015573503,
      "num_tokens": 81561490.0,
      "step": 37235
    },
    {
      "entropy": 5.607593441009522,
      "epoch": 3.6959110758237395,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036868572860644845,
      "loss": 4.8847,
      "mean_token_accuracy": 0.19434225857257842,
      "num_tokens": 81572434.0,
      "step": 37240
    },
    {
      "entropy": 5.643085384368897,
      "epoch": 3.6964073044859074,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036865351347538123,
      "loss": 4.9773,
      "mean_token_accuracy": 0.18618564754724504,
      "num_tokens": 81585135.0,
      "step": 37245
    },
    {
      "entropy": 5.600600957870483,
      "epoch": 3.696903533148075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003686212960219088,
      "loss": 4.8816,
      "mean_token_accuracy": 0.20296751260757445,
      "num_tokens": 81596919.0,
      "step": 37250
    },
    {
      "entropy": 5.662802600860596,
      "epoch": 3.6973997618102423,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036858907624682987,
      "loss": 4.9972,
      "mean_token_accuracy": 0.19102999866008757,
      "num_tokens": 81608335.0,
      "step": 37255
    },
    {
      "entropy": 5.6697920799255375,
      "epoch": 3.69789599047241,
      "grad_norm": 1.25,
      "learning_rate": 0.0003685568541509435,
      "loss": 4.9733,
      "mean_token_accuracy": 0.18982930034399031,
      "num_tokens": 81618940.0,
      "step": 37260
    },
    {
      "entropy": 5.662750482559204,
      "epoch": 3.6983922191345773,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036852462973504884,
      "loss": 4.9895,
      "mean_token_accuracy": 0.18960260450839997,
      "num_tokens": 81629215.0,
      "step": 37265
    },
    {
      "entropy": 5.700026512145996,
      "epoch": 3.6988884477967447,
      "grad_norm": 1.125,
      "learning_rate": 0.0003684924029999449,
      "loss": 5.0621,
      "mean_token_accuracy": 0.190073461830616,
      "num_tokens": 81642201.0,
      "step": 37270
    },
    {
      "entropy": 5.685253000259399,
      "epoch": 3.6993846764589122,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036846017394643064,
      "loss": 4.8948,
      "mean_token_accuracy": 0.1969260573387146,
      "num_tokens": 81652307.0,
      "step": 37275
    },
    {
      "entropy": 5.697873306274414,
      "epoch": 3.6998809051210797,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003684279425753057,
      "loss": 5.0189,
      "mean_token_accuracy": 0.19211190342903137,
      "num_tokens": 81662904.0,
      "step": 37280
    },
    {
      "entropy": 5.702719402313233,
      "epoch": 3.700377133783247,
      "grad_norm": 1.125,
      "learning_rate": 0.00036839570888736893,
      "loss": 5.0687,
      "mean_token_accuracy": 0.18486027270555497,
      "num_tokens": 81674456.0,
      "step": 37285
    },
    {
      "entropy": 5.6762392044067385,
      "epoch": 3.7008733624454146,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003683634728834199,
      "loss": 4.9364,
      "mean_token_accuracy": 0.1885800138115883,
      "num_tokens": 81685128.0,
      "step": 37290
    },
    {
      "entropy": 5.651086807250977,
      "epoch": 3.701369591107582,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036833123456425794,
      "loss": 4.9984,
      "mean_token_accuracy": 0.19505852609872817,
      "num_tokens": 81696661.0,
      "step": 37295
    },
    {
      "entropy": 5.74805793762207,
      "epoch": 3.70186581976975,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003682989939306824,
      "loss": 5.045,
      "mean_token_accuracy": 0.18423343747854232,
      "num_tokens": 81708095.0,
      "step": 37300
    },
    {
      "entropy": 5.749905157089233,
      "epoch": 3.7023620484319175,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003682667509834929,
      "loss": 5.0987,
      "mean_token_accuracy": 0.17950325161218644,
      "num_tokens": 81719798.0,
      "step": 37305
    },
    {
      "entropy": 5.6588091373443605,
      "epoch": 3.702858277094085,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036823450572348896,
      "loss": 4.9307,
      "mean_token_accuracy": 0.19204130619764329,
      "num_tokens": 81731041.0,
      "step": 37310
    },
    {
      "entropy": 5.7075989723205565,
      "epoch": 3.7033545057562525,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003682022581514702,
      "loss": 4.9686,
      "mean_token_accuracy": 0.18935954868793486,
      "num_tokens": 81742198.0,
      "step": 37315
    },
    {
      "entropy": 5.669908285140991,
      "epoch": 3.70385073441842,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036817000826823616,
      "loss": 4.9445,
      "mean_token_accuracy": 0.19149478673934936,
      "num_tokens": 81752824.0,
      "step": 37320
    },
    {
      "entropy": 5.646805477142334,
      "epoch": 3.7043469630805874,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036813775607458677,
      "loss": 4.961,
      "mean_token_accuracy": 0.19342196732759476,
      "num_tokens": 81764785.0,
      "step": 37325
    },
    {
      "entropy": 5.637331581115722,
      "epoch": 3.7048431917427553,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003681055015713217,
      "loss": 4.896,
      "mean_token_accuracy": 0.20317145586013793,
      "num_tokens": 81776755.0,
      "step": 37330
    },
    {
      "entropy": 5.7273622989654545,
      "epoch": 3.705339420404923,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003680732447592408,
      "loss": 4.9647,
      "mean_token_accuracy": 0.19176949113607406,
      "num_tokens": 81788088.0,
      "step": 37335
    },
    {
      "entropy": 5.673963928222657,
      "epoch": 3.7058356490670903,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036804098563914407,
      "loss": 5.0421,
      "mean_token_accuracy": 0.18584298640489577,
      "num_tokens": 81800114.0,
      "step": 37340
    },
    {
      "entropy": 5.702544689178467,
      "epoch": 3.7063318777292578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003680087242118313,
      "loss": 4.9907,
      "mean_token_accuracy": 0.18669967651367186,
      "num_tokens": 81811309.0,
      "step": 37345
    },
    {
      "entropy": 5.691465520858765,
      "epoch": 3.7068281063914252,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003679764604781026,
      "loss": 5.0493,
      "mean_token_accuracy": 0.18941646665334702,
      "num_tokens": 81821570.0,
      "step": 37350
    },
    {
      "entropy": 5.714661979675293,
      "epoch": 3.7073243350535927,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036794419443875805,
      "loss": 4.9294,
      "mean_token_accuracy": 0.19769223034381866,
      "num_tokens": 81832674.0,
      "step": 37355
    },
    {
      "entropy": 5.729397392272949,
      "epoch": 3.70782056371576,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003679119260945977,
      "loss": 5.0754,
      "mean_token_accuracy": 0.18319352716207504,
      "num_tokens": 81843463.0,
      "step": 37360
    },
    {
      "entropy": 5.679284143447876,
      "epoch": 3.7083167923779277,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003678796554464219,
      "loss": 4.9807,
      "mean_token_accuracy": 0.19995462596416474,
      "num_tokens": 81853512.0,
      "step": 37365
    },
    {
      "entropy": 5.7428717613220215,
      "epoch": 3.708813021040095,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036784738249503063,
      "loss": 5.0739,
      "mean_token_accuracy": 0.1849709689617157,
      "num_tokens": 81864703.0,
      "step": 37370
    },
    {
      "entropy": 5.6581205368042,
      "epoch": 3.7093092497022626,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003678151072412245,
      "loss": 4.9201,
      "mean_token_accuracy": 0.19674964398145675,
      "num_tokens": 81875387.0,
      "step": 37375
    },
    {
      "entropy": 5.6630213260650635,
      "epoch": 3.70980547836443,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036778282968580355,
      "loss": 4.9286,
      "mean_token_accuracy": 0.19357228130102158,
      "num_tokens": 81885908.0,
      "step": 37380
    },
    {
      "entropy": 5.746851062774658,
      "epoch": 3.710301707026598,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003677505498295685,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18802593350410463,
      "num_tokens": 81898256.0,
      "step": 37385
    },
    {
      "entropy": 5.628878355026245,
      "epoch": 3.7107979356887655,
      "grad_norm": 1.25,
      "learning_rate": 0.00036771826767331955,
      "loss": 4.9159,
      "mean_token_accuracy": 0.19700418412685394,
      "num_tokens": 81908993.0,
      "step": 37390
    },
    {
      "entropy": 5.607645702362061,
      "epoch": 3.711294164350933,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003676859832178575,
      "loss": 4.9505,
      "mean_token_accuracy": 0.19363687187433243,
      "num_tokens": 81919189.0,
      "step": 37395
    },
    {
      "entropy": 5.672506427764892,
      "epoch": 3.7117903930131004,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003676536964639827,
      "loss": 4.9302,
      "mean_token_accuracy": 0.19443976432085036,
      "num_tokens": 81929213.0,
      "step": 37400
    },
    {
      "entropy": 5.668020009994507,
      "epoch": 3.712286621675268,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003676214074124958,
      "loss": 4.978,
      "mean_token_accuracy": 0.18668819665908815,
      "num_tokens": 81940735.0,
      "step": 37405
    },
    {
      "entropy": 5.611828851699829,
      "epoch": 3.7127828503374354,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036758911606419773,
      "loss": 4.9361,
      "mean_token_accuracy": 0.18911614269018173,
      "num_tokens": 81951962.0,
      "step": 37410
    },
    {
      "entropy": 5.694208097457886,
      "epoch": 3.713279078999603,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003675568224198889,
      "loss": 4.94,
      "mean_token_accuracy": 0.18925395905971526,
      "num_tokens": 81963328.0,
      "step": 37415
    },
    {
      "entropy": 5.649926328659058,
      "epoch": 3.7137753076617708,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003675245264803705,
      "loss": 4.97,
      "mean_token_accuracy": 0.1897622674703598,
      "num_tokens": 81975318.0,
      "step": 37420
    },
    {
      "entropy": 5.664273691177368,
      "epoch": 3.7142715363239382,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003674922282464432,
      "loss": 4.9769,
      "mean_token_accuracy": 0.18867579102516174,
      "num_tokens": 81986036.0,
      "step": 37425
    },
    {
      "entropy": 5.780053520202637,
      "epoch": 3.7147677649861057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003674599277189078,
      "loss": 5.0872,
      "mean_token_accuracy": 0.17926553785800933,
      "num_tokens": 81998709.0,
      "step": 37430
    },
    {
      "entropy": 5.755236959457397,
      "epoch": 3.715263993648273,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036742762489856554,
      "loss": 5.0601,
      "mean_token_accuracy": 0.18243886679410934,
      "num_tokens": 82010436.0,
      "step": 37435
    },
    {
      "entropy": 5.7101037979125975,
      "epoch": 3.7157602223104407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003673953197862173,
      "loss": 5.0416,
      "mean_token_accuracy": 0.18469264954328538,
      "num_tokens": 82022194.0,
      "step": 37440
    },
    {
      "entropy": 5.68030276298523,
      "epoch": 3.716256450972608,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036736301238266434,
      "loss": 4.9976,
      "mean_token_accuracy": 0.18853570967912675,
      "num_tokens": 82033105.0,
      "step": 37445
    },
    {
      "entropy": 5.636376571655274,
      "epoch": 3.7167526796347756,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036733070268870754,
      "loss": 4.9351,
      "mean_token_accuracy": 0.18599424362182618,
      "num_tokens": 82043161.0,
      "step": 37450
    },
    {
      "entropy": 5.629429769515991,
      "epoch": 3.717248908296943,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003672983907051483,
      "loss": 4.9085,
      "mean_token_accuracy": 0.19930257350206376,
      "num_tokens": 82053889.0,
      "step": 37455
    },
    {
      "entropy": 5.695936822891236,
      "epoch": 3.7177451369591106,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00036726607643278793,
      "loss": 4.9907,
      "mean_token_accuracy": 0.1851869523525238,
      "num_tokens": 82066230.0,
      "step": 37460
    },
    {
      "entropy": 5.676793622970581,
      "epoch": 3.718241365621278,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036723375987242763,
      "loss": 4.9542,
      "mean_token_accuracy": 0.19120925664901733,
      "num_tokens": 82077655.0,
      "step": 37465
    },
    {
      "entropy": 5.750347137451172,
      "epoch": 3.718737594283446,
      "grad_norm": 1.25,
      "learning_rate": 0.00036720144102486886,
      "loss": 5.0407,
      "mean_token_accuracy": 0.18791109770536424,
      "num_tokens": 82089765.0,
      "step": 37470
    },
    {
      "entropy": 5.781395196914673,
      "epoch": 3.7192338229456134,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036716911989091304,
      "loss": 5.073,
      "mean_token_accuracy": 0.18082498461008073,
      "num_tokens": 82101095.0,
      "step": 37475
    },
    {
      "entropy": 5.641116619110107,
      "epoch": 3.719730051607781,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036713679647136163,
      "loss": 4.8933,
      "mean_token_accuracy": 0.20002776235342026,
      "num_tokens": 82111741.0,
      "step": 37480
    },
    {
      "entropy": 5.67317943572998,
      "epoch": 3.7202262802699484,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003671044707670161,
      "loss": 5.02,
      "mean_token_accuracy": 0.18753331899642944,
      "num_tokens": 82122749.0,
      "step": 37485
    },
    {
      "entropy": 5.658057117462159,
      "epoch": 3.720722508932116,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036707214277867834,
      "loss": 4.9466,
      "mean_token_accuracy": 0.1895798847079277,
      "num_tokens": 82132032.0,
      "step": 37490
    },
    {
      "entropy": 5.661992359161377,
      "epoch": 3.7212187375942833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003670398125071497,
      "loss": 5.011,
      "mean_token_accuracy": 0.18705885708332062,
      "num_tokens": 82143169.0,
      "step": 37495
    },
    {
      "entropy": 5.74250111579895,
      "epoch": 3.721714966256451,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036700747995323214,
      "loss": 5.0471,
      "mean_token_accuracy": 0.187747985124588,
      "num_tokens": 82154412.0,
      "step": 37500
    },
    {
      "entropy": 5.644176483154297,
      "epoch": 3.7222111949186187,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003669751451177273,
      "loss": 4.9557,
      "mean_token_accuracy": 0.1912587732076645,
      "num_tokens": 82165473.0,
      "step": 37505
    },
    {
      "entropy": 5.6764472961425785,
      "epoch": 3.722707423580786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000366942808001437,
      "loss": 4.9891,
      "mean_token_accuracy": 0.18891876190900803,
      "num_tokens": 82175716.0,
      "step": 37510
    },
    {
      "entropy": 5.688989591598511,
      "epoch": 3.7232036522429537,
      "grad_norm": 1.25,
      "learning_rate": 0.00036691046860516327,
      "loss": 5.0229,
      "mean_token_accuracy": 0.19166693687438965,
      "num_tokens": 82185981.0,
      "step": 37515
    },
    {
      "entropy": 5.729449510574341,
      "epoch": 3.723699880905121,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036687812692970796,
      "loss": 5.0466,
      "mean_token_accuracy": 0.18174461871385575,
      "num_tokens": 82197336.0,
      "step": 37520
    },
    {
      "entropy": 5.725175905227661,
      "epoch": 3.7241961095672886,
      "grad_norm": 1.125,
      "learning_rate": 0.000366845782975873,
      "loss": 5.0211,
      "mean_token_accuracy": 0.18990744799375534,
      "num_tokens": 82207895.0,
      "step": 37525
    },
    {
      "entropy": 5.655369901657105,
      "epoch": 3.724692338229456,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036681343674446054,
      "loss": 4.9359,
      "mean_token_accuracy": 0.19857245683670044,
      "num_tokens": 82218618.0,
      "step": 37530
    },
    {
      "entropy": 5.641321754455566,
      "epoch": 3.7251885668916236,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003667810882362728,
      "loss": 4.9918,
      "mean_token_accuracy": 0.18688336610794068,
      "num_tokens": 82229803.0,
      "step": 37535
    },
    {
      "entropy": 5.641556453704834,
      "epoch": 3.725684795553791,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003667487374521117,
      "loss": 4.962,
      "mean_token_accuracy": 0.1919868618249893,
      "num_tokens": 82239448.0,
      "step": 37540
    },
    {
      "entropy": 5.691271114349365,
      "epoch": 3.7261810242159585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003667163843927797,
      "loss": 4.9857,
      "mean_token_accuracy": 0.18740689605474473,
      "num_tokens": 82250093.0,
      "step": 37545
    },
    {
      "entropy": 5.673308372497559,
      "epoch": 3.726677252878126,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036668402905907897,
      "loss": 4.9744,
      "mean_token_accuracy": 0.1879489466547966,
      "num_tokens": 82260479.0,
      "step": 37550
    },
    {
      "entropy": 5.711430454254151,
      "epoch": 3.727173481540294,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036665167145181186,
      "loss": 5.0246,
      "mean_token_accuracy": 0.1863517478108406,
      "num_tokens": 82272482.0,
      "step": 37555
    },
    {
      "entropy": 5.677920246124268,
      "epoch": 3.7276697102024614,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036661931157178085,
      "loss": 4.9809,
      "mean_token_accuracy": 0.18816535025835038,
      "num_tokens": 82282456.0,
      "step": 37560
    },
    {
      "entropy": 5.766162490844726,
      "epoch": 3.728165938864629,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036658694941978825,
      "loss": 5.0565,
      "mean_token_accuracy": 0.18600278347730637,
      "num_tokens": 82292832.0,
      "step": 37565
    },
    {
      "entropy": 5.715641689300537,
      "epoch": 3.7286621675267964,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036655458499663675,
      "loss": 4.9707,
      "mean_token_accuracy": 0.1873958244919777,
      "num_tokens": 82302841.0,
      "step": 37570
    },
    {
      "entropy": 5.674414777755738,
      "epoch": 3.729158396188964,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003665222183031288,
      "loss": 4.9924,
      "mean_token_accuracy": 0.18933159708976746,
      "num_tokens": 82314092.0,
      "step": 37575
    },
    {
      "entropy": 5.726155614852905,
      "epoch": 3.7296546248511313,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000366489849340067,
      "loss": 5.1112,
      "mean_token_accuracy": 0.17798553109169007,
      "num_tokens": 82325153.0,
      "step": 37580
    },
    {
      "entropy": 5.7833061695098875,
      "epoch": 3.730150853513299,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036645747810825415,
      "loss": 5.0267,
      "mean_token_accuracy": 0.18943500518798828,
      "num_tokens": 82336463.0,
      "step": 37585
    },
    {
      "entropy": 5.686757898330688,
      "epoch": 3.7306470821754667,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036642510460849285,
      "loss": 4.9892,
      "mean_token_accuracy": 0.19403524845838546,
      "num_tokens": 82347314.0,
      "step": 37590
    },
    {
      "entropy": 5.677128839492798,
      "epoch": 3.731143310837634,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000366392728841586,
      "loss": 5.0075,
      "mean_token_accuracy": 0.1899850144982338,
      "num_tokens": 82358582.0,
      "step": 37595
    },
    {
      "entropy": 5.701727819442749,
      "epoch": 3.7316395394998016,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003663603508083364,
      "loss": 5.0277,
      "mean_token_accuracy": 0.18665633499622344,
      "num_tokens": 82369132.0,
      "step": 37600
    },
    {
      "entropy": 5.711297225952149,
      "epoch": 3.732135768161969,
      "grad_norm": 1.296875,
      "learning_rate": 0.000366327970509547,
      "loss": 5.0575,
      "mean_token_accuracy": 0.17991290390491485,
      "num_tokens": 82380011.0,
      "step": 37605
    },
    {
      "entropy": 5.675677156448364,
      "epoch": 3.7326319968241366,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036629558794602063,
      "loss": 5.0645,
      "mean_token_accuracy": 0.19002895653247834,
      "num_tokens": 82391113.0,
      "step": 37610
    },
    {
      "entropy": 5.725074386596679,
      "epoch": 3.733128225486304,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036626320311856046,
      "loss": 5.0124,
      "mean_token_accuracy": 0.18654888719320298,
      "num_tokens": 82402780.0,
      "step": 37615
    },
    {
      "entropy": 5.6536050796508786,
      "epoch": 3.7336244541484715,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036623081602796953,
      "loss": 4.9968,
      "mean_token_accuracy": 0.18407551050186158,
      "num_tokens": 82412005.0,
      "step": 37620
    },
    {
      "entropy": 5.65929102897644,
      "epoch": 3.734120682810639,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036619842667505096,
      "loss": 4.9805,
      "mean_token_accuracy": 0.1907728686928749,
      "num_tokens": 82420868.0,
      "step": 37625
    },
    {
      "entropy": 5.707687902450561,
      "epoch": 3.7346169114728065,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036616603506060786,
      "loss": 4.9831,
      "mean_token_accuracy": 0.1887217402458191,
      "num_tokens": 82430397.0,
      "step": 37630
    },
    {
      "entropy": 5.724308109283447,
      "epoch": 3.735113140134974,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003661336411854436,
      "loss": 4.9765,
      "mean_token_accuracy": 0.1970253735780716,
      "num_tokens": 82440567.0,
      "step": 37635
    },
    {
      "entropy": 5.727404499053955,
      "epoch": 3.7356093687971414,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003661012450503613,
      "loss": 4.9819,
      "mean_token_accuracy": 0.18958361595869064,
      "num_tokens": 82450920.0,
      "step": 37640
    },
    {
      "entropy": 5.648462104797363,
      "epoch": 3.7361055974593094,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00036606884665616455,
      "loss": 4.9906,
      "mean_token_accuracy": 0.1884343221783638,
      "num_tokens": 82461741.0,
      "step": 37645
    },
    {
      "entropy": 5.674598503112793,
      "epoch": 3.736601826121477,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003660364460036566,
      "loss": 4.9847,
      "mean_token_accuracy": 0.18985320031642913,
      "num_tokens": 82472360.0,
      "step": 37650
    },
    {
      "entropy": 5.620030403137207,
      "epoch": 3.7370980547836443,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003660040430936409,
      "loss": 4.9493,
      "mean_token_accuracy": 0.1812801405787468,
      "num_tokens": 82482758.0,
      "step": 37655
    },
    {
      "entropy": 5.667005491256714,
      "epoch": 3.737594283445812,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036597163792692107,
      "loss": 4.9019,
      "mean_token_accuracy": 0.192095747590065,
      "num_tokens": 82494059.0,
      "step": 37660
    },
    {
      "entropy": 5.708583927154541,
      "epoch": 3.7380905121079793,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036593923050430067,
      "loss": 5.0659,
      "mean_token_accuracy": 0.18552182316780091,
      "num_tokens": 82505503.0,
      "step": 37665
    },
    {
      "entropy": 5.72133321762085,
      "epoch": 3.7385867407701467,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036590682082658334,
      "loss": 5.0762,
      "mean_token_accuracy": 0.18932214230298997,
      "num_tokens": 82516669.0,
      "step": 37670
    },
    {
      "entropy": 5.6826153755187985,
      "epoch": 3.7390829694323147,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00036587440889457274,
      "loss": 4.945,
      "mean_token_accuracy": 0.18906888663768767,
      "num_tokens": 82526717.0,
      "step": 37675
    },
    {
      "entropy": 5.740807247161865,
      "epoch": 3.739579198094482,
      "grad_norm": 1.1875,
      "learning_rate": 0.00036584199470907254,
      "loss": 5.0604,
      "mean_token_accuracy": 0.1853153422474861,
      "num_tokens": 82536617.0,
      "step": 37680
    },
    {
      "entropy": 5.698573255538941,
      "epoch": 3.7400754267566496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003658095782708866,
      "loss": 5.0631,
      "mean_token_accuracy": 0.18197844922542572,
      "num_tokens": 82547858.0,
      "step": 37685
    },
    {
      "entropy": 5.671398305892945,
      "epoch": 3.740571655418817,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036577715958081885,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18688004463911057,
      "num_tokens": 82559334.0,
      "step": 37690
    },
    {
      "entropy": 5.677141857147217,
      "epoch": 3.7410678840809846,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003657447386396732,
      "loss": 4.981,
      "mean_token_accuracy": 0.19339718967676162,
      "num_tokens": 82570687.0,
      "step": 37695
    },
    {
      "entropy": 5.739489507675171,
      "epoch": 3.741564112743152,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003657123154482534,
      "loss": 5.0804,
      "mean_token_accuracy": 0.18433899134397508,
      "num_tokens": 82582030.0,
      "step": 37700
    },
    {
      "entropy": 5.618896484375,
      "epoch": 3.7420603414053195,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036567989000736386,
      "loss": 4.8954,
      "mean_token_accuracy": 0.1924571841955185,
      "num_tokens": 82592828.0,
      "step": 37705
    },
    {
      "entropy": 5.663987922668457,
      "epoch": 3.742556570067487,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036564746231780836,
      "loss": 4.9611,
      "mean_token_accuracy": 0.19277577102184296,
      "num_tokens": 82603133.0,
      "step": 37710
    },
    {
      "entropy": 5.640584182739258,
      "epoch": 3.7430527987296545,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003656150323803911,
      "loss": 4.9353,
      "mean_token_accuracy": 0.18851163238286972,
      "num_tokens": 82612889.0,
      "step": 37715
    },
    {
      "entropy": 5.631088018417358,
      "epoch": 3.743549027391822,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036558260019591633,
      "loss": 4.9816,
      "mean_token_accuracy": 0.18878804743289948,
      "num_tokens": 82623647.0,
      "step": 37720
    },
    {
      "entropy": 5.668596839904785,
      "epoch": 3.7440452560539894,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036555016576518816,
      "loss": 5.0116,
      "mean_token_accuracy": 0.18616752475500106,
      "num_tokens": 82633875.0,
      "step": 37725
    },
    {
      "entropy": 5.6780198097229,
      "epoch": 3.7445414847161573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003655177290890111,
      "loss": 5.0088,
      "mean_token_accuracy": 0.18909218460321425,
      "num_tokens": 82644596.0,
      "step": 37730
    },
    {
      "entropy": 5.691309928894043,
      "epoch": 3.745037713378325,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036548529016818936,
      "loss": 5.0048,
      "mean_token_accuracy": 0.1928626134991646,
      "num_tokens": 82656073.0,
      "step": 37735
    },
    {
      "entropy": 5.744250869750976,
      "epoch": 3.7455339420404923,
      "grad_norm": 1.390625,
      "learning_rate": 0.0003654528490035274,
      "loss": 5.0485,
      "mean_token_accuracy": 0.18422882854938508,
      "num_tokens": 82666644.0,
      "step": 37740
    },
    {
      "entropy": 5.674280214309692,
      "epoch": 3.7460301707026598,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003654204055958298,
      "loss": 4.9358,
      "mean_token_accuracy": 0.1911829009652138,
      "num_tokens": 82677812.0,
      "step": 37745
    },
    {
      "entropy": 5.66747407913208,
      "epoch": 3.7465263993648272,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036538795994590093,
      "loss": 5.0022,
      "mean_token_accuracy": 0.18828352987766267,
      "num_tokens": 82689215.0,
      "step": 37750
    },
    {
      "entropy": 5.687344598770141,
      "epoch": 3.7470226280269947,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036535551205454535,
      "loss": 5.0373,
      "mean_token_accuracy": 0.18873483091592788,
      "num_tokens": 82701115.0,
      "step": 37755
    },
    {
      "entropy": 5.685602617263794,
      "epoch": 3.7475188566891626,
      "grad_norm": 1.375,
      "learning_rate": 0.00036532306192256774,
      "loss": 5.0152,
      "mean_token_accuracy": 0.19167723804712294,
      "num_tokens": 82712064.0,
      "step": 37760
    },
    {
      "entropy": 5.697951030731201,
      "epoch": 3.74801508535133,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0003652906095507729,
      "loss": 5.0307,
      "mean_token_accuracy": 0.18437916785478592,
      "num_tokens": 82721515.0,
      "step": 37765
    },
    {
      "entropy": 5.701739835739136,
      "epoch": 3.7485113140134976,
      "grad_norm": 1.359375,
      "learning_rate": 0.00036525815493996557,
      "loss": 5.0799,
      "mean_token_accuracy": 0.1816338837146759,
      "num_tokens": 82731185.0,
      "step": 37770
    },
    {
      "entropy": 5.663391017913819,
      "epoch": 3.749007542675665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036522569809095034,
      "loss": 4.9625,
      "mean_token_accuracy": 0.19000980257987976,
      "num_tokens": 82740609.0,
      "step": 37775
    },
    {
      "entropy": 5.791698789596557,
      "epoch": 3.7495037713378325,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036519323900453235,
      "loss": 5.1694,
      "mean_token_accuracy": 0.18492878526449202,
      "num_tokens": 82752737.0,
      "step": 37780
    },
    {
      "entropy": 5.74487533569336,
      "epoch": 3.75,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036516077768151636,
      "loss": 5.07,
      "mean_token_accuracy": 0.18235134482383727,
      "num_tokens": 82763585.0,
      "step": 37785
    },
    {
      "entropy": 5.665353012084961,
      "epoch": 3.7504962286621675,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003651283141227073,
      "loss": 4.9666,
      "mean_token_accuracy": 0.18830618709325792,
      "num_tokens": 82774173.0,
      "step": 37790
    },
    {
      "entropy": 5.669465017318726,
      "epoch": 3.750992457324335,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003650958483289102,
      "loss": 5.0246,
      "mean_token_accuracy": 0.19315276741981507,
      "num_tokens": 82786279.0,
      "step": 37795
    },
    {
      "entropy": 5.629930067062378,
      "epoch": 3.7514886859865024,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003650633803009303,
      "loss": 4.9462,
      "mean_token_accuracy": 0.18727725148200988,
      "num_tokens": 82798087.0,
      "step": 37800
    },
    {
      "entropy": 5.732688760757446,
      "epoch": 3.75198491464867,
      "grad_norm": 1.3125,
      "learning_rate": 0.00036503091003957264,
      "loss": 4.9935,
      "mean_token_accuracy": 0.19988790303468704,
      "num_tokens": 82809712.0,
      "step": 37805
    },
    {
      "entropy": 5.619228839874268,
      "epoch": 3.7524811433108374,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036499843754564235,
      "loss": 4.9211,
      "mean_token_accuracy": 0.19861538261175155,
      "num_tokens": 82821894.0,
      "step": 37810
    },
    {
      "entropy": 5.685818004608154,
      "epoch": 3.7529773719730053,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003649659628199448,
      "loss": 4.9677,
      "mean_token_accuracy": 0.19649874716997145,
      "num_tokens": 82833094.0,
      "step": 37815
    },
    {
      "entropy": 5.641027021408081,
      "epoch": 3.7534736006351728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003649334858632852,
      "loss": 4.9565,
      "mean_token_accuracy": 0.1941981077194214,
      "num_tokens": 82843690.0,
      "step": 37820
    },
    {
      "entropy": 5.693260669708252,
      "epoch": 3.7539698292973402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000364901006676469,
      "loss": 5.0091,
      "mean_token_accuracy": 0.18994226157665253,
      "num_tokens": 82855810.0,
      "step": 37825
    },
    {
      "entropy": 5.669996976852417,
      "epoch": 3.7544660579595077,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036486852526030147,
      "loss": 4.9057,
      "mean_token_accuracy": 0.1965497076511383,
      "num_tokens": 82866836.0,
      "step": 37830
    },
    {
      "entropy": 5.689925575256348,
      "epoch": 3.754962286621675,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003648360416155882,
      "loss": 5.0271,
      "mean_token_accuracy": 0.19307169914245606,
      "num_tokens": 82877645.0,
      "step": 37835
    },
    {
      "entropy": 5.693251848220825,
      "epoch": 3.7554585152838427,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003648035557431347,
      "loss": 5.0046,
      "mean_token_accuracy": 0.18994068056344987,
      "num_tokens": 82888638.0,
      "step": 37840
    },
    {
      "entropy": 5.645466089248657,
      "epoch": 3.7559547439460106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036477106764374646,
      "loss": 4.9754,
      "mean_token_accuracy": 0.19453135281801223,
      "num_tokens": 82898550.0,
      "step": 37845
    },
    {
      "entropy": 5.71600112915039,
      "epoch": 3.756450972608178,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003647385773182292,
      "loss": 5.0215,
      "mean_token_accuracy": 0.1934506192803383,
      "num_tokens": 82908926.0,
      "step": 37850
    },
    {
      "entropy": 5.7631604194641115,
      "epoch": 3.7569472012703455,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036470608476738865,
      "loss": 5.066,
      "mean_token_accuracy": 0.1954063132405281,
      "num_tokens": 82921294.0,
      "step": 37855
    },
    {
      "entropy": 5.748652505874634,
      "epoch": 3.757443429932513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003646735899920304,
      "loss": 5.0341,
      "mean_token_accuracy": 0.18821719884872437,
      "num_tokens": 82931315.0,
      "step": 37860
    },
    {
      "entropy": 5.719536542892456,
      "epoch": 3.7579396585946805,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036464109299296047,
      "loss": 4.9998,
      "mean_token_accuracy": 0.18553048521280288,
      "num_tokens": 82942385.0,
      "step": 37865
    },
    {
      "entropy": 5.681225728988648,
      "epoch": 3.758435887256848,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003646085937709845,
      "loss": 5.0229,
      "mean_token_accuracy": 0.18970461338758468,
      "num_tokens": 82952803.0,
      "step": 37870
    },
    {
      "entropy": 5.680458879470825,
      "epoch": 3.7589321159190154,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003645760923269085,
      "loss": 5.0049,
      "mean_token_accuracy": 0.18190168142318724,
      "num_tokens": 82964143.0,
      "step": 37875
    },
    {
      "entropy": 5.701639938354492,
      "epoch": 3.759428344581183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003645435886615385,
      "loss": 4.9446,
      "mean_token_accuracy": 0.18910832852125167,
      "num_tokens": 82975800.0,
      "step": 37880
    },
    {
      "entropy": 5.719209480285644,
      "epoch": 3.7599245732433504,
      "grad_norm": 1.28125,
      "learning_rate": 0.00036451108277568044,
      "loss": 5.0282,
      "mean_token_accuracy": 0.1848834902048111,
      "num_tokens": 82987462.0,
      "step": 37885
    },
    {
      "entropy": 5.63855299949646,
      "epoch": 3.760420801905518,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036447857467014036,
      "loss": 4.9274,
      "mean_token_accuracy": 0.19703784584999084,
      "num_tokens": 82996754.0,
      "step": 37890
    },
    {
      "entropy": 5.65360746383667,
      "epoch": 3.7609170305676853,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003644460643457245,
      "loss": 4.9656,
      "mean_token_accuracy": 0.18938966244459152,
      "num_tokens": 83009026.0,
      "step": 37895
    },
    {
      "entropy": 5.744311046600342,
      "epoch": 3.7614132592298533,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036441355180323885,
      "loss": 5.0753,
      "mean_token_accuracy": 0.18117535263299941,
      "num_tokens": 83020451.0,
      "step": 37900
    },
    {
      "entropy": 5.6752537250518795,
      "epoch": 3.7619094878920207,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003643810370434899,
      "loss": 4.9414,
      "mean_token_accuracy": 0.19451479613780975,
      "num_tokens": 83030822.0,
      "step": 37905
    },
    {
      "entropy": 5.677247858047485,
      "epoch": 3.762405716554188,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003643485200672838,
      "loss": 4.9788,
      "mean_token_accuracy": 0.19817968755960463,
      "num_tokens": 83042441.0,
      "step": 37910
    },
    {
      "entropy": 5.684012413024902,
      "epoch": 3.7629019452163557,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036431600087542695,
      "loss": 5.0041,
      "mean_token_accuracy": 0.18753558099269868,
      "num_tokens": 83054386.0,
      "step": 37915
    },
    {
      "entropy": 5.70242862701416,
      "epoch": 3.763398173878523,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003642834794687257,
      "loss": 5.0043,
      "mean_token_accuracy": 0.18379670530557632,
      "num_tokens": 83066221.0,
      "step": 37920
    },
    {
      "entropy": 5.673607349395752,
      "epoch": 3.7638944025406906,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003642509558479865,
      "loss": 5.0405,
      "mean_token_accuracy": 0.1879077062010765,
      "num_tokens": 83078080.0,
      "step": 37925
    },
    {
      "entropy": 5.709010601043701,
      "epoch": 3.7643906312028586,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036421843001401607,
      "loss": 4.9969,
      "mean_token_accuracy": 0.19908546209335326,
      "num_tokens": 83089486.0,
      "step": 37930
    },
    {
      "entropy": 5.742274761199951,
      "epoch": 3.764886859865026,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036418590196762076,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18581126928329467,
      "num_tokens": 83101253.0,
      "step": 37935
    },
    {
      "entropy": 5.68352313041687,
      "epoch": 3.7653830885271935,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036415337170960714,
      "loss": 4.9352,
      "mean_token_accuracy": 0.19539244920015336,
      "num_tokens": 83112098.0,
      "step": 37940
    },
    {
      "entropy": 5.683777379989624,
      "epoch": 3.765879317189361,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036412083924078223,
      "loss": 4.9553,
      "mean_token_accuracy": 0.18848340958356857,
      "num_tokens": 83122047.0,
      "step": 37945
    },
    {
      "entropy": 5.684715461730957,
      "epoch": 3.7663755458515285,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0003640883045619523,
      "loss": 4.9834,
      "mean_token_accuracy": 0.18607122749090194,
      "num_tokens": 83132525.0,
      "step": 37950
    },
    {
      "entropy": 5.683867311477661,
      "epoch": 3.766871774513696,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003640557676739245,
      "loss": 5.0269,
      "mean_token_accuracy": 0.18493961542844772,
      "num_tokens": 83143274.0,
      "step": 37955
    },
    {
      "entropy": 5.729437017440796,
      "epoch": 3.7673680031758634,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003640232285775056,
      "loss": 5.066,
      "mean_token_accuracy": 0.19041728228330612,
      "num_tokens": 83154451.0,
      "step": 37960
    },
    {
      "entropy": 5.659969806671143,
      "epoch": 3.767864231838031,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036399068727350245,
      "loss": 4.9886,
      "mean_token_accuracy": 0.18119220286607743,
      "num_tokens": 83165535.0,
      "step": 37965
    },
    {
      "entropy": 5.623429870605468,
      "epoch": 3.7683604605001984,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00036395814376272194,
      "loss": 4.8889,
      "mean_token_accuracy": 0.20223002582788469,
      "num_tokens": 83175955.0,
      "step": 37970
    },
    {
      "entropy": 5.695733976364136,
      "epoch": 3.768856689162366,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003639255980459713,
      "loss": 5.0537,
      "mean_token_accuracy": 0.17866081148386,
      "num_tokens": 83186200.0,
      "step": 37975
    },
    {
      "entropy": 5.744284677505493,
      "epoch": 3.7693529178245333,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003638930501240572,
      "loss": 5.0144,
      "mean_token_accuracy": 0.18177436292171478,
      "num_tokens": 83197121.0,
      "step": 37980
    },
    {
      "entropy": 5.64689621925354,
      "epoch": 3.769849146486701,
      "grad_norm": 1.125,
      "learning_rate": 0.00036386049999778724,
      "loss": 4.9571,
      "mean_token_accuracy": 0.1942042216658592,
      "num_tokens": 83208580.0,
      "step": 37985
    },
    {
      "entropy": 5.660580205917358,
      "epoch": 3.7703453751488687,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036382794766796824,
      "loss": 4.8968,
      "mean_token_accuracy": 0.19654601216316223,
      "num_tokens": 83218589.0,
      "step": 37990
    },
    {
      "entropy": 5.646065330505371,
      "epoch": 3.770841603811036,
      "grad_norm": 1.25,
      "learning_rate": 0.00036379539313540757,
      "loss": 4.9993,
      "mean_token_accuracy": 0.19035040885210036,
      "num_tokens": 83229712.0,
      "step": 37995
    },
    {
      "entropy": 5.625918912887573,
      "epoch": 3.7713378324732036,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003637628364009125,
      "loss": 4.9397,
      "mean_token_accuracy": 0.1988181874155998,
      "num_tokens": 83240327.0,
      "step": 38000
    },
    {
      "entropy": 5.605123615264892,
      "epoch": 3.771834061135371,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036373027746529037,
      "loss": 4.9604,
      "mean_token_accuracy": 0.19425656497478486,
      "num_tokens": 83251966.0,
      "step": 38005
    },
    {
      "entropy": 5.658561706542969,
      "epoch": 3.7723302897975386,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003636977163293485,
      "loss": 4.9102,
      "mean_token_accuracy": 0.19077570885419845,
      "num_tokens": 83261948.0,
      "step": 38010
    },
    {
      "entropy": 5.694021606445313,
      "epoch": 3.772826518459706,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003636651529938944,
      "loss": 5.0085,
      "mean_token_accuracy": 0.1869366943836212,
      "num_tokens": 83273172.0,
      "step": 38015
    },
    {
      "entropy": 5.593960618972778,
      "epoch": 3.773322747121874,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003636325874597356,
      "loss": 4.9318,
      "mean_token_accuracy": 0.19103471785783768,
      "num_tokens": 83283836.0,
      "step": 38020
    },
    {
      "entropy": 5.627335166931152,
      "epoch": 3.7738189757840415,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036360001972767956,
      "loss": 5.0198,
      "mean_token_accuracy": 0.19203227758407593,
      "num_tokens": 83294206.0,
      "step": 38025
    },
    {
      "entropy": 5.699480485916138,
      "epoch": 3.774315204446209,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036356744979853406,
      "loss": 4.9974,
      "mean_token_accuracy": 0.19265709221363067,
      "num_tokens": 83305066.0,
      "step": 38030
    },
    {
      "entropy": 5.639671850204468,
      "epoch": 3.7748114331083764,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036353487767310655,
      "loss": 4.9326,
      "mean_token_accuracy": 0.1921784460544586,
      "num_tokens": 83316169.0,
      "step": 38035
    },
    {
      "entropy": 5.736663007736206,
      "epoch": 3.775307661770544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003635023033522049,
      "loss": 4.9998,
      "mean_token_accuracy": 0.19305215775966644,
      "num_tokens": 83327702.0,
      "step": 38040
    },
    {
      "entropy": 5.683598041534424,
      "epoch": 3.7758038904327114,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003634697268366368,
      "loss": 4.9712,
      "mean_token_accuracy": 0.18876330852508544,
      "num_tokens": 83339951.0,
      "step": 38045
    },
    {
      "entropy": 5.6821324825286865,
      "epoch": 3.776300119094879,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003634371481272101,
      "loss": 4.9576,
      "mean_token_accuracy": 0.18940298557281493,
      "num_tokens": 83351869.0,
      "step": 38050
    },
    {
      "entropy": 5.6793934345245365,
      "epoch": 3.7767963477570463,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003634045672247327,
      "loss": 5.029,
      "mean_token_accuracy": 0.18559196591377258,
      "num_tokens": 83362715.0,
      "step": 38055
    },
    {
      "entropy": 5.658660840988159,
      "epoch": 3.777292576419214,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003633719841300126,
      "loss": 4.9437,
      "mean_token_accuracy": 0.19309339672327042,
      "num_tokens": 83372988.0,
      "step": 38060
    },
    {
      "entropy": 5.616121244430542,
      "epoch": 3.7777888050813813,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003633393988438576,
      "loss": 4.9356,
      "mean_token_accuracy": 0.19359594583511353,
      "num_tokens": 83383172.0,
      "step": 38065
    },
    {
      "entropy": 5.603335285186768,
      "epoch": 3.778285033743549,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003633068113670759,
      "loss": 4.9338,
      "mean_token_accuracy": 0.19010725766420364,
      "num_tokens": 83395402.0,
      "step": 38070
    },
    {
      "entropy": 5.698479127883911,
      "epoch": 3.7787812624057167,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036327422170047565,
      "loss": 5.0929,
      "mean_token_accuracy": 0.18677461594343187,
      "num_tokens": 83406936.0,
      "step": 38075
    },
    {
      "entropy": 5.73558874130249,
      "epoch": 3.779277491067884,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003632416298448648,
      "loss": 5.0828,
      "mean_token_accuracy": 0.18044131845235825,
      "num_tokens": 83420043.0,
      "step": 38080
    },
    {
      "entropy": 5.690941572189331,
      "epoch": 3.7797737197300516,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00036320903580105176,
      "loss": 4.9755,
      "mean_token_accuracy": 0.1972062587738037,
      "num_tokens": 83430083.0,
      "step": 38085
    },
    {
      "entropy": 5.654227256774902,
      "epoch": 3.780269948392219,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036317643956984466,
      "loss": 4.888,
      "mean_token_accuracy": 0.19775771647691726,
      "num_tokens": 83440992.0,
      "step": 38090
    },
    {
      "entropy": 5.709534072875977,
      "epoch": 3.7807661770543866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036314384115205184,
      "loss": 4.9937,
      "mean_token_accuracy": 0.19158503413200378,
      "num_tokens": 83452534.0,
      "step": 38095
    },
    {
      "entropy": 5.623320293426514,
      "epoch": 3.781262405716554,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003631112405484817,
      "loss": 4.906,
      "mean_token_accuracy": 0.19468116760253906,
      "num_tokens": 83463064.0,
      "step": 38100
    },
    {
      "entropy": 5.714684677124024,
      "epoch": 3.781758634378722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036307863775994275,
      "loss": 5.0131,
      "mean_token_accuracy": 0.18903470933437347,
      "num_tokens": 83474426.0,
      "step": 38105
    },
    {
      "entropy": 5.679113435745239,
      "epoch": 3.7822548630408894,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036304603278724325,
      "loss": 5.0223,
      "mean_token_accuracy": 0.18584718108177184,
      "num_tokens": 83485896.0,
      "step": 38110
    },
    {
      "entropy": 5.696738338470459,
      "epoch": 3.782751091703057,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036301342563119194,
      "loss": 5.0334,
      "mean_token_accuracy": 0.1873920366168022,
      "num_tokens": 83496257.0,
      "step": 38115
    },
    {
      "entropy": 5.691024303436279,
      "epoch": 3.7832473203652244,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003629808162925973,
      "loss": 4.9823,
      "mean_token_accuracy": 0.19248959869146348,
      "num_tokens": 83506500.0,
      "step": 38120
    },
    {
      "entropy": 5.750270795822144,
      "epoch": 3.783743549027392,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00036294820477226797,
      "loss": 4.9894,
      "mean_token_accuracy": 0.1877605527639389,
      "num_tokens": 83517226.0,
      "step": 38125
    },
    {
      "entropy": 5.631587314605713,
      "epoch": 3.7842397776895593,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003629155910710127,
      "loss": 4.9755,
      "mean_token_accuracy": 0.19441553354263305,
      "num_tokens": 83527307.0,
      "step": 38130
    },
    {
      "entropy": 5.667460203170776,
      "epoch": 3.784736006351727,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003628829751896403,
      "loss": 4.9848,
      "mean_token_accuracy": 0.19308729618787765,
      "num_tokens": 83539252.0,
      "step": 38135
    },
    {
      "entropy": 5.703604602813721,
      "epoch": 3.7852322350138943,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00036285035712895934,
      "loss": 5.0079,
      "mean_token_accuracy": 0.18861342817544938,
      "num_tokens": 83551791.0,
      "step": 38140
    },
    {
      "entropy": 5.662324476242065,
      "epoch": 3.7857284636760618,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036281773688977894,
      "loss": 4.9372,
      "mean_token_accuracy": 0.18884475380182267,
      "num_tokens": 83561545.0,
      "step": 38145
    },
    {
      "entropy": 5.62384443283081,
      "epoch": 3.7862246923382292,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003627851144729078,
      "loss": 4.9228,
      "mean_token_accuracy": 0.19547986090183259,
      "num_tokens": 83572050.0,
      "step": 38150
    },
    {
      "entropy": 5.656531429290771,
      "epoch": 3.786720921000397,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036275248987915497,
      "loss": 5.0259,
      "mean_token_accuracy": 0.19181205183267594,
      "num_tokens": 83583177.0,
      "step": 38155
    },
    {
      "entropy": 5.677211141586303,
      "epoch": 3.7872171496625646,
      "grad_norm": 1.375,
      "learning_rate": 0.00036271986310932957,
      "loss": 4.9709,
      "mean_token_accuracy": 0.19102171510457994,
      "num_tokens": 83592906.0,
      "step": 38160
    },
    {
      "entropy": 5.663853025436401,
      "epoch": 3.787713378324732,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003626872341642406,
      "loss": 5.0343,
      "mean_token_accuracy": 0.18943773508071898,
      "num_tokens": 83602658.0,
      "step": 38165
    },
    {
      "entropy": 5.680257081985474,
      "epoch": 3.7882096069868996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003626546030446971,
      "loss": 5.0142,
      "mean_token_accuracy": 0.19323610663414,
      "num_tokens": 83612774.0,
      "step": 38170
    },
    {
      "entropy": 5.6992048740386965,
      "epoch": 3.788705835649067,
      "grad_norm": 1.328125,
      "learning_rate": 0.00036262196975150835,
      "loss": 4.9474,
      "mean_token_accuracy": 0.19885327368974687,
      "num_tokens": 83624284.0,
      "step": 38175
    },
    {
      "entropy": 5.705643558502198,
      "epoch": 3.7892020643112345,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003625893342854836,
      "loss": 5.0487,
      "mean_token_accuracy": 0.18507340550422668,
      "num_tokens": 83636850.0,
      "step": 38180
    },
    {
      "entropy": 5.699598979949951,
      "epoch": 3.789698292973402,
      "grad_norm": 1.171875,
      "learning_rate": 0.00036255669664743195,
      "loss": 5.0086,
      "mean_token_accuracy": 0.18667099475860596,
      "num_tokens": 83647562.0,
      "step": 38185
    },
    {
      "entropy": 5.664904975891114,
      "epoch": 3.79019452163557,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00036252405683816306,
      "loss": 4.9637,
      "mean_token_accuracy": 0.19515821784734727,
      "num_tokens": 83659020.0,
      "step": 38190
    },
    {
      "entropy": 5.689760446548462,
      "epoch": 3.7906907502977374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003624914148584861,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18574869334697724,
      "num_tokens": 83670162.0,
      "step": 38195
    },
    {
      "entropy": 5.612149620056153,
      "epoch": 3.791186978959905,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003624587707092105,
      "loss": 4.9126,
      "mean_token_accuracy": 0.18896340131759642,
      "num_tokens": 83680609.0,
      "step": 38200
    },
    {
      "entropy": 5.685749340057373,
      "epoch": 3.7916832076220723,
      "grad_norm": 1.296875,
      "learning_rate": 0.00036242612439114593,
      "loss": 4.9506,
      "mean_token_accuracy": 0.19175257086753844,
      "num_tokens": 83690912.0,
      "step": 38205
    },
    {
      "entropy": 5.695581531524658,
      "epoch": 3.79217943628424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036239347590510186,
      "loss": 5.0378,
      "mean_token_accuracy": 0.1836041897535324,
      "num_tokens": 83702022.0,
      "step": 38210
    },
    {
      "entropy": 5.748956537246704,
      "epoch": 3.7926756649464073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003623608252518879,
      "loss": 5.0766,
      "mean_token_accuracy": 0.18258372843265533,
      "num_tokens": 83713836.0,
      "step": 38215
    },
    {
      "entropy": 5.716650342941284,
      "epoch": 3.7931718936085748,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003623281724323137,
      "loss": 5.0689,
      "mean_token_accuracy": 0.186294724047184,
      "num_tokens": 83725038.0,
      "step": 38220
    },
    {
      "entropy": 5.654089117050171,
      "epoch": 3.7936681222707422,
      "grad_norm": 1.21875,
      "learning_rate": 0.000362295517447189,
      "loss": 4.9278,
      "mean_token_accuracy": 0.18927232027053834,
      "num_tokens": 83734984.0,
      "step": 38225
    },
    {
      "entropy": 5.709114456176758,
      "epoch": 3.7941643509329097,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003622628602973236,
      "loss": 4.9925,
      "mean_token_accuracy": 0.19087498933076857,
      "num_tokens": 83746323.0,
      "step": 38230
    },
    {
      "entropy": 5.6189357280731205,
      "epoch": 3.794660579595077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036223020098352724,
      "loss": 4.9264,
      "mean_token_accuracy": 0.18888114392757416,
      "num_tokens": 83758334.0,
      "step": 38235
    },
    {
      "entropy": 5.68665714263916,
      "epoch": 3.7951568082572447,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003621975395066099,
      "loss": 4.9503,
      "mean_token_accuracy": 0.1913865849375725,
      "num_tokens": 83769022.0,
      "step": 38240
    },
    {
      "entropy": 5.737388563156128,
      "epoch": 3.7956530369194126,
      "grad_norm": 1.546875,
      "learning_rate": 0.00036216487586738144,
      "loss": 5.0333,
      "mean_token_accuracy": 0.18896856307983398,
      "num_tokens": 83780682.0,
      "step": 38245
    },
    {
      "entropy": 5.697950029373169,
      "epoch": 3.79614926558158,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000362132210066652,
      "loss": 5.0109,
      "mean_token_accuracy": 0.18569873422384262,
      "num_tokens": 83790936.0,
      "step": 38250
    },
    {
      "entropy": 5.708381414413452,
      "epoch": 3.7966454942437475,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036209954210523137,
      "loss": 5.0565,
      "mean_token_accuracy": 0.18269238024950027,
      "num_tokens": 83802498.0,
      "step": 38255
    },
    {
      "entropy": 5.621304368972778,
      "epoch": 3.797141722905915,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003620668719839298,
      "loss": 4.9544,
      "mean_token_accuracy": 0.1915568932890892,
      "num_tokens": 83813269.0,
      "step": 38260
    },
    {
      "entropy": 5.6427075386047365,
      "epoch": 3.7976379515680825,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036203419970355747,
      "loss": 4.9703,
      "mean_token_accuracy": 0.18901339471340178,
      "num_tokens": 83822692.0,
      "step": 38265
    },
    {
      "entropy": 5.7171190738677975,
      "epoch": 3.79813418023025,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003620015252649245,
      "loss": 5.0491,
      "mean_token_accuracy": 0.18906441777944566,
      "num_tokens": 83834375.0,
      "step": 38270
    },
    {
      "entropy": 5.759387874603272,
      "epoch": 3.798630408892418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036196884866884115,
      "loss": 5.1062,
      "mean_token_accuracy": 0.18442731350660324,
      "num_tokens": 83846328.0,
      "step": 38275
    },
    {
      "entropy": 5.703217029571533,
      "epoch": 3.7991266375545854,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036193616991611784,
      "loss": 4.9808,
      "mean_token_accuracy": 0.18418368846178054,
      "num_tokens": 83857086.0,
      "step": 38280
    },
    {
      "entropy": 5.75643482208252,
      "epoch": 3.799622866216753,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00036190348900756475,
      "loss": 5.0503,
      "mean_token_accuracy": 0.18451854288578035,
      "num_tokens": 83867152.0,
      "step": 38285
    },
    {
      "entropy": 5.7145918846130375,
      "epoch": 3.8001190948789203,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003618708059439924,
      "loss": 4.9927,
      "mean_token_accuracy": 0.1916322410106659,
      "num_tokens": 83878960.0,
      "step": 38290
    },
    {
      "entropy": 5.6095921993255615,
      "epoch": 3.8006153235410878,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003618381207262113,
      "loss": 4.928,
      "mean_token_accuracy": 0.1942228302359581,
      "num_tokens": 83890656.0,
      "step": 38295
    },
    {
      "entropy": 5.655356311798096,
      "epoch": 3.8011115522032553,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003618054333550319,
      "loss": 4.9911,
      "mean_token_accuracy": 0.19104042649269104,
      "num_tokens": 83901099.0,
      "step": 38300
    },
    {
      "entropy": 5.668024921417237,
      "epoch": 3.8016077808654227,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036177274383126487,
      "loss": 5.0274,
      "mean_token_accuracy": 0.18756632059812545,
      "num_tokens": 83912016.0,
      "step": 38305
    },
    {
      "entropy": 5.728113031387329,
      "epoch": 3.80210400952759,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003617400521557207,
      "loss": 5.0404,
      "mean_token_accuracy": 0.18447503447532654,
      "num_tokens": 83923198.0,
      "step": 38310
    },
    {
      "entropy": 5.69767279624939,
      "epoch": 3.8026002381897577,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0003617073583292101,
      "loss": 5.058,
      "mean_token_accuracy": 0.18377214670181274,
      "num_tokens": 83935310.0,
      "step": 38315
    },
    {
      "entropy": 5.647762298583984,
      "epoch": 3.803096466851925,
      "grad_norm": 1.234375,
      "learning_rate": 0.00036167466235254394,
      "loss": 4.969,
      "mean_token_accuracy": 0.19064048379659654,
      "num_tokens": 83947104.0,
      "step": 38320
    },
    {
      "entropy": 5.667890977859497,
      "epoch": 3.8035926955140926,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00036164196422653293,
      "loss": 4.9772,
      "mean_token_accuracy": 0.18688036799430846,
      "num_tokens": 83957263.0,
      "step": 38325
    },
    {
      "entropy": 5.650736618041992,
      "epoch": 3.8040889241762605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036160926395198786,
      "loss": 4.9333,
      "mean_token_accuracy": 0.19535560607910157,
      "num_tokens": 83970680.0,
      "step": 38330
    },
    {
      "entropy": 5.708948707580566,
      "epoch": 3.804585152838428,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003615765615297197,
      "loss": 5.0428,
      "mean_token_accuracy": 0.18624064773321153,
      "num_tokens": 83982456.0,
      "step": 38335
    },
    {
      "entropy": 5.677844572067261,
      "epoch": 3.8050813815005955,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003615438569605394,
      "loss": 5.0642,
      "mean_token_accuracy": 0.18611662238836288,
      "num_tokens": 83993303.0,
      "step": 38340
    },
    {
      "entropy": 5.722970342636108,
      "epoch": 3.805577610162763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036151115024525777,
      "loss": 5.0209,
      "mean_token_accuracy": 0.18362841308116912,
      "num_tokens": 84004701.0,
      "step": 38345
    },
    {
      "entropy": 5.738949203491211,
      "epoch": 3.8060738388249304,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036147844138468616,
      "loss": 5.067,
      "mean_token_accuracy": 0.18781895637512208,
      "num_tokens": 84014779.0,
      "step": 38350
    },
    {
      "entropy": 5.692936038970947,
      "epoch": 3.806570067487098,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003614457303796356,
      "loss": 5.0751,
      "mean_token_accuracy": 0.18204795569181442,
      "num_tokens": 84025309.0,
      "step": 38355
    },
    {
      "entropy": 5.767366361618042,
      "epoch": 3.807066296149266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003614130172309171,
      "loss": 5.1565,
      "mean_token_accuracy": 0.18307781368494033,
      "num_tokens": 84037406.0,
      "step": 38360
    },
    {
      "entropy": 5.749300527572632,
      "epoch": 3.8075625248114333,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003613803019393419,
      "loss": 5.0438,
      "mean_token_accuracy": 0.18471099436283112,
      "num_tokens": 84048439.0,
      "step": 38365
    },
    {
      "entropy": 5.682438087463379,
      "epoch": 3.808058753473601,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036134758450572153,
      "loss": 5.0278,
      "mean_token_accuracy": 0.1856648236513138,
      "num_tokens": 84059681.0,
      "step": 38370
    },
    {
      "entropy": 5.753436517715454,
      "epoch": 3.8085549821357683,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000361314864930867,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1854643553495407,
      "num_tokens": 84069945.0,
      "step": 38375
    },
    {
      "entropy": 5.826923036575318,
      "epoch": 3.8090512107979357,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003612821432155898,
      "loss": 5.1437,
      "mean_token_accuracy": 0.17586909383535385,
      "num_tokens": 84082440.0,
      "step": 38380
    },
    {
      "entropy": 5.715146112442016,
      "epoch": 3.809547439460103,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003612494193607015,
      "loss": 4.9369,
      "mean_token_accuracy": 0.19857465773820876,
      "num_tokens": 84093431.0,
      "step": 38385
    },
    {
      "entropy": 5.7042278289794925,
      "epoch": 3.8100436681222707,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003612166933670134,
      "loss": 5.064,
      "mean_token_accuracy": 0.18410520851612092,
      "num_tokens": 84105904.0,
      "step": 38390
    },
    {
      "entropy": 5.706853866577148,
      "epoch": 3.810539896784438,
      "grad_norm": 1.125,
      "learning_rate": 0.000361183965235337,
      "loss": 5.074,
      "mean_token_accuracy": 0.181926628947258,
      "num_tokens": 84117967.0,
      "step": 38395
    },
    {
      "entropy": 5.700658321380615,
      "epoch": 3.8110361254466056,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003611512349664841,
      "loss": 4.9239,
      "mean_token_accuracy": 0.1899405911564827,
      "num_tokens": 84129287.0,
      "step": 38400
    },
    {
      "entropy": 5.764480924606323,
      "epoch": 3.811532354108773,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003611185025612661,
      "loss": 5.0451,
      "mean_token_accuracy": 0.18505332320928575,
      "num_tokens": 84140291.0,
      "step": 38405
    },
    {
      "entropy": 5.769438934326172,
      "epoch": 3.8120285827709406,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003610857680204949,
      "loss": 5.0422,
      "mean_token_accuracy": 0.1887755051255226,
      "num_tokens": 84150955.0,
      "step": 38410
    },
    {
      "entropy": 5.64604172706604,
      "epoch": 3.8125248114331085,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00036105303134498207,
      "loss": 4.9283,
      "mean_token_accuracy": 0.19350748062133788,
      "num_tokens": 84161665.0,
      "step": 38415
    },
    {
      "entropy": 5.6261663913726805,
      "epoch": 3.813021040095276,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00036102029253553945,
      "loss": 5.0183,
      "mean_token_accuracy": 0.18962059020996094,
      "num_tokens": 84172225.0,
      "step": 38420
    },
    {
      "entropy": 5.67357907295227,
      "epoch": 3.8135172687574435,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003609875515929791,
      "loss": 4.9863,
      "mean_token_accuracy": 0.18812275826931,
      "num_tokens": 84181877.0,
      "step": 38425
    },
    {
      "entropy": 5.656916618347168,
      "epoch": 3.814013497419611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003609548085181126,
      "loss": 4.9565,
      "mean_token_accuracy": 0.19362206608057023,
      "num_tokens": 84193342.0,
      "step": 38430
    },
    {
      "entropy": 5.644864702224732,
      "epoch": 3.8145097260817784,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003609220633117521,
      "loss": 4.9602,
      "mean_token_accuracy": 0.1894407331943512,
      "num_tokens": 84204719.0,
      "step": 38435
    },
    {
      "entropy": 5.6556174755096436,
      "epoch": 3.815005954743946,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003608893159747097,
      "loss": 4.9314,
      "mean_token_accuracy": 0.19347410202026366,
      "num_tokens": 84214931.0,
      "step": 38440
    },
    {
      "entropy": 5.646397542953491,
      "epoch": 3.815502183406114,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036085656650779716,
      "loss": 4.9797,
      "mean_token_accuracy": 0.18978420346975328,
      "num_tokens": 84227429.0,
      "step": 38445
    },
    {
      "entropy": 5.665806770324707,
      "epoch": 3.8159984120682813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00036082381491182685,
      "loss": 4.9768,
      "mean_token_accuracy": 0.1881168380379677,
      "num_tokens": 84236917.0,
      "step": 38450
    },
    {
      "entropy": 5.671115684509277,
      "epoch": 3.8164946407304488,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003607910611876109,
      "loss": 4.9934,
      "mean_token_accuracy": 0.19001419395208358,
      "num_tokens": 84247964.0,
      "step": 38455
    },
    {
      "entropy": 5.710459804534912,
      "epoch": 3.8169908693926162,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003607583053359615,
      "loss": 5.075,
      "mean_token_accuracy": 0.18318301141262056,
      "num_tokens": 84259559.0,
      "step": 38460
    },
    {
      "entropy": 5.731273651123047,
      "epoch": 3.8174870980547837,
      "grad_norm": 1.125,
      "learning_rate": 0.0003607255473576908,
      "loss": 4.9393,
      "mean_token_accuracy": 0.19668716937303543,
      "num_tokens": 84270849.0,
      "step": 38465
    },
    {
      "entropy": 5.714383554458618,
      "epoch": 3.817983326716951,
      "grad_norm": 1.203125,
      "learning_rate": 0.00036069278725361144,
      "loss": 4.9544,
      "mean_token_accuracy": 0.19328290820121766,
      "num_tokens": 84281977.0,
      "step": 38470
    },
    {
      "entropy": 5.664256238937378,
      "epoch": 3.8184795553791187,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036066002502453556,
      "loss": 4.9898,
      "mean_token_accuracy": 0.18632269650697708,
      "num_tokens": 84294289.0,
      "step": 38475
    },
    {
      "entropy": 5.66128511428833,
      "epoch": 3.818975784041286,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003606272606712755,
      "loss": 4.9825,
      "mean_token_accuracy": 0.19364003241062164,
      "num_tokens": 84304114.0,
      "step": 38480
    },
    {
      "entropy": 5.600553512573242,
      "epoch": 3.8194720127034536,
      "grad_norm": 1.265625,
      "learning_rate": 0.00036059449419464397,
      "loss": 4.917,
      "mean_token_accuracy": 0.18918263614177705,
      "num_tokens": 84314467.0,
      "step": 38485
    },
    {
      "entropy": 5.703945112228394,
      "epoch": 3.819968241365621,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003605617255954535,
      "loss": 5.0922,
      "mean_token_accuracy": 0.1828413039445877,
      "num_tokens": 84325565.0,
      "step": 38490
    },
    {
      "entropy": 5.650237274169922,
      "epoch": 3.8204644700277886,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003605289548745164,
      "loss": 4.9668,
      "mean_token_accuracy": 0.18675912767648697,
      "num_tokens": 84336578.0,
      "step": 38495
    },
    {
      "entropy": 5.694436073303223,
      "epoch": 3.8209606986899565,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003604961820326457,
      "loss": 4.9796,
      "mean_token_accuracy": 0.18808730095624923,
      "num_tokens": 84347447.0,
      "step": 38500
    },
    {
      "entropy": 5.615996694564819,
      "epoch": 3.821456927352124,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036046340707065387,
      "loss": 4.958,
      "mean_token_accuracy": 0.19276791363954543,
      "num_tokens": 84359527.0,
      "step": 38505
    },
    {
      "entropy": 5.680104303359985,
      "epoch": 3.8219531560142914,
      "grad_norm": 1.25,
      "learning_rate": 0.0003604306299893537,
      "loss": 4.9938,
      "mean_token_accuracy": 0.18402765542268754,
      "num_tokens": 84369653.0,
      "step": 38510
    },
    {
      "entropy": 5.627724885940552,
      "epoch": 3.822449384676459,
      "grad_norm": 1.21875,
      "learning_rate": 0.00036039785078955793,
      "loss": 4.899,
      "mean_token_accuracy": 0.19652344137430192,
      "num_tokens": 84378881.0,
      "step": 38515
    },
    {
      "entropy": 5.725387096405029,
      "epoch": 3.8229456133386264,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003603650694720796,
      "loss": 5.0843,
      "mean_token_accuracy": 0.18688304871320724,
      "num_tokens": 84390382.0,
      "step": 38520
    },
    {
      "entropy": 5.695966196060181,
      "epoch": 3.823441842000794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00036033228603773136,
      "loss": 4.9793,
      "mean_token_accuracy": 0.19224643111228942,
      "num_tokens": 84401032.0,
      "step": 38525
    },
    {
      "entropy": 5.7271346092224125,
      "epoch": 3.8239380706629618,
      "grad_norm": 1.15625,
      "learning_rate": 0.00036029950048732635,
      "loss": 5.1337,
      "mean_token_accuracy": 0.178554168343544,
      "num_tokens": 84411934.0,
      "step": 38530
    },
    {
      "entropy": 5.689135217666626,
      "epoch": 3.8244342993251292,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036026671282167753,
      "loss": 4.9727,
      "mean_token_accuracy": 0.19025251269340515,
      "num_tokens": 84423291.0,
      "step": 38535
    },
    {
      "entropy": 5.66993522644043,
      "epoch": 3.8249305279872967,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000360233923041598,
      "loss": 5.0459,
      "mean_token_accuracy": 0.18194108158349992,
      "num_tokens": 84436451.0,
      "step": 38540
    },
    {
      "entropy": 5.70587911605835,
      "epoch": 3.825426756649464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00036020113114790085,
      "loss": 4.9877,
      "mean_token_accuracy": 0.19264593571424485,
      "num_tokens": 84447933.0,
      "step": 38545
    },
    {
      "entropy": 5.7077075958251955,
      "epoch": 3.8259229853116317,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003601683371413992,
      "loss": 4.9708,
      "mean_token_accuracy": 0.18468220978975297,
      "num_tokens": 84458135.0,
      "step": 38550
    },
    {
      "entropy": 5.7843334674835205,
      "epoch": 3.826419213973799,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00036013554102290636,
      "loss": 5.0424,
      "mean_token_accuracy": 0.18335156440734862,
      "num_tokens": 84468620.0,
      "step": 38555
    },
    {
      "entropy": 5.6732258796691895,
      "epoch": 3.8269154426359666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003601027427932355,
      "loss": 4.9135,
      "mean_token_accuracy": 0.19180147498846054,
      "num_tokens": 84479891.0,
      "step": 38560
    },
    {
      "entropy": 5.655559825897217,
      "epoch": 3.827411671298134,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003600699424532,
      "loss": 5.0144,
      "mean_token_accuracy": 0.1914273425936699,
      "num_tokens": 84490621.0,
      "step": 38565
    },
    {
      "entropy": 5.7090390682220455,
      "epoch": 3.8279078999603016,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00036003714000361326,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20029377192258835,
      "num_tokens": 84500940.0,
      "step": 38570
    },
    {
      "entropy": 5.661738872528076,
      "epoch": 3.828404128622469,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00036000433544528866,
      "loss": 5.0189,
      "mean_token_accuracy": 0.19223816692829132,
      "num_tokens": 84511891.0,
      "step": 38575
    },
    {
      "entropy": 5.695192432403564,
      "epoch": 3.8289003572846365,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003599715287790397,
      "loss": 5.0167,
      "mean_token_accuracy": 0.18927165120840073,
      "num_tokens": 84522546.0,
      "step": 38580
    },
    {
      "entropy": 5.7642906665802,
      "epoch": 3.8293965859468044,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035993872000567994,
      "loss": 5.0115,
      "mean_token_accuracy": 0.18701623678207396,
      "num_tokens": 84532281.0,
      "step": 38585
    },
    {
      "entropy": 5.718983507156372,
      "epoch": 3.829892814608972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035990590912602304,
      "loss": 4.9696,
      "mean_token_accuracy": 0.18823987990617752,
      "num_tokens": 84542477.0,
      "step": 38590
    },
    {
      "entropy": 5.711442661285401,
      "epoch": 3.8303890432711394,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035987309614088245,
      "loss": 4.9716,
      "mean_token_accuracy": 0.19119134843349456,
      "num_tokens": 84552780.0,
      "step": 38595
    },
    {
      "entropy": 5.750794506072998,
      "epoch": 3.830885271933307,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035984028105107195,
      "loss": 5.0464,
      "mean_token_accuracy": 0.1860672801733017,
      "num_tokens": 84564895.0,
      "step": 38600
    },
    {
      "entropy": 5.668489027023315,
      "epoch": 3.8313815005954743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035980746385740536,
      "loss": 4.9285,
      "mean_token_accuracy": 0.19312204122543336,
      "num_tokens": 84575680.0,
      "step": 38605
    },
    {
      "entropy": 5.630154228210449,
      "epoch": 3.831877729257642,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035977464456069636,
      "loss": 4.9632,
      "mean_token_accuracy": 0.1943509265780449,
      "num_tokens": 84586865.0,
      "step": 38610
    },
    {
      "entropy": 5.653676986694336,
      "epoch": 3.8323739579198093,
      "grad_norm": 1.296875,
      "learning_rate": 0.00035974182316175887,
      "loss": 4.9904,
      "mean_token_accuracy": 0.19321919679641725,
      "num_tokens": 84598285.0,
      "step": 38615
    },
    {
      "entropy": 5.7231245040893555,
      "epoch": 3.832870186581977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003597089996614068,
      "loss": 5.046,
      "mean_token_accuracy": 0.18474243581295013,
      "num_tokens": 84609481.0,
      "step": 38620
    },
    {
      "entropy": 5.806306076049805,
      "epoch": 3.8333664152441447,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035967617406045406,
      "loss": 5.1028,
      "mean_token_accuracy": 0.18346186429262162,
      "num_tokens": 84619719.0,
      "step": 38625
    },
    {
      "entropy": 5.726664113998413,
      "epoch": 3.833862643906312,
      "grad_norm": 1.34375,
      "learning_rate": 0.00035964334635971455,
      "loss": 4.979,
      "mean_token_accuracy": 0.17998279333114625,
      "num_tokens": 84630074.0,
      "step": 38630
    },
    {
      "entropy": 5.642797470092773,
      "epoch": 3.8343588725684796,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003596105165600025,
      "loss": 4.9381,
      "mean_token_accuracy": 0.19225842207670213,
      "num_tokens": 84640508.0,
      "step": 38635
    },
    {
      "entropy": 5.669848489761352,
      "epoch": 3.834855101230647,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00035957768466213203,
      "loss": 4.9578,
      "mean_token_accuracy": 0.19714051187038423,
      "num_tokens": 84651824.0,
      "step": 38640
    },
    {
      "entropy": 5.684962797164917,
      "epoch": 3.8353513298928146,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035954485066691717,
      "loss": 5.0431,
      "mean_token_accuracy": 0.19003735184669496,
      "num_tokens": 84662107.0,
      "step": 38645
    },
    {
      "entropy": 5.67089262008667,
      "epoch": 3.835847558554982,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003595120145751723,
      "loss": 4.9277,
      "mean_token_accuracy": 0.19505662918090821,
      "num_tokens": 84674192.0,
      "step": 38650
    },
    {
      "entropy": 5.676529121398926,
      "epoch": 3.8363437872171495,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035947917638771146,
      "loss": 4.9767,
      "mean_token_accuracy": 0.1928801566362381,
      "num_tokens": 84684294.0,
      "step": 38655
    },
    {
      "entropy": 5.6524711608886715,
      "epoch": 3.836840015879317,
      "grad_norm": 1.3125,
      "learning_rate": 0.00035944633610534904,
      "loss": 4.9848,
      "mean_token_accuracy": 0.1869989216327667,
      "num_tokens": 84695934.0,
      "step": 38660
    },
    {
      "entropy": 5.748883676528931,
      "epoch": 3.8373362445414845,
      "grad_norm": 1.125,
      "learning_rate": 0.00035941349372889956,
      "loss": 5.0224,
      "mean_token_accuracy": 0.18692255020141602,
      "num_tokens": 84707549.0,
      "step": 38665
    },
    {
      "entropy": 5.683405637741089,
      "epoch": 3.8378324732036524,
      "grad_norm": 1.390625,
      "learning_rate": 0.00035938064925917726,
      "loss": 4.8909,
      "mean_token_accuracy": 0.20334823578596115,
      "num_tokens": 84717864.0,
      "step": 38670
    },
    {
      "entropy": 5.6314881324768065,
      "epoch": 3.83832870186582,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003593478026969967,
      "loss": 4.945,
      "mean_token_accuracy": 0.18515675365924836,
      "num_tokens": 84728328.0,
      "step": 38675
    },
    {
      "entropy": 5.753503847122192,
      "epoch": 3.8388249305279873,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035931495404317243,
      "loss": 5.0831,
      "mean_token_accuracy": 0.17918664664030076,
      "num_tokens": 84739037.0,
      "step": 38680
    },
    {
      "entropy": 5.644786071777344,
      "epoch": 3.839321159190155,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000359282103298519,
      "loss": 4.9945,
      "mean_token_accuracy": 0.19317708909511566,
      "num_tokens": 84750356.0,
      "step": 38685
    },
    {
      "entropy": 5.684923315048218,
      "epoch": 3.8398173878523223,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035924925046385085,
      "loss": 4.941,
      "mean_token_accuracy": 0.1899753764271736,
      "num_tokens": 84759622.0,
      "step": 38690
    },
    {
      "entropy": 5.644970417022705,
      "epoch": 3.8403136165144898,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00035921639553998297,
      "loss": 4.9871,
      "mean_token_accuracy": 0.18864582628011703,
      "num_tokens": 84771602.0,
      "step": 38695
    },
    {
      "entropy": 5.702092027664184,
      "epoch": 3.8408098451766572,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035918353852772996,
      "loss": 4.9939,
      "mean_token_accuracy": 0.19025272876024246,
      "num_tokens": 84781730.0,
      "step": 38700
    },
    {
      "entropy": 5.705162811279297,
      "epoch": 3.841306073838825,
      "grad_norm": 1.125,
      "learning_rate": 0.0003591506794279066,
      "loss": 4.9552,
      "mean_token_accuracy": 0.19505424946546554,
      "num_tokens": 84793846.0,
      "step": 38705
    },
    {
      "entropy": 5.7501842975616455,
      "epoch": 3.8418023025009926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003591178182413276,
      "loss": 5.1162,
      "mean_token_accuracy": 0.1862150326371193,
      "num_tokens": 84804704.0,
      "step": 38710
    },
    {
      "entropy": 5.675026750564575,
      "epoch": 3.84229853116316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035908495496880804,
      "loss": 4.9355,
      "mean_token_accuracy": 0.1934064283967018,
      "num_tokens": 84816052.0,
      "step": 38715
    },
    {
      "entropy": 5.711042976379394,
      "epoch": 3.8427947598253276,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003590520896111628,
      "loss": 5.0619,
      "mean_token_accuracy": 0.18905302435159682,
      "num_tokens": 84827412.0,
      "step": 38720
    },
    {
      "entropy": 5.739316082000732,
      "epoch": 3.843290988487495,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035901922216920683,
      "loss": 5.0393,
      "mean_token_accuracy": 0.18742400258779526,
      "num_tokens": 84840003.0,
      "step": 38725
    },
    {
      "entropy": 5.69931583404541,
      "epoch": 3.8437872171496625,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003589863526437552,
      "loss": 4.9388,
      "mean_token_accuracy": 0.19522304385900496,
      "num_tokens": 84850385.0,
      "step": 38730
    },
    {
      "entropy": 5.689484310150147,
      "epoch": 3.84428344581183,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003589534810356231,
      "loss": 4.9377,
      "mean_token_accuracy": 0.18918393105268477,
      "num_tokens": 84860628.0,
      "step": 38735
    },
    {
      "entropy": 5.628170108795166,
      "epoch": 3.8447796744739975,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00035892060734562544,
      "loss": 4.9136,
      "mean_token_accuracy": 0.19919871985912324,
      "num_tokens": 84870654.0,
      "step": 38740
    },
    {
      "entropy": 5.632048034667969,
      "epoch": 3.845275903136165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00035888773157457756,
      "loss": 4.9536,
      "mean_token_accuracy": 0.18737971037626266,
      "num_tokens": 84882928.0,
      "step": 38745
    },
    {
      "entropy": 5.740181541442871,
      "epoch": 3.8457721317983324,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003588548537232947,
      "loss": 5.0288,
      "mean_token_accuracy": 0.19292248636484147,
      "num_tokens": 84893626.0,
      "step": 38750
    },
    {
      "entropy": 5.678687286376953,
      "epoch": 3.8462683604605004,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035882197379259224,
      "loss": 5.013,
      "mean_token_accuracy": 0.18576325327157975,
      "num_tokens": 84904442.0,
      "step": 38755
    },
    {
      "entropy": 5.720851182937622,
      "epoch": 3.846764589122668,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003587890917832854,
      "loss": 5.0266,
      "mean_token_accuracy": 0.19180604219436645,
      "num_tokens": 84914444.0,
      "step": 38760
    },
    {
      "entropy": 5.753353786468506,
      "epoch": 3.8472608177848353,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003587562076961897,
      "loss": 5.0501,
      "mean_token_accuracy": 0.18523770272731782,
      "num_tokens": 84925274.0,
      "step": 38765
    },
    {
      "entropy": 5.695265531539917,
      "epoch": 3.847757046447003,
      "grad_norm": 1.328125,
      "learning_rate": 0.00035872332153212046,
      "loss": 4.9515,
      "mean_token_accuracy": 0.191641141474247,
      "num_tokens": 84936236.0,
      "step": 38770
    },
    {
      "entropy": 5.744184827804565,
      "epoch": 3.8482532751091703,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035869043329189314,
      "loss": 5.0473,
      "mean_token_accuracy": 0.18473320007324218,
      "num_tokens": 84947895.0,
      "step": 38775
    },
    {
      "entropy": 5.671549224853516,
      "epoch": 3.8487495037713377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003586575429763236,
      "loss": 4.9693,
      "mean_token_accuracy": 0.19224686175584793,
      "num_tokens": 84958584.0,
      "step": 38780
    },
    {
      "entropy": 5.708817672729492,
      "epoch": 3.849245732433505,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003586246505862272,
      "loss": 5.0401,
      "mean_token_accuracy": 0.1932062342762947,
      "num_tokens": 84970135.0,
      "step": 38785
    },
    {
      "entropy": 5.786244201660156,
      "epoch": 3.849741961095673,
      "grad_norm": 1.359375,
      "learning_rate": 0.00035859175612241957,
      "loss": 5.0301,
      "mean_token_accuracy": 0.18722635358572007,
      "num_tokens": 84982086.0,
      "step": 38790
    },
    {
      "entropy": 5.696982526779175,
      "epoch": 3.8502381897578406,
      "grad_norm": 1.25,
      "learning_rate": 0.00035855885958571666,
      "loss": 4.9862,
      "mean_token_accuracy": 0.1923239678144455,
      "num_tokens": 84994399.0,
      "step": 38795
    },
    {
      "entropy": 5.74739031791687,
      "epoch": 3.850734418420008,
      "grad_norm": 1.3125,
      "learning_rate": 0.000358525960976934,
      "loss": 5.0925,
      "mean_token_accuracy": 0.1845914363861084,
      "num_tokens": 85004248.0,
      "step": 38800
    },
    {
      "entropy": 5.749304485321045,
      "epoch": 3.8512306470821756,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003584930602968875,
      "loss": 5.0974,
      "mean_token_accuracy": 0.18413357436656952,
      "num_tokens": 85015306.0,
      "step": 38805
    },
    {
      "entropy": 5.676632499694824,
      "epoch": 3.851726875744343,
      "grad_norm": 1.21875,
      "learning_rate": 0.00035846015754639304,
      "loss": 4.9021,
      "mean_token_accuracy": 0.1928565412759781,
      "num_tokens": 85025566.0,
      "step": 38810
    },
    {
      "entropy": 5.70184178352356,
      "epoch": 3.8522231044065105,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003584272527262665,
      "loss": 5.0285,
      "mean_token_accuracy": 0.19358834475278855,
      "num_tokens": 85036692.0,
      "step": 38815
    },
    {
      "entropy": 5.725148487091064,
      "epoch": 3.852719333068678,
      "grad_norm": 1.4375,
      "learning_rate": 0.00035839434583732384,
      "loss": 5.0597,
      "mean_token_accuracy": 0.18432803452014923,
      "num_tokens": 85046712.0,
      "step": 38820
    },
    {
      "entropy": 5.728960037231445,
      "epoch": 3.8532155617308455,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035836143688038107,
      "loss": 5.0471,
      "mean_token_accuracy": 0.18732592910528184,
      "num_tokens": 85056122.0,
      "step": 38825
    },
    {
      "entropy": 5.708226251602173,
      "epoch": 3.853711790393013,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035832852585625436,
      "loss": 5.0386,
      "mean_token_accuracy": 0.18439806401729583,
      "num_tokens": 85068201.0,
      "step": 38830
    },
    {
      "entropy": 5.734686422348022,
      "epoch": 3.8542080190551804,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003582956127657598,
      "loss": 5.0295,
      "mean_token_accuracy": 0.19539177566766738,
      "num_tokens": 85079248.0,
      "step": 38835
    },
    {
      "entropy": 5.639464855194092,
      "epoch": 3.8547042477173483,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003582626976097134,
      "loss": 4.9047,
      "mean_token_accuracy": 0.19349705874919892,
      "num_tokens": 85091602.0,
      "step": 38840
    },
    {
      "entropy": 5.664415884017944,
      "epoch": 3.855200476379516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003582297803889316,
      "loss": 4.919,
      "mean_token_accuracy": 0.19270336329936982,
      "num_tokens": 85101540.0,
      "step": 38845
    },
    {
      "entropy": 5.645927762985229,
      "epoch": 3.8556967050416833,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003581968611042306,
      "loss": 4.9825,
      "mean_token_accuracy": 0.19130364060401917,
      "num_tokens": 85112856.0,
      "step": 38850
    },
    {
      "entropy": 5.666553688049317,
      "epoch": 3.8561929337038507,
      "grad_norm": 1.125,
      "learning_rate": 0.00035816393975642675,
      "loss": 4.9879,
      "mean_token_accuracy": 0.18555186986923217,
      "num_tokens": 85123269.0,
      "step": 38855
    },
    {
      "entropy": 5.614957618713379,
      "epoch": 3.856689162366018,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00035813101634633626,
      "loss": 4.8807,
      "mean_token_accuracy": 0.20119670331478118,
      "num_tokens": 85134175.0,
      "step": 38860
    },
    {
      "entropy": 5.710721349716186,
      "epoch": 3.8571853910281857,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00035809809087477586,
      "loss": 5.0036,
      "mean_token_accuracy": 0.19484832435846328,
      "num_tokens": 85146808.0,
      "step": 38865
    },
    {
      "entropy": 5.686022472381592,
      "epoch": 3.857681619690353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003580651633425617,
      "loss": 4.9648,
      "mean_token_accuracy": 0.18568150103092193,
      "num_tokens": 85158651.0,
      "step": 38870
    },
    {
      "entropy": 5.765532922744751,
      "epoch": 3.858177848352521,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003580322337505106,
      "loss": 5.0291,
      "mean_token_accuracy": 0.18581603318452836,
      "num_tokens": 85169152.0,
      "step": 38875
    },
    {
      "entropy": 5.662467622756958,
      "epoch": 3.8586740770146886,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00035799930209943906,
      "loss": 5.0071,
      "mean_token_accuracy": 0.18097514808177947,
      "num_tokens": 85180516.0,
      "step": 38880
    },
    {
      "entropy": 5.668847942352295,
      "epoch": 3.859170305676856,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003579663683901636,
      "loss": 4.9814,
      "mean_token_accuracy": 0.1881064772605896,
      "num_tokens": 85192109.0,
      "step": 38885
    },
    {
      "entropy": 5.694802618026733,
      "epoch": 3.8596665343390235,
      "grad_norm": 1.125,
      "learning_rate": 0.00035793343262350094,
      "loss": 4.9997,
      "mean_token_accuracy": 0.18725039064884186,
      "num_tokens": 85203005.0,
      "step": 38890
    },
    {
      "entropy": 5.697049951553344,
      "epoch": 3.860162763001191,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003579004948002679,
      "loss": 4.9908,
      "mean_token_accuracy": 0.19509754925966263,
      "num_tokens": 85214374.0,
      "step": 38895
    },
    {
      "entropy": 5.605131149291992,
      "epoch": 3.8606589916633585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00035786755492128134,
      "loss": 4.9009,
      "mean_token_accuracy": 0.19983323365449907,
      "num_tokens": 85225056.0,
      "step": 38900
    },
    {
      "entropy": 5.727624225616455,
      "epoch": 3.861155220325526,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0003578346129873578,
      "loss": 5.1243,
      "mean_token_accuracy": 0.1813351407647133,
      "num_tokens": 85236154.0,
      "step": 38905
    },
    {
      "entropy": 5.684330224990845,
      "epoch": 3.8616514489876934,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00035780166899931437,
      "loss": 4.9664,
      "mean_token_accuracy": 0.18837082684040068,
      "num_tokens": 85246757.0,
      "step": 38910
    },
    {
      "entropy": 5.6761997699737545,
      "epoch": 3.862147677649861,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003577687229579681,
      "loss": 4.9289,
      "mean_token_accuracy": 0.19295624792575836,
      "num_tokens": 85258066.0,
      "step": 38915
    },
    {
      "entropy": 5.70129075050354,
      "epoch": 3.8626439063120284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003577357748641357,
      "loss": 5.0035,
      "mean_token_accuracy": 0.18793220371007918,
      "num_tokens": 85268204.0,
      "step": 38920
    },
    {
      "entropy": 5.608046579360962,
      "epoch": 3.863140134974196,
      "grad_norm": 1.125,
      "learning_rate": 0.0003577028247186344,
      "loss": 4.9176,
      "mean_token_accuracy": 0.19651535302400588,
      "num_tokens": 85278808.0,
      "step": 38925
    },
    {
      "entropy": 5.688019466400147,
      "epoch": 3.8636363636363638,
      "grad_norm": 1.1875,
      "learning_rate": 0.00035766987252228133,
      "loss": 4.9936,
      "mean_token_accuracy": 0.1865213543176651,
      "num_tokens": 85290082.0,
      "step": 38930
    },
    {
      "entropy": 5.649104070663452,
      "epoch": 3.8641325922985312,
      "grad_norm": 1.25,
      "learning_rate": 0.0003576369182758935,
      "loss": 4.9832,
      "mean_token_accuracy": 0.1884508565068245,
      "num_tokens": 85300313.0,
      "step": 38935
    },
    {
      "entropy": 5.648344802856445,
      "epoch": 3.8646288209606987,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003576039619802882,
      "loss": 4.8809,
      "mean_token_accuracy": 0.19807495474815368,
      "num_tokens": 85310967.0,
      "step": 38940
    },
    {
      "entropy": 5.701713705062867,
      "epoch": 3.865125049622866,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0003575710036362826,
      "loss": 4.9632,
      "mean_token_accuracy": 0.19082336723804474,
      "num_tokens": 85321835.0,
      "step": 38945
    },
    {
      "entropy": 5.695753526687622,
      "epoch": 3.8656212782850337,
      "grad_norm": 1.265625,
      "learning_rate": 0.0003575380432446941,
      "loss": 5.0017,
      "mean_token_accuracy": 0.18775007575750352,
      "num_tokens": 85334193.0,
      "step": 38950
    },
    {
      "entropy": 5.717475366592407,
      "epoch": 3.866117506947201,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003575050808063399,
      "loss": 5.0082,
      "mean_token_accuracy": 0.19062376618385315,
      "num_tokens": 85344864.0,
      "step": 38955
    },
    {
      "entropy": 5.6810815811157225,
      "epoch": 3.866613735609369,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00035747211632203757,
      "loss": 4.9358,
      "mean_token_accuracy": 0.19496876150369644,
      "num_tokens": 85356183.0,
      "step": 38960
    },
    {
      "entropy": 5.698201036453247,
      "epoch": 3.8671099642715365,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003574391497926044,
      "loss": 4.9105,
      "mean_token_accuracy": 0.1989549934864044,
      "num_tokens": 85367157.0,
      "step": 38965
    },
    {
      "entropy": 5.732802963256836,
      "epoch": 3.867606192933704,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000357406181218858,
      "loss": 5.0487,
      "mean_token_accuracy": 0.1918008252978325,
      "num_tokens": 85378995.0,
      "step": 38970
    },
    {
      "entropy": 5.658675622940064,
      "epoch": 3.8681024215958715,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003573732106016158,
      "loss": 4.9848,
      "mean_token_accuracy": 0.19440695494413376,
      "num_tokens": 85389949.0,
      "step": 38975
    },
    {
      "entropy": 5.676565933227539,
      "epoch": 3.868598650258039,
      "grad_norm": 1.25,
      "learning_rate": 0.0003573402379416955,
      "loss": 5.0158,
      "mean_token_accuracy": 0.18288587629795075,
      "num_tokens": 85401261.0,
      "step": 38980
    },
    {
      "entropy": 5.7748706340789795,
      "epoch": 3.8690948789202064,
      "grad_norm": 1.171875,
      "learning_rate": 0.00035730726323991474,
      "loss": 5.0452,
      "mean_token_accuracy": 0.1998060092329979,
      "num_tokens": 85413446.0,
      "step": 38985
    },
    {
      "entropy": 5.696991348266602,
      "epoch": 3.869591107582374,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003572742864970913,
      "loss": 4.9805,
      "mean_token_accuracy": 0.18888451159000397,
      "num_tokens": 85423819.0,
      "step": 38990
    },
    {
      "entropy": 5.704282903671265,
      "epoch": 3.8700873362445414,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003572413077140427,
      "loss": 4.9649,
      "mean_token_accuracy": 0.1944171130657196,
      "num_tokens": 85435129.0,
      "step": 38995
    },
    {
      "entropy": 5.704431343078613,
      "epoch": 3.870583564906709,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000357208326891587,
      "loss": 4.9446,
      "mean_token_accuracy": 0.19752982556819915,
      "num_tokens": 85445845.0,
      "step": 39000
    },
    {
      "epoch": 3.870583564906709,
      "eval_entropy": 5.433925486370004,
      "eval_loss": 5.11483907699585,
      "eval_mean_token_accuracy": 0.18974152148959472,
      "eval_num_tokens": 85445845.0,
      "eval_runtime": 20.2732,
      "eval_samples_per_second": 1660.763,
      "eval_steps_per_second": 207.614,
      "step": 39000
    }
  ],
  "logging_steps": 5,
  "max_steps": 100750,
  "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": 2.127781039767552e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}