{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.885747259088286,
  "eval_steps": 3000,
  "global_step": 51000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.691951370239257,
      "epoch": 0.0005770340450086555,
      "grad_norm": 12.5,
      "learning_rate": 2e-06,
      "loss": 10.8557,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11955.0,
      "step": 5
    },
    {
      "entropy": 10.691877937316894,
      "epoch": 0.001154068090017311,
      "grad_norm": 12.8125,
      "learning_rate": 4.5e-06,
      "loss": 10.7945,
      "mean_token_accuracy": 0.00010020040208473802,
      "num_tokens": 24473.0,
      "step": 10
    },
    {
      "entropy": 10.690855121612548,
      "epoch": 0.0017311021350259665,
      "grad_norm": 10.5625,
      "learning_rate": 7e-06,
      "loss": 10.5617,
      "mean_token_accuracy": 0.02503108715172857,
      "num_tokens": 36716.0,
      "step": 15
    },
    {
      "entropy": 10.675894641876221,
      "epoch": 0.002308136180034622,
      "grad_norm": 6.40625,
      "learning_rate": 9.5e-06,
      "loss": 10.2237,
      "mean_token_accuracy": 0.045600682869553565,
      "num_tokens": 49747.0,
      "step": 20
    },
    {
      "entropy": 10.583820247650147,
      "epoch": 0.0028851702250432777,
      "grad_norm": 4.53125,
      "learning_rate": 1.2e-05,
      "loss": 9.9687,
      "mean_token_accuracy": 0.04857061766088009,
      "num_tokens": 61991.0,
      "step": 25
    },
    {
      "entropy": 10.477548027038575,
      "epoch": 0.003462204270051933,
      "grad_norm": 2.734375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 9.8415,
      "mean_token_accuracy": 0.05598664358258247,
      "num_tokens": 74918.0,
      "step": 30
    },
    {
      "entropy": 10.462637138366699,
      "epoch": 0.004039238315060588,
      "grad_norm": 2.8125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 9.8059,
      "mean_token_accuracy": 0.054788753390312195,
      "num_tokens": 87833.0,
      "step": 35
    },
    {
      "entropy": 10.446255970001221,
      "epoch": 0.004616272360069244,
      "grad_norm": 2.546875,
      "learning_rate": 1.95e-05,
      "loss": 9.7253,
      "mean_token_accuracy": 0.05731462575495243,
      "num_tokens": 101031.0,
      "step": 40
    },
    {
      "entropy": 10.412657737731934,
      "epoch": 0.0051933064050779,
      "grad_norm": 2.6875,
      "learning_rate": 2.2e-05,
      "loss": 9.6744,
      "mean_token_accuracy": 0.06321047246456146,
      "num_tokens": 114085.0,
      "step": 45
    },
    {
      "entropy": 10.344977378845215,
      "epoch": 0.005770340450086555,
      "grad_norm": 2.296875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.682,
      "mean_token_accuracy": 0.05514660738408565,
      "num_tokens": 126190.0,
      "step": 50
    },
    {
      "entropy": 10.360903549194337,
      "epoch": 0.006347374495095211,
      "grad_norm": 2.546875,
      "learning_rate": 2.7e-05,
      "loss": 9.5899,
      "mean_token_accuracy": 0.07075095213949681,
      "num_tokens": 138933.0,
      "step": 55
    },
    {
      "entropy": 10.256150436401366,
      "epoch": 0.006924408540103866,
      "grad_norm": 2.109375,
      "learning_rate": 2.95e-05,
      "loss": 9.4931,
      "mean_token_accuracy": 0.07884349599480629,
      "num_tokens": 150542.0,
      "step": 60
    },
    {
      "entropy": 10.159530735015869,
      "epoch": 0.0075014425851125215,
      "grad_norm": 2.234375,
      "learning_rate": 3.2e-05,
      "loss": 9.4791,
      "mean_token_accuracy": 0.07697005346417427,
      "num_tokens": 163507.0,
      "step": 65
    },
    {
      "entropy": 10.226024723052978,
      "epoch": 0.008078476630121177,
      "grad_norm": 2.640625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.4185,
      "mean_token_accuracy": 0.08218079656362534,
      "num_tokens": 176668.0,
      "step": 70
    },
    {
      "entropy": 10.084021472930909,
      "epoch": 0.008655510675129832,
      "grad_norm": 2.078125,
      "learning_rate": 3.7e-05,
      "loss": 9.3547,
      "mean_token_accuracy": 0.08463524356484413,
      "num_tokens": 191413.0,
      "step": 75
    },
    {
      "entropy": 10.201306247711182,
      "epoch": 0.009232544720138488,
      "grad_norm": 2.328125,
      "learning_rate": 3.95e-05,
      "loss": 9.2867,
      "mean_token_accuracy": 0.08322363793849945,
      "num_tokens": 204176.0,
      "step": 80
    },
    {
      "entropy": 10.13297176361084,
      "epoch": 0.009809578765147143,
      "grad_norm": 1.8359375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.2251,
      "mean_token_accuracy": 0.08606483414769173,
      "num_tokens": 216851.0,
      "step": 85
    },
    {
      "entropy": 10.010610389709473,
      "epoch": 0.0103866128101558,
      "grad_norm": 2.265625,
      "learning_rate": 4.45e-05,
      "loss": 9.1974,
      "mean_token_accuracy": 0.08009482100605965,
      "num_tokens": 229621.0,
      "step": 90
    },
    {
      "entropy": 10.165260124206544,
      "epoch": 0.010963646855164455,
      "grad_norm": 2.6875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 9.151,
      "mean_token_accuracy": 0.08369554057717324,
      "num_tokens": 242074.0,
      "step": 95
    },
    {
      "entropy": 9.886237907409669,
      "epoch": 0.01154068090017311,
      "grad_norm": 1.6328125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 9.0049,
      "mean_token_accuracy": 0.08854745402932167,
      "num_tokens": 254094.0,
      "step": 100
    },
    {
      "entropy": 9.991595554351807,
      "epoch": 0.012117714945181766,
      "grad_norm": 1.921875,
      "learning_rate": 5.2e-05,
      "loss": 9.0197,
      "mean_token_accuracy": 0.08869979456067086,
      "num_tokens": 267560.0,
      "step": 105
    },
    {
      "entropy": 9.877109718322753,
      "epoch": 0.012694748990190421,
      "grad_norm": 2.09375,
      "learning_rate": 5.45e-05,
      "loss": 9.0088,
      "mean_token_accuracy": 0.08046346381306649,
      "num_tokens": 280717.0,
      "step": 110
    },
    {
      "entropy": 9.986603260040283,
      "epoch": 0.013271783035199077,
      "grad_norm": 1.84375,
      "learning_rate": 5.7e-05,
      "loss": 8.956,
      "mean_token_accuracy": 0.08400088958442212,
      "num_tokens": 293856.0,
      "step": 115
    },
    {
      "entropy": 9.746489906311036,
      "epoch": 0.013848817080207732,
      "grad_norm": 1.609375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.7825,
      "mean_token_accuracy": 0.0846656046807766,
      "num_tokens": 305191.0,
      "step": 120
    },
    {
      "entropy": 9.743055725097657,
      "epoch": 0.014425851125216388,
      "grad_norm": 1.484375,
      "learning_rate": 6.2e-05,
      "loss": 8.8098,
      "mean_token_accuracy": 0.08250505477190018,
      "num_tokens": 318544.0,
      "step": 125
    },
    {
      "entropy": 9.622491931915283,
      "epoch": 0.015002885170225043,
      "grad_norm": 1.5859375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.631,
      "mean_token_accuracy": 0.09809689894318581,
      "num_tokens": 331445.0,
      "step": 130
    },
    {
      "entropy": 9.531793117523193,
      "epoch": 0.015579919215233698,
      "grad_norm": 1.53125,
      "learning_rate": 6.7e-05,
      "loss": 8.69,
      "mean_token_accuracy": 0.08484943211078644,
      "num_tokens": 344775.0,
      "step": 135
    },
    {
      "entropy": 9.43226900100708,
      "epoch": 0.016156953260242354,
      "grad_norm": 1.2421875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 8.5877,
      "mean_token_accuracy": 0.08672104477882385,
      "num_tokens": 357804.0,
      "step": 140
    },
    {
      "entropy": 9.345358848571777,
      "epoch": 0.01673398730525101,
      "grad_norm": 1.3828125,
      "learning_rate": 7.2e-05,
      "loss": 8.5103,
      "mean_token_accuracy": 0.08910816386342049,
      "num_tokens": 369826.0,
      "step": 145
    },
    {
      "entropy": 9.061049365997315,
      "epoch": 0.017311021350259664,
      "grad_norm": 1.734375,
      "learning_rate": 7.45e-05,
      "loss": 8.4734,
      "mean_token_accuracy": 0.0947182148694992,
      "num_tokens": 382768.0,
      "step": 150
    },
    {
      "entropy": 9.184530448913574,
      "epoch": 0.01788805539526832,
      "grad_norm": 1.4921875,
      "learning_rate": 7.7e-05,
      "loss": 8.477,
      "mean_token_accuracy": 0.09176877290010452,
      "num_tokens": 395910.0,
      "step": 155
    },
    {
      "entropy": 9.0551118850708,
      "epoch": 0.018465089440276975,
      "grad_norm": 1.15625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 8.4796,
      "mean_token_accuracy": 0.08274488113820552,
      "num_tokens": 408140.0,
      "step": 160
    },
    {
      "entropy": 8.977632331848145,
      "epoch": 0.019042123485285632,
      "grad_norm": 1.515625,
      "learning_rate": 8.2e-05,
      "loss": 8.3492,
      "mean_token_accuracy": 0.09365171417593957,
      "num_tokens": 420486.0,
      "step": 165
    },
    {
      "entropy": 8.88683271408081,
      "epoch": 0.019619157530294286,
      "grad_norm": 1.2265625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 8.3127,
      "mean_token_accuracy": 0.10231351554393768,
      "num_tokens": 433207.0,
      "step": 170
    },
    {
      "entropy": 8.718389701843261,
      "epoch": 0.020196191575302943,
      "grad_norm": 1.1015625,
      "learning_rate": 8.7e-05,
      "loss": 8.2991,
      "mean_token_accuracy": 0.09770049974322319,
      "num_tokens": 444880.0,
      "step": 175
    },
    {
      "entropy": 8.861179733276368,
      "epoch": 0.0207732256203116,
      "grad_norm": 1.59375,
      "learning_rate": 8.95e-05,
      "loss": 8.3278,
      "mean_token_accuracy": 0.09125575348734856,
      "num_tokens": 456650.0,
      "step": 180
    },
    {
      "entropy": 8.653267669677735,
      "epoch": 0.021350259665320254,
      "grad_norm": 1.2109375,
      "learning_rate": 9.2e-05,
      "loss": 8.3105,
      "mean_token_accuracy": 0.09449072554707527,
      "num_tokens": 469667.0,
      "step": 185
    },
    {
      "entropy": 8.639202690124511,
      "epoch": 0.02192729371032891,
      "grad_norm": 1.109375,
      "learning_rate": 9.45e-05,
      "loss": 8.2226,
      "mean_token_accuracy": 0.10481953471899033,
      "num_tokens": 481952.0,
      "step": 190
    },
    {
      "entropy": 8.621485424041747,
      "epoch": 0.022504327755337564,
      "grad_norm": 1.296875,
      "learning_rate": 9.7e-05,
      "loss": 8.2791,
      "mean_token_accuracy": 0.10009086430072785,
      "num_tokens": 494352.0,
      "step": 195
    },
    {
      "entropy": 8.494072914123535,
      "epoch": 0.02308136180034622,
      "grad_norm": 1.203125,
      "learning_rate": 9.95e-05,
      "loss": 8.19,
      "mean_token_accuracy": 0.10785712376236915,
      "num_tokens": 506937.0,
      "step": 200
    },
    {
      "entropy": 8.508214092254638,
      "epoch": 0.023658395845354875,
      "grad_norm": 1.203125,
      "learning_rate": 0.000102,
      "loss": 8.1846,
      "mean_token_accuracy": 0.10090541988611221,
      "num_tokens": 519552.0,
      "step": 205
    },
    {
      "entropy": 8.504057216644288,
      "epoch": 0.024235429890363532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00010449999999999999,
      "loss": 8.2334,
      "mean_token_accuracy": 0.09878813922405243,
      "num_tokens": 531359.0,
      "step": 210
    },
    {
      "entropy": 8.462166213989258,
      "epoch": 0.024812463935372186,
      "grad_norm": 1.4375,
      "learning_rate": 0.000107,
      "loss": 8.2473,
      "mean_token_accuracy": 0.1007633239030838,
      "num_tokens": 544197.0,
      "step": 215
    },
    {
      "entropy": 8.630421257019043,
      "epoch": 0.025389497980380843,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0001095,
      "loss": 8.2016,
      "mean_token_accuracy": 0.09566057249903678,
      "num_tokens": 556818.0,
      "step": 220
    },
    {
      "entropy": 8.372712326049804,
      "epoch": 0.025966532025389497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000112,
      "loss": 8.2354,
      "mean_token_accuracy": 0.10132482126355172,
      "num_tokens": 569363.0,
      "step": 225
    },
    {
      "entropy": 8.531051921844483,
      "epoch": 0.026543566070398154,
      "grad_norm": 1.21875,
      "learning_rate": 0.0001145,
      "loss": 8.2111,
      "mean_token_accuracy": 0.1007345326244831,
      "num_tokens": 581645.0,
      "step": 230
    },
    {
      "entropy": 8.464646530151366,
      "epoch": 0.027120600115406807,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00011700000000000001,
      "loss": 8.1687,
      "mean_token_accuracy": 0.09699883535504342,
      "num_tokens": 593501.0,
      "step": 235
    },
    {
      "entropy": 8.380043411254883,
      "epoch": 0.027697634160415464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00011949999999999999,
      "loss": 8.1632,
      "mean_token_accuracy": 0.10044692009687424,
      "num_tokens": 606328.0,
      "step": 240
    },
    {
      "entropy": 8.322643375396728,
      "epoch": 0.02827466820542412,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000122,
      "loss": 8.0529,
      "mean_token_accuracy": 0.11430942863225937,
      "num_tokens": 619003.0,
      "step": 245
    },
    {
      "entropy": 8.437573051452636,
      "epoch": 0.028851702250432775,
      "grad_norm": 1.484375,
      "learning_rate": 0.0001245,
      "loss": 8.1974,
      "mean_token_accuracy": 0.09745758473873138,
      "num_tokens": 632910.0,
      "step": 250
    },
    {
      "entropy": 8.345236682891846,
      "epoch": 0.029428736295441432,
      "grad_norm": 1.28125,
      "learning_rate": 0.000127,
      "loss": 8.1263,
      "mean_token_accuracy": 0.1067387342453003,
      "num_tokens": 645959.0,
      "step": 255
    },
    {
      "entropy": 8.347197151184082,
      "epoch": 0.030005770340450086,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001295,
      "loss": 8.1006,
      "mean_token_accuracy": 0.10821648016571998,
      "num_tokens": 658620.0,
      "step": 260
    },
    {
      "entropy": 8.356802654266357,
      "epoch": 0.030582804385458743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000132,
      "loss": 8.1138,
      "mean_token_accuracy": 0.1112379103899002,
      "num_tokens": 670804.0,
      "step": 265
    },
    {
      "entropy": 8.397929668426514,
      "epoch": 0.031159838430467397,
      "grad_norm": 1.25,
      "learning_rate": 0.00013450000000000002,
      "loss": 8.1186,
      "mean_token_accuracy": 0.1025885134935379,
      "num_tokens": 683656.0,
      "step": 270
    },
    {
      "entropy": 8.3319091796875,
      "epoch": 0.031736872475476054,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00013700000000000002,
      "loss": 8.093,
      "mean_token_accuracy": 0.10706457123160362,
      "num_tokens": 696629.0,
      "step": 275
    },
    {
      "entropy": 8.304152488708496,
      "epoch": 0.03231390652048471,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001395,
      "loss": 8.113,
      "mean_token_accuracy": 0.10793379172682763,
      "num_tokens": 709053.0,
      "step": 280
    },
    {
      "entropy": 8.2885817527771,
      "epoch": 0.03289094056549336,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00014199999999999998,
      "loss": 8.0615,
      "mean_token_accuracy": 0.10749010294675827,
      "num_tokens": 721947.0,
      "step": 285
    },
    {
      "entropy": 8.336502933502198,
      "epoch": 0.03346797461050202,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0001445,
      "loss": 8.0686,
      "mean_token_accuracy": 0.10405683368444443,
      "num_tokens": 734495.0,
      "step": 290
    },
    {
      "entropy": 8.229808235168457,
      "epoch": 0.034045008655510675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000147,
      "loss": 8.0824,
      "mean_token_accuracy": 0.1042616605758667,
      "num_tokens": 746965.0,
      "step": 295
    },
    {
      "entropy": 8.36244821548462,
      "epoch": 0.03462204270051933,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001495,
      "loss": 8.0907,
      "mean_token_accuracy": 0.10315949246287345,
      "num_tokens": 760105.0,
      "step": 300
    },
    {
      "entropy": 8.327516651153564,
      "epoch": 0.03519907674552799,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000152,
      "loss": 7.9907,
      "mean_token_accuracy": 0.11302562728524208,
      "num_tokens": 771765.0,
      "step": 305
    },
    {
      "entropy": 8.284095668792725,
      "epoch": 0.03577611079053664,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00015450000000000001,
      "loss": 8.0752,
      "mean_token_accuracy": 0.10686711072921753,
      "num_tokens": 784895.0,
      "step": 310
    },
    {
      "entropy": 8.20336627960205,
      "epoch": 0.0363531448355453,
      "grad_norm": 1.25,
      "learning_rate": 0.000157,
      "loss": 8.0143,
      "mean_token_accuracy": 0.11653341501951217,
      "num_tokens": 797139.0,
      "step": 315
    },
    {
      "entropy": 8.277932500839233,
      "epoch": 0.03693017888055395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001595,
      "loss": 7.9691,
      "mean_token_accuracy": 0.11882287114858628,
      "num_tokens": 809169.0,
      "step": 320
    },
    {
      "entropy": 8.219804096221925,
      "epoch": 0.03750721292556261,
      "grad_norm": 1.078125,
      "learning_rate": 0.000162,
      "loss": 8.017,
      "mean_token_accuracy": 0.11045632883906364,
      "num_tokens": 821737.0,
      "step": 325
    },
    {
      "entropy": 8.202571201324464,
      "epoch": 0.038084246970571264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.9302,
      "mean_token_accuracy": 0.1168684981763363,
      "num_tokens": 834365.0,
      "step": 330
    },
    {
      "entropy": 8.243558883666992,
      "epoch": 0.03866128101557992,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.9747,
      "mean_token_accuracy": 0.11678506284952164,
      "num_tokens": 846445.0,
      "step": 335
    },
    {
      "entropy": 8.234999084472657,
      "epoch": 0.03923831506058857,
      "grad_norm": 1.1875,
      "learning_rate": 0.00016950000000000003,
      "loss": 8.023,
      "mean_token_accuracy": 0.11369575336575508,
      "num_tokens": 858558.0,
      "step": 340
    },
    {
      "entropy": 8.161874151229858,
      "epoch": 0.03981534910559723,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00017199999999999998,
      "loss": 8.0332,
      "mean_token_accuracy": 0.10968955010175704,
      "num_tokens": 871209.0,
      "step": 345
    },
    {
      "entropy": 8.284571075439453,
      "epoch": 0.040392383150605886,
      "grad_norm": 1.15625,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.9815,
      "mean_token_accuracy": 0.11281129494309425,
      "num_tokens": 883270.0,
      "step": 350
    },
    {
      "entropy": 8.099399852752686,
      "epoch": 0.04096941719561454,
      "grad_norm": 1.1875,
      "learning_rate": 0.000177,
      "loss": 7.9788,
      "mean_token_accuracy": 0.11309987381100654,
      "num_tokens": 895468.0,
      "step": 355
    },
    {
      "entropy": 8.298340988159179,
      "epoch": 0.0415464512406232,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001795,
      "loss": 7.9521,
      "mean_token_accuracy": 0.11634099185466766,
      "num_tokens": 907900.0,
      "step": 360
    },
    {
      "entropy": 8.028721857070924,
      "epoch": 0.042123485285631854,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000182,
      "loss": 7.9507,
      "mean_token_accuracy": 0.11578193232417107,
      "num_tokens": 920036.0,
      "step": 365
    },
    {
      "entropy": 8.210243511199952,
      "epoch": 0.04270051933064051,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001845,
      "loss": 7.9225,
      "mean_token_accuracy": 0.12110625132918358,
      "num_tokens": 932525.0,
      "step": 370
    },
    {
      "entropy": 8.095637941360474,
      "epoch": 0.04327755337564916,
      "grad_norm": 1.21875,
      "learning_rate": 0.000187,
      "loss": 7.8898,
      "mean_token_accuracy": 0.12433087080717087,
      "num_tokens": 943581.0,
      "step": 375
    },
    {
      "entropy": 8.17782621383667,
      "epoch": 0.04385458742065782,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001895,
      "loss": 7.9898,
      "mean_token_accuracy": 0.11301371231675147,
      "num_tokens": 956270.0,
      "step": 380
    },
    {
      "entropy": 8.173714351654052,
      "epoch": 0.044431621465666475,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000192,
      "loss": 7.9475,
      "mean_token_accuracy": 0.11783768609166145,
      "num_tokens": 967631.0,
      "step": 385
    },
    {
      "entropy": 8.170480346679687,
      "epoch": 0.04500865551067513,
      "grad_norm": 1.28125,
      "learning_rate": 0.0001945,
      "loss": 7.9699,
      "mean_token_accuracy": 0.11223287805914879,
      "num_tokens": 980999.0,
      "step": 390
    },
    {
      "entropy": 8.166180324554443,
      "epoch": 0.04558568955568378,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.8761,
      "mean_token_accuracy": 0.11871615797281265,
      "num_tokens": 993404.0,
      "step": 395
    },
    {
      "entropy": 7.9651154518127445,
      "epoch": 0.04616272360069244,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.8892,
      "mean_token_accuracy": 0.11921441331505775,
      "num_tokens": 1006009.0,
      "step": 400
    },
    {
      "entropy": 8.124106407165527,
      "epoch": 0.0467397576457011,
      "grad_norm": 1.015625,
      "learning_rate": 0.000202,
      "loss": 7.9879,
      "mean_token_accuracy": 0.11321467459201813,
      "num_tokens": 1019644.0,
      "step": 405
    },
    {
      "entropy": 8.151873683929443,
      "epoch": 0.04731679169070975,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.8907,
      "mean_token_accuracy": 0.12443416640162468,
      "num_tokens": 1032196.0,
      "step": 410
    },
    {
      "entropy": 7.957633638381958,
      "epoch": 0.04789382573571841,
      "grad_norm": 1.28125,
      "learning_rate": 0.000207,
      "loss": 7.8459,
      "mean_token_accuracy": 0.12779026255011558,
      "num_tokens": 1044211.0,
      "step": 415
    },
    {
      "entropy": 8.143306255340576,
      "epoch": 0.048470859780727064,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002095,
      "loss": 7.8905,
      "mean_token_accuracy": 0.12050737589597701,
      "num_tokens": 1055743.0,
      "step": 420
    },
    {
      "entropy": 8.111260080337525,
      "epoch": 0.04904789382573572,
      "grad_norm": 1.203125,
      "learning_rate": 0.000212,
      "loss": 7.8987,
      "mean_token_accuracy": 0.11890480741858482,
      "num_tokens": 1068550.0,
      "step": 425
    },
    {
      "entropy": 8.020257091522216,
      "epoch": 0.04962492787074437,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002145,
      "loss": 7.8119,
      "mean_token_accuracy": 0.12798867896199226,
      "num_tokens": 1079878.0,
      "step": 430
    },
    {
      "entropy": 8.137621879577637,
      "epoch": 0.05020196191575303,
      "grad_norm": 1.328125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.8687,
      "mean_token_accuracy": 0.11096916049718857,
      "num_tokens": 1093331.0,
      "step": 435
    },
    {
      "entropy": 8.066670083999634,
      "epoch": 0.050778995960761686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002195,
      "loss": 7.8257,
      "mean_token_accuracy": 0.12115365341305732,
      "num_tokens": 1105830.0,
      "step": 440
    },
    {
      "entropy": 7.986959838867188,
      "epoch": 0.05135603000577034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000222,
      "loss": 7.8313,
      "mean_token_accuracy": 0.1250676207244396,
      "num_tokens": 1118071.0,
      "step": 445
    },
    {
      "entropy": 8.007787656784057,
      "epoch": 0.05193306405077899,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002245,
      "loss": 7.8457,
      "mean_token_accuracy": 0.12186303734779358,
      "num_tokens": 1129781.0,
      "step": 450
    },
    {
      "entropy": 8.014291572570801,
      "epoch": 0.052510098095787654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.8951,
      "mean_token_accuracy": 0.11840380728244781,
      "num_tokens": 1142750.0,
      "step": 455
    },
    {
      "entropy": 7.9940962314605715,
      "epoch": 0.05308713214079631,
      "grad_norm": 1.1875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.8294,
      "mean_token_accuracy": 0.12203348949551582,
      "num_tokens": 1154314.0,
      "step": 460
    },
    {
      "entropy": 8.042590236663818,
      "epoch": 0.05366416618580496,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.8571,
      "mean_token_accuracy": 0.12293424680829049,
      "num_tokens": 1167199.0,
      "step": 465
    },
    {
      "entropy": 8.021096801757812,
      "epoch": 0.054241200230813615,
      "grad_norm": 1.453125,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.8511,
      "mean_token_accuracy": 0.12046421691775322,
      "num_tokens": 1179654.0,
      "step": 470
    },
    {
      "entropy": 7.927558755874633,
      "epoch": 0.054818234275822275,
      "grad_norm": 1.21875,
      "learning_rate": 0.000237,
      "loss": 7.8023,
      "mean_token_accuracy": 0.1239857017993927,
      "num_tokens": 1190713.0,
      "step": 475
    },
    {
      "entropy": 8.043779516220093,
      "epoch": 0.05539526832083093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002395,
      "loss": 7.8039,
      "mean_token_accuracy": 0.12089489996433259,
      "num_tokens": 1203881.0,
      "step": 480
    },
    {
      "entropy": 8.005364274978637,
      "epoch": 0.05597230236583958,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000242,
      "loss": 7.8446,
      "mean_token_accuracy": 0.1142806276679039,
      "num_tokens": 1215795.0,
      "step": 485
    },
    {
      "entropy": 7.864235591888428,
      "epoch": 0.05654933641084824,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002445,
      "loss": 7.7685,
      "mean_token_accuracy": 0.12962937504053115,
      "num_tokens": 1227773.0,
      "step": 490
    },
    {
      "entropy": 8.027290201187133,
      "epoch": 0.0571263704558569,
      "grad_norm": 1.109375,
      "learning_rate": 0.000247,
      "loss": 7.7609,
      "mean_token_accuracy": 0.1293689340353012,
      "num_tokens": 1240217.0,
      "step": 495
    },
    {
      "entropy": 8.006452465057373,
      "epoch": 0.05770340450086555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002495,
      "loss": 7.8361,
      "mean_token_accuracy": 0.1246810033917427,
      "num_tokens": 1253324.0,
      "step": 500
    },
    {
      "entropy": 7.870799207687378,
      "epoch": 0.058280438545874204,
      "grad_norm": 1.234375,
      "learning_rate": 0.000252,
      "loss": 7.7464,
      "mean_token_accuracy": 0.11857289001345635,
      "num_tokens": 1266007.0,
      "step": 505
    },
    {
      "entropy": 7.910182189941406,
      "epoch": 0.058857472590882864,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002545,
      "loss": 7.7933,
      "mean_token_accuracy": 0.12473395615816116,
      "num_tokens": 1278403.0,
      "step": 510
    },
    {
      "entropy": 7.910435581207276,
      "epoch": 0.05943450663589152,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000257,
      "loss": 7.8167,
      "mean_token_accuracy": 0.1220628336071968,
      "num_tokens": 1290352.0,
      "step": 515
    },
    {
      "entropy": 7.96209020614624,
      "epoch": 0.06001154068090017,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002595,
      "loss": 7.8138,
      "mean_token_accuracy": 0.11629830971360207,
      "num_tokens": 1302304.0,
      "step": 520
    },
    {
      "entropy": 7.892691946029663,
      "epoch": 0.060588574725908825,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000262,
      "loss": 7.7773,
      "mean_token_accuracy": 0.12471387833356858,
      "num_tokens": 1314014.0,
      "step": 525
    },
    {
      "entropy": 8.03321852684021,
      "epoch": 0.061165608770917486,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.8184,
      "mean_token_accuracy": 0.12079318463802338,
      "num_tokens": 1326751.0,
      "step": 530
    },
    {
      "entropy": 7.839642143249511,
      "epoch": 0.06174264281592614,
      "grad_norm": 1.34375,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.6974,
      "mean_token_accuracy": 0.12328156679868699,
      "num_tokens": 1339070.0,
      "step": 535
    },
    {
      "entropy": 7.8171533107757565,
      "epoch": 0.06231967686093479,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.7089,
      "mean_token_accuracy": 0.12542344257235527,
      "num_tokens": 1351345.0,
      "step": 540
    },
    {
      "entropy": 7.918297529220581,
      "epoch": 0.06289671090594345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.7941,
      "mean_token_accuracy": 0.12069407552480697,
      "num_tokens": 1364331.0,
      "step": 545
    },
    {
      "entropy": 7.783697843551636,
      "epoch": 0.06347374495095211,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002745,
      "loss": 7.6797,
      "mean_token_accuracy": 0.13021565973758698,
      "num_tokens": 1377426.0,
      "step": 550
    },
    {
      "entropy": 7.897759199142456,
      "epoch": 0.06405077899596076,
      "grad_norm": 1.09375,
      "learning_rate": 0.000277,
      "loss": 7.7903,
      "mean_token_accuracy": 0.11785958632826805,
      "num_tokens": 1391553.0,
      "step": 555
    },
    {
      "entropy": 7.917001342773437,
      "epoch": 0.06462781304096941,
      "grad_norm": 1.53125,
      "learning_rate": 0.0002795,
      "loss": 7.7446,
      "mean_token_accuracy": 0.12249603047966957,
      "num_tokens": 1405526.0,
      "step": 560
    },
    {
      "entropy": 7.944843435287476,
      "epoch": 0.06520484708597807,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.7697,
      "mean_token_accuracy": 0.12450801730155944,
      "num_tokens": 1417868.0,
      "step": 565
    },
    {
      "entropy": 7.7399509906768795,
      "epoch": 0.06578188113098672,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002845,
      "loss": 7.6914,
      "mean_token_accuracy": 0.12707902863621712,
      "num_tokens": 1429907.0,
      "step": 570
    },
    {
      "entropy": 7.985376596450806,
      "epoch": 0.06635891517599539,
      "grad_norm": 1.15625,
      "learning_rate": 0.000287,
      "loss": 7.7166,
      "mean_token_accuracy": 0.12132233530282974,
      "num_tokens": 1442552.0,
      "step": 575
    },
    {
      "entropy": 7.768247318267822,
      "epoch": 0.06693594922100404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002895,
      "loss": 7.7624,
      "mean_token_accuracy": 0.12234465405344963,
      "num_tokens": 1455717.0,
      "step": 580
    },
    {
      "entropy": 7.9525719165802,
      "epoch": 0.0675129832660127,
      "grad_norm": 1.109375,
      "learning_rate": 0.000292,
      "loss": 7.7642,
      "mean_token_accuracy": 0.1182574063539505,
      "num_tokens": 1468263.0,
      "step": 585
    },
    {
      "entropy": 7.756762266159058,
      "epoch": 0.06809001731102135,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002945,
      "loss": 7.6998,
      "mean_token_accuracy": 0.12650698721408843,
      "num_tokens": 1481632.0,
      "step": 590
    },
    {
      "entropy": 7.854887342453003,
      "epoch": 0.06866705135603,
      "grad_norm": 1.109375,
      "learning_rate": 0.000297,
      "loss": 7.648,
      "mean_token_accuracy": 0.1267064943909645,
      "num_tokens": 1495121.0,
      "step": 595
    },
    {
      "entropy": 7.762281274795532,
      "epoch": 0.06924408540103866,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002995,
      "loss": 7.7071,
      "mean_token_accuracy": 0.12668090239167212,
      "num_tokens": 1507669.0,
      "step": 600
    },
    {
      "entropy": 7.738871908187866,
      "epoch": 0.06982111944604731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000302,
      "loss": 7.7326,
      "mean_token_accuracy": 0.11923622786998749,
      "num_tokens": 1520707.0,
      "step": 605
    },
    {
      "entropy": 7.851386690139771,
      "epoch": 0.07039815349105598,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003045,
      "loss": 7.6367,
      "mean_token_accuracy": 0.12671747878193856,
      "num_tokens": 1532974.0,
      "step": 610
    },
    {
      "entropy": 7.776523685455322,
      "epoch": 0.07097518753606463,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000307,
      "loss": 7.6379,
      "mean_token_accuracy": 0.12370964214205742,
      "num_tokens": 1544085.0,
      "step": 615
    },
    {
      "entropy": 7.664855432510376,
      "epoch": 0.07155222158107329,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003095,
      "loss": 7.6676,
      "mean_token_accuracy": 0.13088590577244757,
      "num_tokens": 1556945.0,
      "step": 620
    },
    {
      "entropy": 7.785807657241821,
      "epoch": 0.07212925562608194,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000312,
      "loss": 7.6149,
      "mean_token_accuracy": 0.13267466947436332,
      "num_tokens": 1569570.0,
      "step": 625
    },
    {
      "entropy": 7.674111986160279,
      "epoch": 0.0727062896710906,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003145,
      "loss": 7.7054,
      "mean_token_accuracy": 0.12257150337100028,
      "num_tokens": 1582529.0,
      "step": 630
    },
    {
      "entropy": 7.782325267791748,
      "epoch": 0.07328332371609925,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000317,
      "loss": 7.6195,
      "mean_token_accuracy": 0.13299935758113862,
      "num_tokens": 1595976.0,
      "step": 635
    },
    {
      "entropy": 7.767532587051392,
      "epoch": 0.0738603577611079,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003195,
      "loss": 7.6424,
      "mean_token_accuracy": 0.12903562188148499,
      "num_tokens": 1609093.0,
      "step": 640
    },
    {
      "entropy": 7.731075191497803,
      "epoch": 0.07443739180611655,
      "grad_norm": 1.03125,
      "learning_rate": 0.000322,
      "loss": 7.622,
      "mean_token_accuracy": 0.12647998183965684,
      "num_tokens": 1622233.0,
      "step": 645
    },
    {
      "entropy": 7.757028245925904,
      "epoch": 0.07501442585112522,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.6987,
      "mean_token_accuracy": 0.12671734392642975,
      "num_tokens": 1635354.0,
      "step": 650
    },
    {
      "entropy": 7.748657846450806,
      "epoch": 0.07559145989613388,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00032700000000000003,
      "loss": 7.659,
      "mean_token_accuracy": 0.12423627153038978,
      "num_tokens": 1648223.0,
      "step": 655
    },
    {
      "entropy": 7.7475920677185055,
      "epoch": 0.07616849394114253,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.6653,
      "mean_token_accuracy": 0.12359859570860862,
      "num_tokens": 1661652.0,
      "step": 660
    },
    {
      "entropy": 7.648698377609253,
      "epoch": 0.07674552798615118,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.5636,
      "mean_token_accuracy": 0.13243083134293557,
      "num_tokens": 1674778.0,
      "step": 665
    },
    {
      "entropy": 7.787905645370484,
      "epoch": 0.07732256203115984,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.6459,
      "mean_token_accuracy": 0.12540043219923974,
      "num_tokens": 1686966.0,
      "step": 670
    },
    {
      "entropy": 7.660491418838501,
      "epoch": 0.07789959607616849,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000337,
      "loss": 7.623,
      "mean_token_accuracy": 0.13194708675146102,
      "num_tokens": 1699979.0,
      "step": 675
    },
    {
      "entropy": 7.710208940505981,
      "epoch": 0.07847663012117714,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003395,
      "loss": 7.5421,
      "mean_token_accuracy": 0.1282073549926281,
      "num_tokens": 1712306.0,
      "step": 680
    },
    {
      "entropy": 7.66764645576477,
      "epoch": 0.07905366416618581,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000342,
      "loss": 7.6151,
      "mean_token_accuracy": 0.12962080165743828,
      "num_tokens": 1724670.0,
      "step": 685
    },
    {
      "entropy": 7.820890140533447,
      "epoch": 0.07963069821119446,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.6844,
      "mean_token_accuracy": 0.12687815874814987,
      "num_tokens": 1736987.0,
      "step": 690
    },
    {
      "entropy": 7.658799171447754,
      "epoch": 0.08020773225620312,
      "grad_norm": 1.125,
      "learning_rate": 0.000347,
      "loss": 7.6473,
      "mean_token_accuracy": 0.12734541893005372,
      "num_tokens": 1750587.0,
      "step": 695
    },
    {
      "entropy": 7.620698738098144,
      "epoch": 0.08078476630121177,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003495,
      "loss": 7.6201,
      "mean_token_accuracy": 0.12172205820679664,
      "num_tokens": 1763432.0,
      "step": 700
    },
    {
      "entropy": 7.6917338371276855,
      "epoch": 0.08136180034622043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000352,
      "loss": 7.5204,
      "mean_token_accuracy": 0.13158636018633843,
      "num_tokens": 1777705.0,
      "step": 705
    },
    {
      "entropy": 7.608352470397949,
      "epoch": 0.08193883439122908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003545,
      "loss": 7.5904,
      "mean_token_accuracy": 0.13091936558485032,
      "num_tokens": 1790673.0,
      "step": 710
    },
    {
      "entropy": 7.676751232147216,
      "epoch": 0.08251586843623773,
      "grad_norm": 1.078125,
      "learning_rate": 0.000357,
      "loss": 7.6165,
      "mean_token_accuracy": 0.12561856135725974,
      "num_tokens": 1803258.0,
      "step": 715
    },
    {
      "entropy": 7.677963590621948,
      "epoch": 0.0830929024812464,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003595,
      "loss": 7.5943,
      "mean_token_accuracy": 0.1276804856956005,
      "num_tokens": 1816534.0,
      "step": 720
    },
    {
      "entropy": 7.624732446670532,
      "epoch": 0.08366993652625505,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000362,
      "loss": 7.5413,
      "mean_token_accuracy": 0.12490850538015366,
      "num_tokens": 1828335.0,
      "step": 725
    },
    {
      "entropy": 7.605259799957276,
      "epoch": 0.08424697057126371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003645,
      "loss": 7.5646,
      "mean_token_accuracy": 0.13269523456692695,
      "num_tokens": 1839327.0,
      "step": 730
    },
    {
      "entropy": 7.720489597320556,
      "epoch": 0.08482400461627236,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000367,
      "loss": 7.6267,
      "mean_token_accuracy": 0.11925766617059708,
      "num_tokens": 1851168.0,
      "step": 735
    },
    {
      "entropy": 7.6475683689117435,
      "epoch": 0.08540103866128101,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003695,
      "loss": 7.6049,
      "mean_token_accuracy": 0.124794951826334,
      "num_tokens": 1863875.0,
      "step": 740
    },
    {
      "entropy": 7.579368352890015,
      "epoch": 0.08597807270628967,
      "grad_norm": 1.125,
      "learning_rate": 0.000372,
      "loss": 7.5178,
      "mean_token_accuracy": 0.12288314551115036,
      "num_tokens": 1876613.0,
      "step": 745
    },
    {
      "entropy": 7.694609594345093,
      "epoch": 0.08655510675129832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003745,
      "loss": 7.6241,
      "mean_token_accuracy": 0.12489113733172416,
      "num_tokens": 1890311.0,
      "step": 750
    },
    {
      "entropy": 7.602844142913819,
      "epoch": 0.08713214079630698,
      "grad_norm": 1.09375,
      "learning_rate": 0.000377,
      "loss": 7.5948,
      "mean_token_accuracy": 0.12589145004749297,
      "num_tokens": 1902925.0,
      "step": 755
    },
    {
      "entropy": 7.596966028213501,
      "epoch": 0.08770917484131564,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003795,
      "loss": 7.5689,
      "mean_token_accuracy": 0.123939748108387,
      "num_tokens": 1915186.0,
      "step": 760
    },
    {
      "entropy": 7.636845922470092,
      "epoch": 0.0882862088863243,
      "grad_norm": 1.125,
      "learning_rate": 0.000382,
      "loss": 7.571,
      "mean_token_accuracy": 0.13280307427048682,
      "num_tokens": 1927184.0,
      "step": 765
    },
    {
      "entropy": 7.59859070777893,
      "epoch": 0.08886324293133295,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003845,
      "loss": 7.5462,
      "mean_token_accuracy": 0.134385896474123,
      "num_tokens": 1940275.0,
      "step": 770
    },
    {
      "entropy": 7.626807355880738,
      "epoch": 0.0894402769763416,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.5775,
      "mean_token_accuracy": 0.12614732086658478,
      "num_tokens": 1952583.0,
      "step": 775
    },
    {
      "entropy": 7.596640682220459,
      "epoch": 0.09001731102135026,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.4969,
      "mean_token_accuracy": 0.12433894649147988,
      "num_tokens": 1965219.0,
      "step": 780
    },
    {
      "entropy": 7.571691226959229,
      "epoch": 0.09059434506635891,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.5153,
      "mean_token_accuracy": 0.13145391047000884,
      "num_tokens": 1977193.0,
      "step": 785
    },
    {
      "entropy": 7.635157775878906,
      "epoch": 0.09117137911136756,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039450000000000005,
      "loss": 7.4312,
      "mean_token_accuracy": 0.1304060213267803,
      "num_tokens": 1988955.0,
      "step": 790
    },
    {
      "entropy": 7.514168548583984,
      "epoch": 0.09174841315637623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039700000000000005,
      "loss": 7.5695,
      "mean_token_accuracy": 0.12537927627563478,
      "num_tokens": 2001647.0,
      "step": 795
    },
    {
      "entropy": 7.537772274017334,
      "epoch": 0.09232544720138489,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003995,
      "loss": 7.4927,
      "mean_token_accuracy": 0.13620842844247819,
      "num_tokens": 2015185.0,
      "step": 800
    },
    {
      "entropy": 7.663471412658692,
      "epoch": 0.09290248124639354,
      "grad_norm": 1.0,
      "learning_rate": 0.000402,
      "loss": 7.5901,
      "mean_token_accuracy": 0.1272108480334282,
      "num_tokens": 2027257.0,
      "step": 805
    },
    {
      "entropy": 7.390483283996582,
      "epoch": 0.0934795152914022,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004045,
      "loss": 7.4161,
      "mean_token_accuracy": 0.13912721797823907,
      "num_tokens": 2040716.0,
      "step": 810
    },
    {
      "entropy": 7.560983800888062,
      "epoch": 0.09405654933641085,
      "grad_norm": 1.234375,
      "learning_rate": 0.00040699999999999997,
      "loss": 7.55,
      "mean_token_accuracy": 0.12221533954143524,
      "num_tokens": 2053240.0,
      "step": 815
    },
    {
      "entropy": 7.622315168380737,
      "epoch": 0.0946335833814195,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004095,
      "loss": 7.5139,
      "mean_token_accuracy": 0.12666195258498192,
      "num_tokens": 2065693.0,
      "step": 820
    },
    {
      "entropy": 7.4823235988616945,
      "epoch": 0.09521061742642815,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000412,
      "loss": 7.4591,
      "mean_token_accuracy": 0.14022310078144073,
      "num_tokens": 2077649.0,
      "step": 825
    },
    {
      "entropy": 7.467005395889283,
      "epoch": 0.09578765147143682,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004145,
      "loss": 7.5063,
      "mean_token_accuracy": 0.12628407776355743,
      "num_tokens": 2090780.0,
      "step": 830
    },
    {
      "entropy": 7.546123218536377,
      "epoch": 0.09636468551644548,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000417,
      "loss": 7.513,
      "mean_token_accuracy": 0.13211931437253951,
      "num_tokens": 2104096.0,
      "step": 835
    },
    {
      "entropy": 7.543429565429688,
      "epoch": 0.09694171956145413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004195,
      "loss": 7.5109,
      "mean_token_accuracy": 0.12248567417263985,
      "num_tokens": 2117157.0,
      "step": 840
    },
    {
      "entropy": 7.5339704036712645,
      "epoch": 0.09751875360646278,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000422,
      "loss": 7.4826,
      "mean_token_accuracy": 0.12713147699832916,
      "num_tokens": 2129309.0,
      "step": 845
    },
    {
      "entropy": 7.466995716094971,
      "epoch": 0.09809578765147144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004245,
      "loss": 7.5207,
      "mean_token_accuracy": 0.12562475129961967,
      "num_tokens": 2142344.0,
      "step": 850
    },
    {
      "entropy": 7.547919607162475,
      "epoch": 0.09867282169648009,
      "grad_norm": 1.171875,
      "learning_rate": 0.000427,
      "loss": 7.5488,
      "mean_token_accuracy": 0.12197135463356971,
      "num_tokens": 2155059.0,
      "step": 855
    },
    {
      "entropy": 7.569330024719238,
      "epoch": 0.09924985574148874,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004295,
      "loss": 7.5369,
      "mean_token_accuracy": 0.1284255087375641,
      "num_tokens": 2168901.0,
      "step": 860
    },
    {
      "entropy": 7.4198370456695555,
      "epoch": 0.0998268897864974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000432,
      "loss": 7.3584,
      "mean_token_accuracy": 0.13862554877996444,
      "num_tokens": 2180953.0,
      "step": 865
    },
    {
      "entropy": 7.476253461837769,
      "epoch": 0.10040392383150606,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004345,
      "loss": 7.4631,
      "mean_token_accuracy": 0.13572433739900588,
      "num_tokens": 2194097.0,
      "step": 870
    },
    {
      "entropy": 7.479328727722168,
      "epoch": 0.10098095787651472,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000437,
      "loss": 7.4058,
      "mean_token_accuracy": 0.13234824016690255,
      "num_tokens": 2206475.0,
      "step": 875
    },
    {
      "entropy": 7.484198713302613,
      "epoch": 0.10155799192152337,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004395,
      "loss": 7.436,
      "mean_token_accuracy": 0.12529727593064308,
      "num_tokens": 2220006.0,
      "step": 880
    },
    {
      "entropy": 7.488145637512207,
      "epoch": 0.10213502596653203,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000442,
      "loss": 7.4962,
      "mean_token_accuracy": 0.12320048362016678,
      "num_tokens": 2234082.0,
      "step": 885
    },
    {
      "entropy": 7.484939479827881,
      "epoch": 0.10271206001154068,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004445,
      "loss": 7.4908,
      "mean_token_accuracy": 0.12574098110198975,
      "num_tokens": 2246871.0,
      "step": 890
    },
    {
      "entropy": 7.439050626754761,
      "epoch": 0.10328909405654933,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000447,
      "loss": 7.4344,
      "mean_token_accuracy": 0.13258244097232819,
      "num_tokens": 2260172.0,
      "step": 895
    },
    {
      "entropy": 7.573509597778321,
      "epoch": 0.10386612810155799,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044950000000000003,
      "loss": 7.4693,
      "mean_token_accuracy": 0.12150668054819107,
      "num_tokens": 2272848.0,
      "step": 900
    },
    {
      "entropy": 7.414100122451782,
      "epoch": 0.10444316214656665,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045200000000000004,
      "loss": 7.4238,
      "mean_token_accuracy": 0.128422711789608,
      "num_tokens": 2285673.0,
      "step": 905
    },
    {
      "entropy": 7.493473386764526,
      "epoch": 0.10502019619157531,
      "grad_norm": 0.921875,
      "learning_rate": 0.00045450000000000004,
      "loss": 7.47,
      "mean_token_accuracy": 0.12711283564567566,
      "num_tokens": 2298459.0,
      "step": 910
    },
    {
      "entropy": 7.421923589706421,
      "epoch": 0.10559723023658396,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045700000000000005,
      "loss": 7.4389,
      "mean_token_accuracy": 0.1314757324755192,
      "num_tokens": 2311226.0,
      "step": 915
    },
    {
      "entropy": 7.34346604347229,
      "epoch": 0.10617426428159261,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.3694,
      "mean_token_accuracy": 0.1267826184630394,
      "num_tokens": 2323481.0,
      "step": 920
    },
    {
      "entropy": 7.37224383354187,
      "epoch": 0.10675129832660127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000462,
      "loss": 7.4024,
      "mean_token_accuracy": 0.13083600699901582,
      "num_tokens": 2336181.0,
      "step": 925
    },
    {
      "entropy": 7.499636030197143,
      "epoch": 0.10732833237160992,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004645,
      "loss": 7.3534,
      "mean_token_accuracy": 0.13574053421616555,
      "num_tokens": 2348941.0,
      "step": 930
    },
    {
      "entropy": 7.3138683319091795,
      "epoch": 0.10790536641661858,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000467,
      "loss": 7.2945,
      "mean_token_accuracy": 0.1320072427392006,
      "num_tokens": 2361618.0,
      "step": 935
    },
    {
      "entropy": 7.307975149154663,
      "epoch": 0.10848240046162723,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004695,
      "loss": 7.259,
      "mean_token_accuracy": 0.1410675063729286,
      "num_tokens": 2376108.0,
      "step": 940
    },
    {
      "entropy": 7.366646480560303,
      "epoch": 0.1090594345066359,
      "grad_norm": 1.0,
      "learning_rate": 0.000472,
      "loss": 7.3883,
      "mean_token_accuracy": 0.13348079323768616,
      "num_tokens": 2387774.0,
      "step": 945
    },
    {
      "entropy": 7.387966251373291,
      "epoch": 0.10963646855164455,
      "grad_norm": 1.125,
      "learning_rate": 0.0004745,
      "loss": 7.3984,
      "mean_token_accuracy": 0.13679807931184768,
      "num_tokens": 2400816.0,
      "step": 950
    },
    {
      "entropy": 7.225770616531372,
      "epoch": 0.1102135025966532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000477,
      "loss": 7.3332,
      "mean_token_accuracy": 0.12989477887749673,
      "num_tokens": 2412970.0,
      "step": 955
    },
    {
      "entropy": 7.398374080657959,
      "epoch": 0.11079053664166186,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004795,
      "loss": 7.3348,
      "mean_token_accuracy": 0.13138538971543312,
      "num_tokens": 2424622.0,
      "step": 960
    },
    {
      "entropy": 7.423074054718017,
      "epoch": 0.11136757068667051,
      "grad_norm": 0.921875,
      "learning_rate": 0.000482,
      "loss": 7.4257,
      "mean_token_accuracy": 0.13353453874588012,
      "num_tokens": 2437293.0,
      "step": 965
    },
    {
      "entropy": 7.288896751403809,
      "epoch": 0.11194460473167916,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004845,
      "loss": 7.3092,
      "mean_token_accuracy": 0.12915137261152268,
      "num_tokens": 2449934.0,
      "step": 970
    },
    {
      "entropy": 7.371738529205322,
      "epoch": 0.11252163877668782,
      "grad_norm": 0.921875,
      "learning_rate": 0.000487,
      "loss": 7.3638,
      "mean_token_accuracy": 0.1267050661146641,
      "num_tokens": 2463139.0,
      "step": 975
    },
    {
      "entropy": 7.306709432601929,
      "epoch": 0.11309867282169649,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895,
      "loss": 7.4078,
      "mean_token_accuracy": 0.12786821126937867,
      "num_tokens": 2474273.0,
      "step": 980
    },
    {
      "entropy": 7.4030670642852785,
      "epoch": 0.11367570686670514,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000492,
      "loss": 7.3521,
      "mean_token_accuracy": 0.13926174938678743,
      "num_tokens": 2486643.0,
      "step": 985
    },
    {
      "entropy": 7.327881813049316,
      "epoch": 0.1142527409117138,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945,
      "loss": 7.3334,
      "mean_token_accuracy": 0.12991991862654687,
      "num_tokens": 2499018.0,
      "step": 990
    },
    {
      "entropy": 7.229681015014648,
      "epoch": 0.11482977495672245,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497,
      "loss": 7.3209,
      "mean_token_accuracy": 0.13620815202593803,
      "num_tokens": 2510463.0,
      "step": 995
    },
    {
      "entropy": 7.333762550354004,
      "epoch": 0.1154068090017311,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995,
      "loss": 7.3648,
      "mean_token_accuracy": 0.13373479768633842,
      "num_tokens": 2522671.0,
      "step": 1000
    },
    {
      "entropy": 7.2945623874664305,
      "epoch": 0.11598384304673975,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999999975783157,
      "loss": 7.3465,
      "mean_token_accuracy": 0.13163619264960288,
      "num_tokens": 2534508.0,
      "step": 1005
    },
    {
      "entropy": 7.360085344314575,
      "epoch": 0.11656087709174841,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999999877402236,
      "loss": 7.322,
      "mean_token_accuracy": 0.13855402842164039,
      "num_tokens": 2546804.0,
      "step": 1010
    },
    {
      "entropy": 7.284858369827271,
      "epoch": 0.11713791113675708,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999999703343686,
      "loss": 7.3227,
      "mean_token_accuracy": 0.12957534343004226,
      "num_tokens": 2558786.0,
      "step": 1015
    },
    {
      "entropy": 7.372665882110596,
      "epoch": 0.11771494518176573,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999999453607515,
      "loss": 7.2956,
      "mean_token_accuracy": 0.13060993254184722,
      "num_tokens": 2570834.0,
      "step": 1020
    },
    {
      "entropy": 7.1804040431976315,
      "epoch": 0.11829197922677438,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999999128193729,
      "loss": 7.2577,
      "mean_token_accuracy": 0.13870095685124398,
      "num_tokens": 2583342.0,
      "step": 1025
    },
    {
      "entropy": 7.312113237380982,
      "epoch": 0.11886901327178304,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499999872710234,
      "loss": 7.3031,
      "mean_token_accuracy": 0.13401906490325927,
      "num_tokens": 2594878.0,
      "step": 1030
    },
    {
      "entropy": 7.149050855636597,
      "epoch": 0.11944604731679169,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999998250333361,
      "loss": 7.2657,
      "mean_token_accuracy": 0.13629036098718644,
      "num_tokens": 2607070.0,
      "step": 1035
    },
    {
      "entropy": 7.316418886184692,
      "epoch": 0.12002308136180034,
      "grad_norm": 0.8203125,
      "learning_rate": 0.000499999769788681,
      "loss": 7.3088,
      "mean_token_accuracy": 0.13769753351807595,
      "num_tokens": 2620761.0,
      "step": 1040
    },
    {
      "entropy": 7.278573226928711,
      "epoch": 0.120600115406809,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999997069762703,
      "loss": 7.3204,
      "mean_token_accuracy": 0.1292047843337059,
      "num_tokens": 2634161.0,
      "step": 1045
    },
    {
      "entropy": 7.217658758163452,
      "epoch": 0.12117714945181765,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999996365961062,
      "loss": 7.2649,
      "mean_token_accuracy": 0.13989781066775323,
      "num_tokens": 2647459.0,
      "step": 1050
    },
    {
      "entropy": 7.200538969039917,
      "epoch": 0.12175418349682632,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999995586481912,
      "loss": 7.2675,
      "mean_token_accuracy": 0.13167453035712243,
      "num_tokens": 2660840.0,
      "step": 1055
    },
    {
      "entropy": 7.280622291564941,
      "epoch": 0.12233121754183497,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999994731325276,
      "loss": 7.2731,
      "mean_token_accuracy": 0.13342294991016387,
      "num_tokens": 2672673.0,
      "step": 1060
    },
    {
      "entropy": 7.171860647201538,
      "epoch": 0.12290825158684363,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999993800491187,
      "loss": 7.289,
      "mean_token_accuracy": 0.13255034536123275,
      "num_tokens": 2685744.0,
      "step": 1065
    },
    {
      "entropy": 7.219749069213867,
      "epoch": 0.12348528563185228,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999992793979672,
      "loss": 7.199,
      "mean_token_accuracy": 0.14802319854497908,
      "num_tokens": 2697068.0,
      "step": 1070
    },
    {
      "entropy": 7.331407928466797,
      "epoch": 0.12406231967686093,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999991711790769,
      "loss": 7.3567,
      "mean_token_accuracy": 0.12301772236824035,
      "num_tokens": 2710418.0,
      "step": 1075
    },
    {
      "entropy": 7.252064895629883,
      "epoch": 0.12463935372186959,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999990553924511,
      "loss": 7.3035,
      "mean_token_accuracy": 0.12562899366021157,
      "num_tokens": 2722440.0,
      "step": 1080
    },
    {
      "entropy": 7.257036256790161,
      "epoch": 0.12521638776687824,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004999989320380939,
      "loss": 7.2928,
      "mean_token_accuracy": 0.12703390195965766,
      "num_tokens": 2735486.0,
      "step": 1085
    },
    {
      "entropy": 7.245379257202148,
      "epoch": 0.1257934218118869,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999988011160094,
      "loss": 7.2958,
      "mean_token_accuracy": 0.12891678139567375,
      "num_tokens": 2747620.0,
      "step": 1090
    },
    {
      "entropy": 7.229975891113281,
      "epoch": 0.12637045585689555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499998662626202,
      "loss": 7.2443,
      "mean_token_accuracy": 0.13310989141464233,
      "num_tokens": 2760987.0,
      "step": 1095
    },
    {
      "entropy": 7.242767477035523,
      "epoch": 0.12694748990190421,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999985165686764,
      "loss": 7.2233,
      "mean_token_accuracy": 0.13173784613609313,
      "num_tokens": 2773161.0,
      "step": 1100
    },
    {
      "entropy": 7.143377590179443,
      "epoch": 0.12752452394691285,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999983629434373,
      "loss": 7.1702,
      "mean_token_accuracy": 0.1464059606194496,
      "num_tokens": 2786179.0,
      "step": 1105
    },
    {
      "entropy": 7.255331087112427,
      "epoch": 0.12810155799192152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999982017504901,
      "loss": 7.1856,
      "mean_token_accuracy": 0.14088639095425606,
      "num_tokens": 2798859.0,
      "step": 1110
    },
    {
      "entropy": 7.152024507522583,
      "epoch": 0.1286785920369302,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999980329898401,
      "loss": 7.2883,
      "mean_token_accuracy": 0.13371546044945717,
      "num_tokens": 2813048.0,
      "step": 1115
    },
    {
      "entropy": 7.280220174789429,
      "epoch": 0.12925562608193883,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499997856661493,
      "loss": 7.2463,
      "mean_token_accuracy": 0.13465555906295776,
      "num_tokens": 2826032.0,
      "step": 1120
    },
    {
      "entropy": 7.17284836769104,
      "epoch": 0.1298326601269475,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499997672765455,
      "loss": 7.2976,
      "mean_token_accuracy": 0.133281010389328,
      "num_tokens": 2839253.0,
      "step": 1125
    },
    {
      "entropy": 7.185248899459839,
      "epoch": 0.13040969417195614,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004999974813017318,
      "loss": 7.2206,
      "mean_token_accuracy": 0.1415400303900242,
      "num_tokens": 2852542.0,
      "step": 1130
    },
    {
      "entropy": 7.138202762603759,
      "epoch": 0.1309867282169648,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999972822703301,
      "loss": 7.2218,
      "mean_token_accuracy": 0.13877267017960548,
      "num_tokens": 2863967.0,
      "step": 1135
    },
    {
      "entropy": 7.290763425827026,
      "epoch": 0.13156376226197344,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999970756712567,
      "loss": 7.3254,
      "mean_token_accuracy": 0.12893034666776657,
      "num_tokens": 2876141.0,
      "step": 1140
    },
    {
      "entropy": 7.014846658706665,
      "epoch": 0.1321407963069821,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999968615045183,
      "loss": 7.1565,
      "mean_token_accuracy": 0.13931377455592156,
      "num_tokens": 2888395.0,
      "step": 1145
    },
    {
      "entropy": 7.292420244216919,
      "epoch": 0.13271783035199078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999966397701222,
      "loss": 7.3365,
      "mean_token_accuracy": 0.1309019148349762,
      "num_tokens": 2902399.0,
      "step": 1150
    },
    {
      "entropy": 7.157604312896728,
      "epoch": 0.13329486439699942,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999964104680759,
      "loss": 7.0878,
      "mean_token_accuracy": 0.13564511761069298,
      "num_tokens": 2914999.0,
      "step": 1155
    },
    {
      "entropy": 7.144139671325684,
      "epoch": 0.13387189844200809,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999961735983871,
      "loss": 7.0883,
      "mean_token_accuracy": 0.14829797372221948,
      "num_tokens": 2928091.0,
      "step": 1160
    },
    {
      "entropy": 7.185041427612305,
      "epoch": 0.13444893248701673,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999959291610639,
      "loss": 7.207,
      "mean_token_accuracy": 0.13267041072249414,
      "num_tokens": 2940840.0,
      "step": 1165
    },
    {
      "entropy": 7.118538331985474,
      "epoch": 0.1350259665320254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999956771561143,
      "loss": 7.1349,
      "mean_token_accuracy": 0.13792715817689896,
      "num_tokens": 2952980.0,
      "step": 1170
    },
    {
      "entropy": 7.110788869857788,
      "epoch": 0.13560300057703403,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999954175835469,
      "loss": 7.1856,
      "mean_token_accuracy": 0.13442225083708764,
      "num_tokens": 2965437.0,
      "step": 1175
    },
    {
      "entropy": 7.1716221332550045,
      "epoch": 0.1361800346220427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999951504433703,
      "loss": 7.2135,
      "mean_token_accuracy": 0.13230456560850143,
      "num_tokens": 2978077.0,
      "step": 1180
    },
    {
      "entropy": 7.156641292572021,
      "epoch": 0.13675706866705137,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999948757355935,
      "loss": 7.2831,
      "mean_token_accuracy": 0.13261316642165183,
      "num_tokens": 2990498.0,
      "step": 1185
    },
    {
      "entropy": 7.198510980606079,
      "epoch": 0.13733410271206,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499994593460226,
      "loss": 7.2269,
      "mean_token_accuracy": 0.13312480375170707,
      "num_tokens": 3002784.0,
      "step": 1190
    },
    {
      "entropy": 7.238051557540894,
      "epoch": 0.13791113675706868,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999943036172771,
      "loss": 7.1808,
      "mean_token_accuracy": 0.13066396713256836,
      "num_tokens": 3015689.0,
      "step": 1195
    },
    {
      "entropy": 7.092861557006836,
      "epoch": 0.13848817080207732,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999940062067565,
      "loss": 7.1113,
      "mean_token_accuracy": 0.14541881829500197,
      "num_tokens": 3026623.0,
      "step": 1200
    },
    {
      "entropy": 7.083968591690064,
      "epoch": 0.13906520484708598,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999937012286742,
      "loss": 7.1675,
      "mean_token_accuracy": 0.14061277359724045,
      "num_tokens": 3039769.0,
      "step": 1205
    },
    {
      "entropy": 7.057619905471801,
      "epoch": 0.13964223889209462,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999933886830405,
      "loss": 7.1518,
      "mean_token_accuracy": 0.1339143082499504,
      "num_tokens": 3052044.0,
      "step": 1210
    },
    {
      "entropy": 7.185735702514648,
      "epoch": 0.1402192729371033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499993068569866,
      "loss": 7.1173,
      "mean_token_accuracy": 0.13880359381437302,
      "num_tokens": 3063937.0,
      "step": 1215
    },
    {
      "entropy": 7.0055492401123045,
      "epoch": 0.14079630698211196,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999927408891614,
      "loss": 7.0929,
      "mean_token_accuracy": 0.14283955842256546,
      "num_tokens": 3076645.0,
      "step": 1220
    },
    {
      "entropy": 7.1528167724609375,
      "epoch": 0.1413733410271206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999924056409378,
      "loss": 7.1459,
      "mean_token_accuracy": 0.1395370438694954,
      "num_tokens": 3088985.0,
      "step": 1225
    },
    {
      "entropy": 7.103240489959717,
      "epoch": 0.14195037507212926,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999920628252063,
      "loss": 7.1467,
      "mean_token_accuracy": 0.13826994374394416,
      "num_tokens": 3101270.0,
      "step": 1230
    },
    {
      "entropy": 7.075673818588257,
      "epoch": 0.1425274091171379,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999917124419785,
      "loss": 7.1521,
      "mean_token_accuracy": 0.13071254119277,
      "num_tokens": 3115420.0,
      "step": 1235
    },
    {
      "entropy": 7.155364894866944,
      "epoch": 0.14310444316214657,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999913544912664,
      "loss": 7.101,
      "mean_token_accuracy": 0.13510319143533706,
      "num_tokens": 3126176.0,
      "step": 1240
    },
    {
      "entropy": 7.037791395187378,
      "epoch": 0.1436814772071552,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004999909889730817,
      "loss": 7.134,
      "mean_token_accuracy": 0.14009907394647597,
      "num_tokens": 3139136.0,
      "step": 1245
    },
    {
      "entropy": 7.121502780914307,
      "epoch": 0.14425851125216388,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999906158874368,
      "loss": 7.0984,
      "mean_token_accuracy": 0.13803695887327194,
      "num_tokens": 3152036.0,
      "step": 1250
    },
    {
      "entropy": 7.089022874832153,
      "epoch": 0.14483554529717255,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999902352343444,
      "loss": 7.124,
      "mean_token_accuracy": 0.1390008471906185,
      "num_tokens": 3164060.0,
      "step": 1255
    },
    {
      "entropy": 7.158830451965332,
      "epoch": 0.1454125793421812,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999898470138171,
      "loss": 7.2071,
      "mean_token_accuracy": 0.12887537702918053,
      "num_tokens": 3176410.0,
      "step": 1260
    },
    {
      "entropy": 7.16694221496582,
      "epoch": 0.14598961338718985,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999894512258681,
      "loss": 7.2054,
      "mean_token_accuracy": 0.13165885508060454,
      "num_tokens": 3189784.0,
      "step": 1265
    },
    {
      "entropy": 7.0905242443084715,
      "epoch": 0.1465666474321985,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999890478705107,
      "loss": 7.126,
      "mean_token_accuracy": 0.1364673689007759,
      "num_tokens": 3202274.0,
      "step": 1270
    },
    {
      "entropy": 7.164013195037842,
      "epoch": 0.14714368147720716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999886369477585,
      "loss": 7.1967,
      "mean_token_accuracy": 0.1339113764464855,
      "num_tokens": 3215593.0,
      "step": 1275
    },
    {
      "entropy": 7.090557193756103,
      "epoch": 0.1477207155222158,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999882184576251,
      "loss": 7.0904,
      "mean_token_accuracy": 0.13543560802936555,
      "num_tokens": 3228307.0,
      "step": 1280
    },
    {
      "entropy": 7.0394628047943115,
      "epoch": 0.14829774956722447,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004999877924001248,
      "loss": 7.063,
      "mean_token_accuracy": 0.137285415828228,
      "num_tokens": 3241676.0,
      "step": 1285
    },
    {
      "entropy": 7.106666898727417,
      "epoch": 0.1488747836122331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999873587752718,
      "loss": 7.1441,
      "mean_token_accuracy": 0.13408377543091773,
      "num_tokens": 3255190.0,
      "step": 1290
    },
    {
      "entropy": 7.084112691879272,
      "epoch": 0.14945181765724178,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999869175830808,
      "loss": 7.171,
      "mean_token_accuracy": 0.13335621058940889,
      "num_tokens": 3268236.0,
      "step": 1295
    },
    {
      "entropy": 7.132741975784302,
      "epoch": 0.15002885170225044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999864688235666,
      "loss": 7.1832,
      "mean_token_accuracy": 0.13459478169679642,
      "num_tokens": 3281316.0,
      "step": 1300
    },
    {
      "entropy": 7.08173189163208,
      "epoch": 0.15060588574725908,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999860124967442,
      "loss": 7.0801,
      "mean_token_accuracy": 0.1370793603360653,
      "num_tokens": 3293517.0,
      "step": 1305
    },
    {
      "entropy": 7.009785556793213,
      "epoch": 0.15118291979226775,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999855486026291,
      "loss": 7.0642,
      "mean_token_accuracy": 0.14415186047554016,
      "num_tokens": 3306814.0,
      "step": 1310
    },
    {
      "entropy": 7.143966817855835,
      "epoch": 0.1517599538372764,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999850771412369,
      "loss": 7.1255,
      "mean_token_accuracy": 0.1338451124727726,
      "num_tokens": 3318722.0,
      "step": 1315
    },
    {
      "entropy": 6.947969913482666,
      "epoch": 0.15233698788228506,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004999845981125832,
      "loss": 7.0902,
      "mean_token_accuracy": 0.13792371451854707,
      "num_tokens": 3331212.0,
      "step": 1320
    },
    {
      "entropy": 7.112496328353882,
      "epoch": 0.1529140219272937,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999841115166844,
      "loss": 7.065,
      "mean_token_accuracy": 0.14001230895519257,
      "num_tokens": 3343999.0,
      "step": 1325
    },
    {
      "entropy": 6.955641222000122,
      "epoch": 0.15349105597230236,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999836173535568,
      "loss": 7.0087,
      "mean_token_accuracy": 0.14088786989450455,
      "num_tokens": 3356281.0,
      "step": 1330
    },
    {
      "entropy": 7.026442623138427,
      "epoch": 0.15406809001731103,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999831156232169,
      "loss": 7.0683,
      "mean_token_accuracy": 0.1388309493660927,
      "num_tokens": 3369370.0,
      "step": 1335
    },
    {
      "entropy": 7.115357398986816,
      "epoch": 0.15464512406231967,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999826063256818,
      "loss": 7.2104,
      "mean_token_accuracy": 0.13239274099469184,
      "num_tokens": 3382400.0,
      "step": 1340
    },
    {
      "entropy": 7.049170970916748,
      "epoch": 0.15522215810732834,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999820894609683,
      "loss": 7.0738,
      "mean_token_accuracy": 0.13679501563310623,
      "num_tokens": 3393666.0,
      "step": 1345
    },
    {
      "entropy": 7.102601671218872,
      "epoch": 0.15579919215233698,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004999815650290941,
      "loss": 7.0784,
      "mean_token_accuracy": 0.13827622160315514,
      "num_tokens": 3405920.0,
      "step": 1350
    },
    {
      "entropy": 6.988848543167114,
      "epoch": 0.15637622619734565,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999810330300766,
      "loss": 6.9811,
      "mean_token_accuracy": 0.14476096332073213,
      "num_tokens": 3417886.0,
      "step": 1355
    },
    {
      "entropy": 7.066180992126465,
      "epoch": 0.1569532602423543,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499980493463934,
      "loss": 7.0976,
      "mean_token_accuracy": 0.13426982685923577,
      "num_tokens": 3430626.0,
      "step": 1360
    },
    {
      "entropy": 7.005979776382446,
      "epoch": 0.15753029428736295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999799463306841,
      "loss": 6.9849,
      "mean_token_accuracy": 0.13666047602891923,
      "num_tokens": 3443248.0,
      "step": 1365
    },
    {
      "entropy": 7.070165061950684,
      "epoch": 0.15810732833237162,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999793916303455,
      "loss": 7.032,
      "mean_token_accuracy": 0.13812974020838736,
      "num_tokens": 3456163.0,
      "step": 1370
    },
    {
      "entropy": 6.949036455154419,
      "epoch": 0.15868436237738026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999788293629368,
      "loss": 7.0586,
      "mean_token_accuracy": 0.1321362428367138,
      "num_tokens": 3468862.0,
      "step": 1375
    },
    {
      "entropy": 7.017643308639526,
      "epoch": 0.15926139642238893,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499978259528477,
      "loss": 7.0618,
      "mean_token_accuracy": 0.14320022165775298,
      "num_tokens": 3481516.0,
      "step": 1380
    },
    {
      "entropy": 7.152157735824585,
      "epoch": 0.15983843046739757,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999776821269852,
      "loss": 7.1397,
      "mean_token_accuracy": 0.12845932096242904,
      "num_tokens": 3493063.0,
      "step": 1385
    },
    {
      "entropy": 6.983753442764282,
      "epoch": 0.16041546451240624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999770971584807,
      "loss": 6.9992,
      "mean_token_accuracy": 0.13282350823283195,
      "num_tokens": 3505460.0,
      "step": 1390
    },
    {
      "entropy": 7.0605611324310305,
      "epoch": 0.16099249855741488,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999765046229834,
      "loss": 7.1184,
      "mean_token_accuracy": 0.13035993799567222,
      "num_tokens": 3519232.0,
      "step": 1395
    },
    {
      "entropy": 6.989923524856567,
      "epoch": 0.16156953260242354,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499975904520513,
      "loss": 7.0311,
      "mean_token_accuracy": 0.14471797570586203,
      "num_tokens": 3531759.0,
      "step": 1400
    },
    {
      "entropy": 7.006714677810669,
      "epoch": 0.1621465666474322,
      "grad_norm": 0.921875,
      "learning_rate": 0.00049997529685109,
      "loss": 6.9624,
      "mean_token_accuracy": 0.14889438897371293,
      "num_tokens": 3543457.0,
      "step": 1405
    },
    {
      "entropy": 6.985404443740845,
      "epoch": 0.16272360069244085,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999746816147344,
      "loss": 7.1171,
      "mean_token_accuracy": 0.13007233440876007,
      "num_tokens": 3557036.0,
      "step": 1410
    },
    {
      "entropy": 7.086899042129517,
      "epoch": 0.16330063473744952,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999740588114673,
      "loss": 7.0671,
      "mean_token_accuracy": 0.1431369036436081,
      "num_tokens": 3569415.0,
      "step": 1415
    },
    {
      "entropy": 6.992696142196655,
      "epoch": 0.16387766878245816,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999734284413095,
      "loss": 7.0656,
      "mean_token_accuracy": 0.13436094596982,
      "num_tokens": 3581452.0,
      "step": 1420
    },
    {
      "entropy": 7.03533968925476,
      "epoch": 0.16445470282746683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499972790504282,
      "loss": 7.1012,
      "mean_token_accuracy": 0.1372412621974945,
      "num_tokens": 3593482.0,
      "step": 1425
    },
    {
      "entropy": 7.005078983306885,
      "epoch": 0.16503173687247547,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999721450004067,
      "loss": 7.0916,
      "mean_token_accuracy": 0.13415241464972497,
      "num_tokens": 3606062.0,
      "step": 1430
    },
    {
      "entropy": 7.060822534561157,
      "epoch": 0.16560877091748413,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999714919297049,
      "loss": 7.0473,
      "mean_token_accuracy": 0.1380315750837326,
      "num_tokens": 3619210.0,
      "step": 1435
    },
    {
      "entropy": 7.01508846282959,
      "epoch": 0.1661858049624928,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999708312921987,
      "loss": 7.0976,
      "mean_token_accuracy": 0.13783507272601128,
      "num_tokens": 3631952.0,
      "step": 1440
    },
    {
      "entropy": 7.05513277053833,
      "epoch": 0.16676283900750144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999701630879103,
      "loss": 6.9784,
      "mean_token_accuracy": 0.1367719940841198,
      "num_tokens": 3644864.0,
      "step": 1445
    },
    {
      "entropy": 6.986611795425415,
      "epoch": 0.1673398730525101,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999694873168623,
      "loss": 7.046,
      "mean_token_accuracy": 0.13879042416810988,
      "num_tokens": 3657504.0,
      "step": 1450
    },
    {
      "entropy": 7.017226314544677,
      "epoch": 0.16791690709751875,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999688039790773,
      "loss": 6.9585,
      "mean_token_accuracy": 0.14265901520848273,
      "num_tokens": 3670304.0,
      "step": 1455
    },
    {
      "entropy": 6.9451819896698,
      "epoch": 0.16849394114252741,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999681130745784,
      "loss": 6.9718,
      "mean_token_accuracy": 0.13715282902121545,
      "num_tokens": 3681503.0,
      "step": 1460
    },
    {
      "entropy": 6.938751745223999,
      "epoch": 0.16907097518753605,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999674146033888,
      "loss": 7.0007,
      "mean_token_accuracy": 0.13866478949785233,
      "num_tokens": 3693248.0,
      "step": 1465
    },
    {
      "entropy": 7.016212892532349,
      "epoch": 0.16964800923254472,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999667085655318,
      "loss": 7.035,
      "mean_token_accuracy": 0.1379549205303192,
      "num_tokens": 3705285.0,
      "step": 1470
    },
    {
      "entropy": 6.883904600143433,
      "epoch": 0.1702250432775534,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999659949610313,
      "loss": 7.0562,
      "mean_token_accuracy": 0.12795180603861808,
      "num_tokens": 3719025.0,
      "step": 1475
    },
    {
      "entropy": 7.103451251983643,
      "epoch": 0.17080207732256203,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999652737899114,
      "loss": 7.0306,
      "mean_token_accuracy": 0.13902753069996834,
      "num_tokens": 3732054.0,
      "step": 1480
    },
    {
      "entropy": 6.955500936508178,
      "epoch": 0.1713791113675707,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999645450521963,
      "loss": 7.0133,
      "mean_token_accuracy": 0.13680559545755386,
      "num_tokens": 3744080.0,
      "step": 1485
    },
    {
      "entropy": 6.9854591369628904,
      "epoch": 0.17195614541257934,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999638087479104,
      "loss": 7.047,
      "mean_token_accuracy": 0.13523856773972512,
      "num_tokens": 3756371.0,
      "step": 1490
    },
    {
      "entropy": 7.104809236526489,
      "epoch": 0.172533179457588,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999630648770786,
      "loss": 6.9673,
      "mean_token_accuracy": 0.1353852815926075,
      "num_tokens": 3769307.0,
      "step": 1495
    },
    {
      "entropy": 6.814951086044312,
      "epoch": 0.17311021350259664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999623134397257,
      "loss": 6.9706,
      "mean_token_accuracy": 0.14363576769828795,
      "num_tokens": 3782826.0,
      "step": 1500
    },
    {
      "entropy": 7.0104962348937985,
      "epoch": 0.1736872475476053,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999615544358774,
      "loss": 6.9846,
      "mean_token_accuracy": 0.1368210159242153,
      "num_tokens": 3794566.0,
      "step": 1505
    },
    {
      "entropy": 6.955519914627075,
      "epoch": 0.17426428159261395,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999607878655587,
      "loss": 7.0675,
      "mean_token_accuracy": 0.1298111744225025,
      "num_tokens": 3807058.0,
      "step": 1510
    },
    {
      "entropy": 7.028119516372681,
      "epoch": 0.17484131563762262,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999600137287958,
      "loss": 7.0107,
      "mean_token_accuracy": 0.14360469132661818,
      "num_tokens": 3819836.0,
      "step": 1515
    },
    {
      "entropy": 6.98626184463501,
      "epoch": 0.17541834968263129,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999592320256146,
      "loss": 7.0598,
      "mean_token_accuracy": 0.1317824549973011,
      "num_tokens": 3832081.0,
      "step": 1520
    },
    {
      "entropy": 7.0153602123260494,
      "epoch": 0.17599538372763993,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999584427560412,
      "loss": 6.9811,
      "mean_token_accuracy": 0.13875427916646005,
      "num_tokens": 3843279.0,
      "step": 1525
    },
    {
      "entropy": 6.946571159362793,
      "epoch": 0.1765724177726486,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999576459201024,
      "loss": 7.0461,
      "mean_token_accuracy": 0.14072771817445756,
      "num_tokens": 3855479.0,
      "step": 1530
    },
    {
      "entropy": 7.037504243850708,
      "epoch": 0.17714945181765723,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499956841517825,
      "loss": 7.0258,
      "mean_token_accuracy": 0.1374424070119858,
      "num_tokens": 3868111.0,
      "step": 1535
    },
    {
      "entropy": 6.900399398803711,
      "epoch": 0.1777264858626659,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499956029549236,
      "loss": 6.9516,
      "mean_token_accuracy": 0.14264189973473548,
      "num_tokens": 3880062.0,
      "step": 1540
    },
    {
      "entropy": 7.105960988998413,
      "epoch": 0.17830351990767454,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999552100143625,
      "loss": 7.0352,
      "mean_token_accuracy": 0.13576769903302194,
      "num_tokens": 3892293.0,
      "step": 1545
    },
    {
      "entropy": 6.912646913528443,
      "epoch": 0.1788805539526832,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999543829132324,
      "loss": 6.8992,
      "mean_token_accuracy": 0.1437755271792412,
      "num_tokens": 3904549.0,
      "step": 1550
    },
    {
      "entropy": 6.929834699630737,
      "epoch": 0.17945758799769188,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999535482458734,
      "loss": 6.953,
      "mean_token_accuracy": 0.14296042323112487,
      "num_tokens": 3916693.0,
      "step": 1555
    },
    {
      "entropy": 6.884093236923218,
      "epoch": 0.18003462204270052,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999527060123135,
      "loss": 6.9797,
      "mean_token_accuracy": 0.13235038220882417,
      "num_tokens": 3930488.0,
      "step": 1560
    },
    {
      "entropy": 6.985349941253662,
      "epoch": 0.18061165608770918,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999518562125811,
      "loss": 6.961,
      "mean_token_accuracy": 0.14040945768356322,
      "num_tokens": 3942353.0,
      "step": 1565
    },
    {
      "entropy": 6.977359771728516,
      "epoch": 0.18118869013271782,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999509988467047,
      "loss": 7.0072,
      "mean_token_accuracy": 0.13477388694882392,
      "num_tokens": 3953868.0,
      "step": 1570
    },
    {
      "entropy": 6.984608507156372,
      "epoch": 0.1817657241777265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999501339147132,
      "loss": 7.0248,
      "mean_token_accuracy": 0.13513074964284896,
      "num_tokens": 3965148.0,
      "step": 1575
    },
    {
      "entropy": 6.968512153625488,
      "epoch": 0.18234275822273513,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999492614166357,
      "loss": 6.9684,
      "mean_token_accuracy": 0.14738112092018127,
      "num_tokens": 3977323.0,
      "step": 1580
    },
    {
      "entropy": 6.937796115875244,
      "epoch": 0.1829197922677438,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999483813525014,
      "loss": 6.9445,
      "mean_token_accuracy": 0.1459215387701988,
      "num_tokens": 3990076.0,
      "step": 1585
    },
    {
      "entropy": 6.896608018875122,
      "epoch": 0.18349682631275246,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999474937223403,
      "loss": 6.959,
      "mean_token_accuracy": 0.1377126969397068,
      "num_tokens": 4001996.0,
      "step": 1590
    },
    {
      "entropy": 6.98504409790039,
      "epoch": 0.1840738603577611,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999465985261818,
      "loss": 7.0041,
      "mean_token_accuracy": 0.13642505407333375,
      "num_tokens": 4015095.0,
      "step": 1595
    },
    {
      "entropy": 6.90059175491333,
      "epoch": 0.18465089440276977,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999456957640562,
      "loss": 6.8935,
      "mean_token_accuracy": 0.1471956580877304,
      "num_tokens": 4028633.0,
      "step": 1600
    },
    {
      "entropy": 6.932298421859741,
      "epoch": 0.1852279284477784,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999447854359939,
      "loss": 6.9938,
      "mean_token_accuracy": 0.1394735284149647,
      "num_tokens": 4042265.0,
      "step": 1605
    },
    {
      "entropy": 6.876345634460449,
      "epoch": 0.18580496249278708,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999438675420255,
      "loss": 6.954,
      "mean_token_accuracy": 0.1372726485133171,
      "num_tokens": 4054467.0,
      "step": 1610
    },
    {
      "entropy": 6.95422887802124,
      "epoch": 0.18638199653779572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999429420821818,
      "loss": 6.951,
      "mean_token_accuracy": 0.1380107432603836,
      "num_tokens": 4068053.0,
      "step": 1615
    },
    {
      "entropy": 6.854258346557617,
      "epoch": 0.1869590305828044,
      "grad_norm": 1.15625,
      "learning_rate": 0.000499942009056494,
      "loss": 6.9318,
      "mean_token_accuracy": 0.1387524574995041,
      "num_tokens": 4080867.0,
      "step": 1620
    },
    {
      "entropy": 6.933802938461303,
      "epoch": 0.18753606462781305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999410684649935,
      "loss": 6.9481,
      "mean_token_accuracy": 0.14056655392050743,
      "num_tokens": 4093228.0,
      "step": 1625
    },
    {
      "entropy": 7.00413064956665,
      "epoch": 0.1881130986728217,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999401203077119,
      "loss": 6.9716,
      "mean_token_accuracy": 0.14066905975341798,
      "num_tokens": 4104929.0,
      "step": 1630
    },
    {
      "entropy": 6.8644571781158445,
      "epoch": 0.18869013271783036,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499939164584681,
      "loss": 6.9945,
      "mean_token_accuracy": 0.14416123032569886,
      "num_tokens": 4119445.0,
      "step": 1635
    },
    {
      "entropy": 7.075875329971313,
      "epoch": 0.189267166762839,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999382012959331,
      "loss": 6.8961,
      "mean_token_accuracy": 0.14873119294643403,
      "num_tokens": 4131558.0,
      "step": 1640
    },
    {
      "entropy": 6.876895570755005,
      "epoch": 0.18984420080784767,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999372304415005,
      "loss": 7.0549,
      "mean_token_accuracy": 0.12923122197389603,
      "num_tokens": 4144566.0,
      "step": 1645
    },
    {
      "entropy": 6.990751695632935,
      "epoch": 0.1904212348528563,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999362520214159,
      "loss": 6.9381,
      "mean_token_accuracy": 0.14834065288305281,
      "num_tokens": 4157295.0,
      "step": 1650
    },
    {
      "entropy": 6.910133790969849,
      "epoch": 0.19099826889786498,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999352660357122,
      "loss": 6.9455,
      "mean_token_accuracy": 0.13908031955361366,
      "num_tokens": 4170229.0,
      "step": 1655
    },
    {
      "entropy": 6.951679515838623,
      "epoch": 0.19157530294287364,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999342724844226,
      "loss": 6.9645,
      "mean_token_accuracy": 0.13757809549570083,
      "num_tokens": 4182906.0,
      "step": 1660
    },
    {
      "entropy": 6.961617517471313,
      "epoch": 0.19215233698788228,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999332713675804,
      "loss": 6.9304,
      "mean_token_accuracy": 0.14034477695822717,
      "num_tokens": 4195556.0,
      "step": 1665
    },
    {
      "entropy": 6.911066627502441,
      "epoch": 0.19272937103289095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999322626852193,
      "loss": 6.8961,
      "mean_token_accuracy": 0.1394181601703167,
      "num_tokens": 4207616.0,
      "step": 1670
    },
    {
      "entropy": 6.845373630523682,
      "epoch": 0.1933064050778996,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999312464373734,
      "loss": 6.8568,
      "mean_token_accuracy": 0.14717165976762772,
      "num_tokens": 4219299.0,
      "step": 1675
    },
    {
      "entropy": 6.886244630813598,
      "epoch": 0.19388343912290826,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999302226240767,
      "loss": 7.0009,
      "mean_token_accuracy": 0.13946698978543282,
      "num_tokens": 4232098.0,
      "step": 1680
    },
    {
      "entropy": 6.927741765975952,
      "epoch": 0.1944604731679169,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999291912453637,
      "loss": 6.8926,
      "mean_token_accuracy": 0.14187719076871871,
      "num_tokens": 4243879.0,
      "step": 1685
    },
    {
      "entropy": 6.919966363906861,
      "epoch": 0.19503750721292556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999281523012691,
      "loss": 6.9798,
      "mean_token_accuracy": 0.13289057463407516,
      "num_tokens": 4256507.0,
      "step": 1690
    },
    {
      "entropy": 7.115394401550293,
      "epoch": 0.1956145412579342,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999271057918278,
      "loss": 7.0498,
      "mean_token_accuracy": 0.13213248774409295,
      "num_tokens": 4270763.0,
      "step": 1695
    },
    {
      "entropy": 6.83512053489685,
      "epoch": 0.19619157530294287,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999260517170751,
      "loss": 6.9311,
      "mean_token_accuracy": 0.13900210931897164,
      "num_tokens": 4284380.0,
      "step": 1700
    },
    {
      "entropy": 6.999711275100708,
      "epoch": 0.19676860934795154,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999249900770463,
      "loss": 7.0283,
      "mean_token_accuracy": 0.13723405376076697,
      "num_tokens": 4297812.0,
      "step": 1705
    },
    {
      "entropy": 6.927718687057495,
      "epoch": 0.19734564339296018,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999239208717772,
      "loss": 7.0,
      "mean_token_accuracy": 0.14325511828064919,
      "num_tokens": 4310374.0,
      "step": 1710
    },
    {
      "entropy": 6.975845861434936,
      "epoch": 0.19792267743796885,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999228441013037,
      "loss": 6.966,
      "mean_token_accuracy": 0.1380312144756317,
      "num_tokens": 4322330.0,
      "step": 1715
    },
    {
      "entropy": 6.951593542098999,
      "epoch": 0.1984997114829775,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499921759765662,
      "loss": 6.9519,
      "mean_token_accuracy": 0.1390042595565319,
      "num_tokens": 4335204.0,
      "step": 1720
    },
    {
      "entropy": 6.7771580696105955,
      "epoch": 0.19907674552798615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999206678648888,
      "loss": 6.8396,
      "mean_token_accuracy": 0.14716555774211884,
      "num_tokens": 4347339.0,
      "step": 1725
    },
    {
      "entropy": 6.960784196853638,
      "epoch": 0.1996537795729948,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004999195683990206,
      "loss": 6.8991,
      "mean_token_accuracy": 0.14737813025712967,
      "num_tokens": 4360273.0,
      "step": 1730
    },
    {
      "entropy": 6.812373208999634,
      "epoch": 0.20023081361800346,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999184613680945,
      "loss": 6.9116,
      "mean_token_accuracy": 0.14005344808101655,
      "num_tokens": 4373424.0,
      "step": 1735
    },
    {
      "entropy": 6.970333909988403,
      "epoch": 0.20080784766301213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999173467721476,
      "loss": 6.8623,
      "mean_token_accuracy": 0.14696093946695327,
      "num_tokens": 4386125.0,
      "step": 1740
    },
    {
      "entropy": 6.900853633880615,
      "epoch": 0.20138488170802077,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999162246112175,
      "loss": 6.9299,
      "mean_token_accuracy": 0.1397725947201252,
      "num_tokens": 4399214.0,
      "step": 1745
    },
    {
      "entropy": 6.842698955535889,
      "epoch": 0.20196191575302944,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999150948853419,
      "loss": 6.8108,
      "mean_token_accuracy": 0.14288675636053086,
      "num_tokens": 4411282.0,
      "step": 1750
    },
    {
      "entropy": 6.895815658569336,
      "epoch": 0.20253894979803808,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999139575945587,
      "loss": 6.904,
      "mean_token_accuracy": 0.14115357473492623,
      "num_tokens": 4423448.0,
      "step": 1755
    },
    {
      "entropy": 6.8710291385650635,
      "epoch": 0.20311598384304674,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999128127389065,
      "loss": 6.8318,
      "mean_token_accuracy": 0.14915311187505723,
      "num_tokens": 4436498.0,
      "step": 1760
    },
    {
      "entropy": 6.864768648147583,
      "epoch": 0.20369301788805538,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999116603184233,
      "loss": 6.8424,
      "mean_token_accuracy": 0.14298070967197418,
      "num_tokens": 4448641.0,
      "step": 1765
    },
    {
      "entropy": 6.82587513923645,
      "epoch": 0.20427005193306405,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999105003331482,
      "loss": 6.8389,
      "mean_token_accuracy": 0.14010173305869103,
      "num_tokens": 4460539.0,
      "step": 1770
    },
    {
      "entropy": 6.890678787231446,
      "epoch": 0.20484708597807272,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999093327831202,
      "loss": 6.9094,
      "mean_token_accuracy": 0.1397670365869999,
      "num_tokens": 4472487.0,
      "step": 1775
    },
    {
      "entropy": 6.8611537456512455,
      "epoch": 0.20542412002308136,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999081576683784,
      "loss": 6.8784,
      "mean_token_accuracy": 0.14253682196140288,
      "num_tokens": 4485164.0,
      "step": 1780
    },
    {
      "entropy": 6.964418983459472,
      "epoch": 0.20600115406809003,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004999069749889626,
      "loss": 6.9038,
      "mean_token_accuracy": 0.14524734169244766,
      "num_tokens": 4497708.0,
      "step": 1785
    },
    {
      "entropy": 6.834679794311524,
      "epoch": 0.20657818811309867,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999057847449123,
      "loss": 6.9527,
      "mean_token_accuracy": 0.13614820539951325,
      "num_tokens": 4510648.0,
      "step": 1790
    },
    {
      "entropy": 6.901812505722046,
      "epoch": 0.20715522215810733,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999045869362678,
      "loss": 6.8605,
      "mean_token_accuracy": 0.14650894105434417,
      "num_tokens": 4522699.0,
      "step": 1795
    },
    {
      "entropy": 6.82709813117981,
      "epoch": 0.20773225620311597,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499903381563069,
      "loss": 6.921,
      "mean_token_accuracy": 0.14181671142578126,
      "num_tokens": 4536060.0,
      "step": 1800
    },
    {
      "entropy": 6.863535785675049,
      "epoch": 0.20830929024812464,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999021686253568,
      "loss": 6.8218,
      "mean_token_accuracy": 0.14091982021927835,
      "num_tokens": 4549876.0,
      "step": 1805
    },
    {
      "entropy": 6.900723028182983,
      "epoch": 0.2088863242931333,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999009481231719,
      "loss": 6.872,
      "mean_token_accuracy": 0.1442640721797943,
      "num_tokens": 4563638.0,
      "step": 1810
    },
    {
      "entropy": 6.806870174407959,
      "epoch": 0.20946335833814195,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998997200565553,
      "loss": 6.8087,
      "mean_token_accuracy": 0.14233467727899551,
      "num_tokens": 4575396.0,
      "step": 1815
    },
    {
      "entropy": 6.781218814849853,
      "epoch": 0.21004039238315061,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998984844255483,
      "loss": 6.8239,
      "mean_token_accuracy": 0.14259946793317796,
      "num_tokens": 4587261.0,
      "step": 1820
    },
    {
      "entropy": 6.937310457229614,
      "epoch": 0.21061742642815925,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004998972412301925,
      "loss": 6.9373,
      "mean_token_accuracy": 0.1397809252142906,
      "num_tokens": 4600086.0,
      "step": 1825
    },
    {
      "entropy": 6.912576341629029,
      "epoch": 0.21119446047316792,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998959904705297,
      "loss": 6.9093,
      "mean_token_accuracy": 0.1342283457517624,
      "num_tokens": 4612320.0,
      "step": 1830
    },
    {
      "entropy": 6.850874280929565,
      "epoch": 0.21177149451817656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998947321466021,
      "loss": 6.8522,
      "mean_token_accuracy": 0.14495468139648438,
      "num_tokens": 4623545.0,
      "step": 1835
    },
    {
      "entropy": 6.87814154624939,
      "epoch": 0.21234852856318523,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998934662584518,
      "loss": 6.7709,
      "mean_token_accuracy": 0.15582974776625633,
      "num_tokens": 4637223.0,
      "step": 1840
    },
    {
      "entropy": 6.624226713180542,
      "epoch": 0.2129255626081939,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998921928061214,
      "loss": 6.8385,
      "mean_token_accuracy": 0.14391470998525618,
      "num_tokens": 4649748.0,
      "step": 1845
    },
    {
      "entropy": 6.985252904891968,
      "epoch": 0.21350259665320254,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998909117896539,
      "loss": 6.8723,
      "mean_token_accuracy": 0.14348524510860444,
      "num_tokens": 4662441.0,
      "step": 1850
    },
    {
      "entropy": 6.706543636322022,
      "epoch": 0.2140796306982112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998896232090922,
      "loss": 6.8359,
      "mean_token_accuracy": 0.15377265065908433,
      "num_tokens": 4675546.0,
      "step": 1855
    },
    {
      "entropy": 7.046459770202636,
      "epoch": 0.21465666474321984,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998883270644798,
      "loss": 6.9101,
      "mean_token_accuracy": 0.1373351812362671,
      "num_tokens": 4687601.0,
      "step": 1860
    },
    {
      "entropy": 6.7164732933044435,
      "epoch": 0.2152336987882285,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004998870233558602,
      "loss": 6.7657,
      "mean_token_accuracy": 0.15242065340280533,
      "num_tokens": 4699336.0,
      "step": 1865
    },
    {
      "entropy": 6.901521301269531,
      "epoch": 0.21581073283323715,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998857120832774,
      "loss": 6.9738,
      "mean_token_accuracy": 0.1383403241634369,
      "num_tokens": 4714312.0,
      "step": 1870
    },
    {
      "entropy": 6.770059871673584,
      "epoch": 0.21638776687824582,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998843932467751,
      "loss": 6.6833,
      "mean_token_accuracy": 0.15454858019948006,
      "num_tokens": 4726436.0,
      "step": 1875
    },
    {
      "entropy": 6.858100318908692,
      "epoch": 0.21696480092325446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998830668463982,
      "loss": 6.9327,
      "mean_token_accuracy": 0.13296440094709397,
      "num_tokens": 4739821.0,
      "step": 1880
    },
    {
      "entropy": 6.93988528251648,
      "epoch": 0.21754183496826313,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998817328821909,
      "loss": 6.9047,
      "mean_token_accuracy": 0.1361124962568283,
      "num_tokens": 4752701.0,
      "step": 1885
    },
    {
      "entropy": 6.844181823730469,
      "epoch": 0.2181188690132718,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998803913541983,
      "loss": 6.9331,
      "mean_token_accuracy": 0.1368408054113388,
      "num_tokens": 4766755.0,
      "step": 1890
    },
    {
      "entropy": 6.866988563537598,
      "epoch": 0.21869590305828043,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998790422624654,
      "loss": 6.8454,
      "mean_token_accuracy": 0.14372632652521133,
      "num_tokens": 4780201.0,
      "step": 1895
    },
    {
      "entropy": 6.863262557983399,
      "epoch": 0.2192729371032891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998776856070376,
      "loss": 6.8284,
      "mean_token_accuracy": 0.14132185727357865,
      "num_tokens": 4792488.0,
      "step": 1900
    },
    {
      "entropy": 6.798328876495361,
      "epoch": 0.21984997114829774,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998763213879606,
      "loss": 6.7934,
      "mean_token_accuracy": 0.14561547935009003,
      "num_tokens": 4805154.0,
      "step": 1905
    },
    {
      "entropy": 6.830894565582275,
      "epoch": 0.2204270051933064,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998749496052803,
      "loss": 6.852,
      "mean_token_accuracy": 0.14222206994891168,
      "num_tokens": 4817947.0,
      "step": 1910
    },
    {
      "entropy": 6.834425973892212,
      "epoch": 0.22100403923831505,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998735702590426,
      "loss": 6.8408,
      "mean_token_accuracy": 0.14236815497279168,
      "num_tokens": 4830296.0,
      "step": 1915
    },
    {
      "entropy": 6.832809066772461,
      "epoch": 0.22158107328332372,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998721833492942,
      "loss": 6.8318,
      "mean_token_accuracy": 0.13888390138745307,
      "num_tokens": 4841210.0,
      "step": 1920
    },
    {
      "entropy": 6.879108953475952,
      "epoch": 0.22215810732833238,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998707888760816,
      "loss": 6.9203,
      "mean_token_accuracy": 0.13413121029734612,
      "num_tokens": 4854406.0,
      "step": 1925
    },
    {
      "entropy": 6.848855876922608,
      "epoch": 0.22273514137334102,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998693868394516,
      "loss": 6.8873,
      "mean_token_accuracy": 0.1403192825615406,
      "num_tokens": 4868156.0,
      "step": 1930
    },
    {
      "entropy": 6.837898826599121,
      "epoch": 0.2233121754183497,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998679772394516,
      "loss": 6.814,
      "mean_token_accuracy": 0.14790826439857482,
      "num_tokens": 4880314.0,
      "step": 1935
    },
    {
      "entropy": 6.8341789722442625,
      "epoch": 0.22388920946335833,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998665600761289,
      "loss": 6.8263,
      "mean_token_accuracy": 0.14194524735212327,
      "num_tokens": 4892652.0,
      "step": 1940
    },
    {
      "entropy": 6.928193855285644,
      "epoch": 0.224466243508367,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499865135349531,
      "loss": 6.9246,
      "mean_token_accuracy": 0.13543495833873748,
      "num_tokens": 4905791.0,
      "step": 1945
    },
    {
      "entropy": 6.824593591690063,
      "epoch": 0.22504327755337564,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998637030597061,
      "loss": 6.8886,
      "mean_token_accuracy": 0.13973383381962776,
      "num_tokens": 4917359.0,
      "step": 1950
    },
    {
      "entropy": 6.9390894889831545,
      "epoch": 0.2256203115983843,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998622632067022,
      "loss": 6.8356,
      "mean_token_accuracy": 0.14431015402078629,
      "num_tokens": 4928762.0,
      "step": 1955
    },
    {
      "entropy": 6.8002971649169925,
      "epoch": 0.22619734564339297,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998608157905678,
      "loss": 6.8618,
      "mean_token_accuracy": 0.1413520596921444,
      "num_tokens": 4940879.0,
      "step": 1960
    },
    {
      "entropy": 6.9403890609741214,
      "epoch": 0.2267743796884016,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998593608113515,
      "loss": 6.8601,
      "mean_token_accuracy": 0.14300046414136885,
      "num_tokens": 4952979.0,
      "step": 1965
    },
    {
      "entropy": 6.7911515712738035,
      "epoch": 0.22735141373341028,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998578982691024,
      "loss": 6.7712,
      "mean_token_accuracy": 0.15122427493333818,
      "num_tokens": 4965465.0,
      "step": 1970
    },
    {
      "entropy": 6.746047019958496,
      "epoch": 0.22792844777841892,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998564281638694,
      "loss": 6.848,
      "mean_token_accuracy": 0.14356525242328644,
      "num_tokens": 4977960.0,
      "step": 1975
    },
    {
      "entropy": 6.9141035079956055,
      "epoch": 0.2285054818234276,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004998549504957023,
      "loss": 6.8461,
      "mean_token_accuracy": 0.14194221496582032,
      "num_tokens": 4990732.0,
      "step": 1980
    },
    {
      "entropy": 6.763252067565918,
      "epoch": 0.22908251586843623,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998534652646506,
      "loss": 6.7825,
      "mean_token_accuracy": 0.14249248653650284,
      "num_tokens": 5002482.0,
      "step": 1985
    },
    {
      "entropy": 6.819217205047607,
      "epoch": 0.2296595499134449,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998519724707642,
      "loss": 6.7532,
      "mean_token_accuracy": 0.14924108535051345,
      "num_tokens": 5014932.0,
      "step": 1990
    },
    {
      "entropy": 6.760645580291748,
      "epoch": 0.23023658395845356,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998504721140934,
      "loss": 6.8991,
      "mean_token_accuracy": 0.13749758303165435,
      "num_tokens": 5027661.0,
      "step": 1995
    },
    {
      "entropy": 6.762338304519654,
      "epoch": 0.2308136180034622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998489641946887,
      "loss": 6.6503,
      "mean_token_accuracy": 0.1527012199163437,
      "num_tokens": 5040657.0,
      "step": 2000
    },
    {
      "entropy": 6.837014961242676,
      "epoch": 0.23139065204847087,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998474487126009,
      "loss": 6.932,
      "mean_token_accuracy": 0.1339501306414604,
      "num_tokens": 5054408.0,
      "step": 2005
    },
    {
      "entropy": 6.857686519622803,
      "epoch": 0.2319676860934795,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998459256678806,
      "loss": 6.817,
      "mean_token_accuracy": 0.14147326201200486,
      "num_tokens": 5065499.0,
      "step": 2010
    },
    {
      "entropy": 6.737187385559082,
      "epoch": 0.23254472013848818,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998443950605796,
      "loss": 6.7662,
      "mean_token_accuracy": 0.15133574232459068,
      "num_tokens": 5077009.0,
      "step": 2015
    },
    {
      "entropy": 6.883497905731201,
      "epoch": 0.23312175418349682,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998428568907488,
      "loss": 6.9356,
      "mean_token_accuracy": 0.1379622347652912,
      "num_tokens": 5089462.0,
      "step": 2020
    },
    {
      "entropy": 6.845494270324707,
      "epoch": 0.23369878822850548,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998413111584403,
      "loss": 6.849,
      "mean_token_accuracy": 0.14125331565737725,
      "num_tokens": 5101816.0,
      "step": 2025
    },
    {
      "entropy": 6.83185510635376,
      "epoch": 0.23427582227351415,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998397578637059,
      "loss": 6.8147,
      "mean_token_accuracy": 0.13760584592819214,
      "num_tokens": 5115085.0,
      "step": 2030
    },
    {
      "entropy": 6.816512393951416,
      "epoch": 0.2348528563185228,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499838197006598,
      "loss": 6.8383,
      "mean_token_accuracy": 0.13956220373511313,
      "num_tokens": 5127705.0,
      "step": 2035
    },
    {
      "entropy": 6.756067323684692,
      "epoch": 0.23542989036353146,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000499836628587169,
      "loss": 6.8004,
      "mean_token_accuracy": 0.14978178068995476,
      "num_tokens": 5140509.0,
      "step": 2040
    },
    {
      "entropy": 6.873353624343872,
      "epoch": 0.2360069244085401,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998350526054716,
      "loss": 6.8414,
      "mean_token_accuracy": 0.13831417188048362,
      "num_tokens": 5153538.0,
      "step": 2045
    },
    {
      "entropy": 6.819164085388183,
      "epoch": 0.23658395845354876,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998334690615591,
      "loss": 6.7717,
      "mean_token_accuracy": 0.14658078253269197,
      "num_tokens": 5166181.0,
      "step": 2050
    },
    {
      "entropy": 6.798212099075317,
      "epoch": 0.2371609924985574,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998318779554844,
      "loss": 6.799,
      "mean_token_accuracy": 0.14503582268953324,
      "num_tokens": 5177209.0,
      "step": 2055
    },
    {
      "entropy": 6.752374887466431,
      "epoch": 0.23773802654356607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998302792873012,
      "loss": 6.8182,
      "mean_token_accuracy": 0.13951095193624496,
      "num_tokens": 5189786.0,
      "step": 2060
    },
    {
      "entropy": 6.845754146575928,
      "epoch": 0.23831506058857474,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998286730570631,
      "loss": 6.7955,
      "mean_token_accuracy": 0.14597882628440856,
      "num_tokens": 5202629.0,
      "step": 2065
    },
    {
      "entropy": 6.795567512512207,
      "epoch": 0.23889209463358338,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998270592648245,
      "loss": 6.8691,
      "mean_token_accuracy": 0.13765914365649223,
      "num_tokens": 5215339.0,
      "step": 2070
    },
    {
      "entropy": 6.792061185836792,
      "epoch": 0.23946912867859205,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004998254379106394,
      "loss": 6.7339,
      "mean_token_accuracy": 0.14747889786958696,
      "num_tokens": 5227666.0,
      "step": 2075
    },
    {
      "entropy": 6.6639073371887205,
      "epoch": 0.2400461627236007,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998238089945622,
      "loss": 6.7198,
      "mean_token_accuracy": 0.14811492711305618,
      "num_tokens": 5238200.0,
      "step": 2080
    },
    {
      "entropy": 6.769328355789185,
      "epoch": 0.24062319676860935,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998221725166479,
      "loss": 6.6725,
      "mean_token_accuracy": 0.14989647567272185,
      "num_tokens": 5250023.0,
      "step": 2085
    },
    {
      "entropy": 6.771991300582886,
      "epoch": 0.241200230813618,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998205284769516,
      "loss": 6.8616,
      "mean_token_accuracy": 0.13556723967194556,
      "num_tokens": 5262634.0,
      "step": 2090
    },
    {
      "entropy": 6.898429107666016,
      "epoch": 0.24177726485862666,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998188768755285,
      "loss": 6.8602,
      "mean_token_accuracy": 0.13672183454036713,
      "num_tokens": 5275739.0,
      "step": 2095
    },
    {
      "entropy": 6.76414852142334,
      "epoch": 0.2423542989036353,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998172177124341,
      "loss": 6.8172,
      "mean_token_accuracy": 0.14185626655817032,
      "num_tokens": 5289223.0,
      "step": 2100
    },
    {
      "entropy": 6.843665933609008,
      "epoch": 0.24293133294864397,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998155509877242,
      "loss": 6.81,
      "mean_token_accuracy": 0.1396335631608963,
      "num_tokens": 5301869.0,
      "step": 2105
    },
    {
      "entropy": 6.808076477050781,
      "epoch": 0.24350836699365264,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499813876701455,
      "loss": 6.8278,
      "mean_token_accuracy": 0.13811877146363258,
      "num_tokens": 5314999.0,
      "step": 2110
    },
    {
      "entropy": 6.79498610496521,
      "epoch": 0.24408540103866128,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004998121948536828,
      "loss": 6.845,
      "mean_token_accuracy": 0.14215158969163894,
      "num_tokens": 5328247.0,
      "step": 2115
    },
    {
      "entropy": 6.880921792984009,
      "epoch": 0.24466243508366994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998105054444639,
      "loss": 6.8275,
      "mean_token_accuracy": 0.13474374860525132,
      "num_tokens": 5341054.0,
      "step": 2120
    },
    {
      "entropy": 6.741239166259765,
      "epoch": 0.24523946912867858,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998088084738555,
      "loss": 6.8547,
      "mean_token_accuracy": 0.14400363937020302,
      "num_tokens": 5353857.0,
      "step": 2125
    },
    {
      "entropy": 6.843392419815063,
      "epoch": 0.24581650317368725,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998071039419144,
      "loss": 6.7824,
      "mean_token_accuracy": 0.13821006268262864,
      "num_tokens": 5366072.0,
      "step": 2130
    },
    {
      "entropy": 6.719191312789917,
      "epoch": 0.2463935372186959,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499805391848698,
      "loss": 6.7717,
      "mean_token_accuracy": 0.14614173099398614,
      "num_tokens": 5379613.0,
      "step": 2135
    },
    {
      "entropy": 6.86360387802124,
      "epoch": 0.24697057126370456,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499803672194264,
      "loss": 6.811,
      "mean_token_accuracy": 0.14077582731842994,
      "num_tokens": 5392394.0,
      "step": 2140
    },
    {
      "entropy": 6.659087896347046,
      "epoch": 0.24754760530871323,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998019449786701,
      "loss": 6.7942,
      "mean_token_accuracy": 0.14860865324735642,
      "num_tokens": 5404127.0,
      "step": 2145
    },
    {
      "entropy": 6.989247369766235,
      "epoch": 0.24812463935372187,
      "grad_norm": 0.875,
      "learning_rate": 0.0004998002102019744,
      "loss": 6.8993,
      "mean_token_accuracy": 0.13390290439128877,
      "num_tokens": 5417317.0,
      "step": 2150
    },
    {
      "entropy": 6.728624773025513,
      "epoch": 0.24870167339873053,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997984678642354,
      "loss": 6.7489,
      "mean_token_accuracy": 0.14663361757993698,
      "num_tokens": 5429991.0,
      "step": 2155
    },
    {
      "entropy": 6.749109506607056,
      "epoch": 0.24927870744373917,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997967179655115,
      "loss": 6.786,
      "mean_token_accuracy": 0.1380963869392872,
      "num_tokens": 5442661.0,
      "step": 2160
    },
    {
      "entropy": 6.798228549957275,
      "epoch": 0.24985574148874784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997949605058617,
      "loss": 6.8134,
      "mean_token_accuracy": 0.14670575857162477,
      "num_tokens": 5455876.0,
      "step": 2165
    },
    {
      "entropy": 6.707323455810547,
      "epoch": 0.2504327755337565,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997931954853451,
      "loss": 6.7693,
      "mean_token_accuracy": 0.14578518271446228,
      "num_tokens": 5468151.0,
      "step": 2170
    },
    {
      "entropy": 6.745931768417359,
      "epoch": 0.2510098095787651,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499791422904021,
      "loss": 6.8043,
      "mean_token_accuracy": 0.15150473788380622,
      "num_tokens": 5481327.0,
      "step": 2175
    },
    {
      "entropy": 6.735119152069092,
      "epoch": 0.2515868436237738,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997896427619491,
      "loss": 6.7461,
      "mean_token_accuracy": 0.1452285200357437,
      "num_tokens": 5493794.0,
      "step": 2180
    },
    {
      "entropy": 6.7988443851470945,
      "epoch": 0.25216387766878245,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997878550591892,
      "loss": 6.7754,
      "mean_token_accuracy": 0.1396859109401703,
      "num_tokens": 5506831.0,
      "step": 2185
    },
    {
      "entropy": 6.80522894859314,
      "epoch": 0.2527409117137911,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004997860597958013,
      "loss": 6.6666,
      "mean_token_accuracy": 0.1507306069135666,
      "num_tokens": 5518988.0,
      "step": 2190
    },
    {
      "entropy": 6.527280187606811,
      "epoch": 0.2533179457587998,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004997842569718461,
      "loss": 6.6262,
      "mean_token_accuracy": 0.14936918169260024,
      "num_tokens": 5530757.0,
      "step": 2195
    },
    {
      "entropy": 6.788536119461059,
      "epoch": 0.25389497980380843,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499782446587384,
      "loss": 6.7494,
      "mean_token_accuracy": 0.14626611173152923,
      "num_tokens": 5542819.0,
      "step": 2200
    },
    {
      "entropy": 6.713362503051758,
      "epoch": 0.25447201384881707,
      "grad_norm": 1.203125,
      "learning_rate": 0.000499780628642476,
      "loss": 6.7897,
      "mean_token_accuracy": 0.14051934331655502,
      "num_tokens": 5555797.0,
      "step": 2205
    },
    {
      "entropy": 6.766080236434936,
      "epoch": 0.2550490478938257,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997788031371834,
      "loss": 6.8895,
      "mean_token_accuracy": 0.13041166514158248,
      "num_tokens": 5570086.0,
      "step": 2210
    },
    {
      "entropy": 6.788276147842407,
      "epoch": 0.2556260819388344,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997769700715672,
      "loss": 6.6845,
      "mean_token_accuracy": 0.1493205651640892,
      "num_tokens": 5582182.0,
      "step": 2215
    },
    {
      "entropy": 6.735091209411621,
      "epoch": 0.25620311598384304,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997751294456892,
      "loss": 6.7832,
      "mean_token_accuracy": 0.14482725113630296,
      "num_tokens": 5595395.0,
      "step": 2220
    },
    {
      "entropy": 6.795186614990234,
      "epoch": 0.2567801500288517,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004997732812596114,
      "loss": 6.7803,
      "mean_token_accuracy": 0.14214650616049768,
      "num_tokens": 5608555.0,
      "step": 2225
    },
    {
      "entropy": 6.733956241607666,
      "epoch": 0.2573571840738604,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997714255133961,
      "loss": 6.753,
      "mean_token_accuracy": 0.1399676337838173,
      "num_tokens": 5620933.0,
      "step": 2230
    },
    {
      "entropy": 6.807656335830688,
      "epoch": 0.257934218118869,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997695622071054,
      "loss": 6.7964,
      "mean_token_accuracy": 0.14257583171129226,
      "num_tokens": 5634061.0,
      "step": 2235
    },
    {
      "entropy": 6.763797903060913,
      "epoch": 0.25851125216387766,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997676913408021,
      "loss": 6.7431,
      "mean_token_accuracy": 0.14566465988755226,
      "num_tokens": 5646617.0,
      "step": 2240
    },
    {
      "entropy": 6.783767938613892,
      "epoch": 0.2590882862088863,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997658129145493,
      "loss": 6.8348,
      "mean_token_accuracy": 0.13886496499180795,
      "num_tokens": 5659361.0,
      "step": 2245
    },
    {
      "entropy": 6.7600174903869625,
      "epoch": 0.259665320253895,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997639269284099,
      "loss": 6.6769,
      "mean_token_accuracy": 0.1541120558977127,
      "num_tokens": 5670591.0,
      "step": 2250
    },
    {
      "entropy": 6.7302642345428465,
      "epoch": 0.26024235429890363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997620333824476,
      "loss": 6.7974,
      "mean_token_accuracy": 0.1425503261387348,
      "num_tokens": 5683211.0,
      "step": 2255
    },
    {
      "entropy": 6.790132999420166,
      "epoch": 0.2608193883439123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997601322767258,
      "loss": 6.7847,
      "mean_token_accuracy": 0.14356143027544022,
      "num_tokens": 5695654.0,
      "step": 2260
    },
    {
      "entropy": 6.707987117767334,
      "epoch": 0.26139642238892097,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997582236113087,
      "loss": 6.6929,
      "mean_token_accuracy": 0.14759473353624344,
      "num_tokens": 5707776.0,
      "step": 2265
    },
    {
      "entropy": 6.729296684265137,
      "epoch": 0.2619734564339296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997563073862603,
      "loss": 6.7372,
      "mean_token_accuracy": 0.1442895159125328,
      "num_tokens": 5719895.0,
      "step": 2270
    },
    {
      "entropy": 6.8028727054595945,
      "epoch": 0.26255049047893825,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997543836016453,
      "loss": 6.7501,
      "mean_token_accuracy": 0.15161172673106194,
      "num_tokens": 5733280.0,
      "step": 2275
    },
    {
      "entropy": 6.734392881393433,
      "epoch": 0.2631275245239469,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997524522575281,
      "loss": 6.7993,
      "mean_token_accuracy": 0.14286206290125847,
      "num_tokens": 5746563.0,
      "step": 2280
    },
    {
      "entropy": 6.888967847824096,
      "epoch": 0.2637045585689556,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997505133539738,
      "loss": 6.8586,
      "mean_token_accuracy": 0.13801560774445534,
      "num_tokens": 5760735.0,
      "step": 2285
    },
    {
      "entropy": 6.7270537376403805,
      "epoch": 0.2642815926139642,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997485668910476,
      "loss": 6.7564,
      "mean_token_accuracy": 0.13578541129827498,
      "num_tokens": 5774304.0,
      "step": 2290
    },
    {
      "entropy": 6.6204423904418945,
      "epoch": 0.26485862665897286,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997466128688151,
      "loss": 6.6497,
      "mean_token_accuracy": 0.15090691298246384,
      "num_tokens": 5786045.0,
      "step": 2295
    },
    {
      "entropy": 6.754159736633301,
      "epoch": 0.26543566070398156,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499744651287342,
      "loss": 6.7144,
      "mean_token_accuracy": 0.1462336763739586,
      "num_tokens": 5798126.0,
      "step": 2300
    },
    {
      "entropy": 6.746666812896729,
      "epoch": 0.2660126947489902,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499742682146694,
      "loss": 6.7266,
      "mean_token_accuracy": 0.15180645808577536,
      "num_tokens": 5809778.0,
      "step": 2305
    },
    {
      "entropy": 6.795130205154419,
      "epoch": 0.26658972879399884,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004997407054469377,
      "loss": 6.8156,
      "mean_token_accuracy": 0.14033974930644036,
      "num_tokens": 5821902.0,
      "step": 2310
    },
    {
      "entropy": 6.782390499114991,
      "epoch": 0.2671667628390075,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997387211881392,
      "loss": 6.7574,
      "mean_token_accuracy": 0.1411186173558235,
      "num_tokens": 5835162.0,
      "step": 2315
    },
    {
      "entropy": 6.764913749694824,
      "epoch": 0.26774379688401617,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004997367293703656,
      "loss": 6.7446,
      "mean_token_accuracy": 0.1470859795808792,
      "num_tokens": 5847708.0,
      "step": 2320
    },
    {
      "entropy": 6.743312931060791,
      "epoch": 0.2683208309290248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997347299936837,
      "loss": 6.7407,
      "mean_token_accuracy": 0.14818446189165116,
      "num_tokens": 5860983.0,
      "step": 2325
    },
    {
      "entropy": 6.69023756980896,
      "epoch": 0.26889786497403345,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997327230581608,
      "loss": 6.7176,
      "mean_token_accuracy": 0.1412323720753193,
      "num_tokens": 5873878.0,
      "step": 2330
    },
    {
      "entropy": 6.642313623428345,
      "epoch": 0.26947489901904215,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997307085638644,
      "loss": 6.7315,
      "mean_token_accuracy": 0.14647015184164047,
      "num_tokens": 5886577.0,
      "step": 2335
    },
    {
      "entropy": 6.8489755153656,
      "epoch": 0.2700519330640508,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004997286865108621,
      "loss": 6.7482,
      "mean_token_accuracy": 0.13658007383346557,
      "num_tokens": 5900373.0,
      "step": 2340
    },
    {
      "entropy": 6.667910242080689,
      "epoch": 0.2706289671090594,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997266568992222,
      "loss": 6.7465,
      "mean_token_accuracy": 0.14726608991622925,
      "num_tokens": 5913038.0,
      "step": 2345
    },
    {
      "entropy": 6.69633846282959,
      "epoch": 0.27120600115406807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997246197290128,
      "loss": 6.6931,
      "mean_token_accuracy": 0.14734635949134828,
      "num_tokens": 5926096.0,
      "step": 2350
    },
    {
      "entropy": 6.720061302185059,
      "epoch": 0.27178303519907676,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997225750003024,
      "loss": 6.7071,
      "mean_token_accuracy": 0.1467483162879944,
      "num_tokens": 5937992.0,
      "step": 2355
    },
    {
      "entropy": 6.702146768569946,
      "epoch": 0.2723600692440854,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997205227131599,
      "loss": 6.7227,
      "mean_token_accuracy": 0.15519675761461257,
      "num_tokens": 5949685.0,
      "step": 2360
    },
    {
      "entropy": 6.744102334976196,
      "epoch": 0.27293710328909404,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997184628676542,
      "loss": 6.7951,
      "mean_token_accuracy": 0.14176806733012198,
      "num_tokens": 5962751.0,
      "step": 2365
    },
    {
      "entropy": 6.628593158721924,
      "epoch": 0.27351413733410274,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004997163954638546,
      "loss": 6.5808,
      "mean_token_accuracy": 0.15153488963842393,
      "num_tokens": 5976043.0,
      "step": 2370
    },
    {
      "entropy": 6.740905237197876,
      "epoch": 0.2740911713791114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997143205018306,
      "loss": 6.7368,
      "mean_token_accuracy": 0.1438068449497223,
      "num_tokens": 5988146.0,
      "step": 2375
    },
    {
      "entropy": 6.690333318710327,
      "epoch": 0.27466820542412,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997122379816523,
      "loss": 6.6951,
      "mean_token_accuracy": 0.14996936321258544,
      "num_tokens": 6000542.0,
      "step": 2380
    },
    {
      "entropy": 6.7154529094696045,
      "epoch": 0.27524523946912866,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004997101479033893,
      "loss": 6.7366,
      "mean_token_accuracy": 0.1433037906885147,
      "num_tokens": 6013574.0,
      "step": 2385
    },
    {
      "entropy": 6.607508945465088,
      "epoch": 0.27582227351413735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004997080502671122,
      "loss": 6.6318,
      "mean_token_accuracy": 0.14444206804037094,
      "num_tokens": 6025391.0,
      "step": 2390
    },
    {
      "entropy": 6.780825185775757,
      "epoch": 0.276399307559146,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997059450728913,
      "loss": 6.7493,
      "mean_token_accuracy": 0.1451184406876564,
      "num_tokens": 6036595.0,
      "step": 2395
    },
    {
      "entropy": 6.7437080383300785,
      "epoch": 0.27697634160415463,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997038323207977,
      "loss": 6.8453,
      "mean_token_accuracy": 0.13703761547803878,
      "num_tokens": 6048852.0,
      "step": 2400
    },
    {
      "entropy": 6.73164849281311,
      "epoch": 0.2775533756491633,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997017120109022,
      "loss": 6.698,
      "mean_token_accuracy": 0.15447451919317245,
      "num_tokens": 6061694.0,
      "step": 2405
    },
    {
      "entropy": 6.718098258972168,
      "epoch": 0.27813040969417197,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996995841432762,
      "loss": 6.7536,
      "mean_token_accuracy": 0.14859242737293243,
      "num_tokens": 6074244.0,
      "step": 2410
    },
    {
      "entropy": 6.787304067611695,
      "epoch": 0.2787074437391806,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996974487179915,
      "loss": 6.7248,
      "mean_token_accuracy": 0.14635652750730516,
      "num_tokens": 6085550.0,
      "step": 2415
    },
    {
      "entropy": 6.705214929580689,
      "epoch": 0.27928447778418924,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996953057351195,
      "loss": 6.774,
      "mean_token_accuracy": 0.14355605840682983,
      "num_tokens": 6096749.0,
      "step": 2420
    },
    {
      "entropy": 6.769416809082031,
      "epoch": 0.27986151182919794,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996931551947327,
      "loss": 6.6649,
      "mean_token_accuracy": 0.1461700275540352,
      "num_tokens": 6109490.0,
      "step": 2425
    },
    {
      "entropy": 6.711434459686279,
      "epoch": 0.2804385458742066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996909970969031,
      "loss": 6.8486,
      "mean_token_accuracy": 0.13653022199869155,
      "num_tokens": 6122656.0,
      "step": 2430
    },
    {
      "entropy": 6.859713554382324,
      "epoch": 0.2810155799192152,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996888314417034,
      "loss": 6.724,
      "mean_token_accuracy": 0.1444801300764084,
      "num_tokens": 6135461.0,
      "step": 2435
    },
    {
      "entropy": 6.648364114761352,
      "epoch": 0.2815926139642239,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996866582292067,
      "loss": 6.7647,
      "mean_token_accuracy": 0.14585992693901062,
      "num_tokens": 6147698.0,
      "step": 2440
    },
    {
      "entropy": 6.834606456756592,
      "epoch": 0.28216964800923255,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996844774594856,
      "loss": 6.7349,
      "mean_token_accuracy": 0.1469629168510437,
      "num_tokens": 6160456.0,
      "step": 2445
    },
    {
      "entropy": 6.665376234054565,
      "epoch": 0.2827466820542412,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996822891326138,
      "loss": 6.7258,
      "mean_token_accuracy": 0.14537576138973235,
      "num_tokens": 6173606.0,
      "step": 2450
    },
    {
      "entropy": 6.720003461837768,
      "epoch": 0.28332371609924983,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996800932486648,
      "loss": 6.7506,
      "mean_token_accuracy": 0.14402114003896713,
      "num_tokens": 6185355.0,
      "step": 2455
    },
    {
      "entropy": 6.787649822235108,
      "epoch": 0.28390075014425853,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996778898077126,
      "loss": 6.7692,
      "mean_token_accuracy": 0.14147617369890214,
      "num_tokens": 6198154.0,
      "step": 2460
    },
    {
      "entropy": 6.731502485275269,
      "epoch": 0.28447778418926717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996756788098309,
      "loss": 6.6544,
      "mean_token_accuracy": 0.1494407206773758,
      "num_tokens": 6209901.0,
      "step": 2465
    },
    {
      "entropy": 6.673964262008667,
      "epoch": 0.2850548182342758,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996734602550945,
      "loss": 6.7203,
      "mean_token_accuracy": 0.14480748698115348,
      "num_tokens": 6222052.0,
      "step": 2470
    },
    {
      "entropy": 6.75030665397644,
      "epoch": 0.2856318522792845,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996712341435779,
      "loss": 6.7232,
      "mean_token_accuracy": 0.15031242072582246,
      "num_tokens": 6235112.0,
      "step": 2475
    },
    {
      "entropy": 6.732296323776245,
      "epoch": 0.28620888632429314,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996690004753559,
      "loss": 6.7491,
      "mean_token_accuracy": 0.14172388911247252,
      "num_tokens": 6248225.0,
      "step": 2480
    },
    {
      "entropy": 6.760669803619384,
      "epoch": 0.2867859203693018,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996667592505037,
      "loss": 6.7068,
      "mean_token_accuracy": 0.14936083406209946,
      "num_tokens": 6259518.0,
      "step": 2485
    },
    {
      "entropy": 6.649824762344361,
      "epoch": 0.2873629544143104,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996645104690967,
      "loss": 6.7054,
      "mean_token_accuracy": 0.13711516559123993,
      "num_tokens": 6271832.0,
      "step": 2490
    },
    {
      "entropy": 6.7701342582702635,
      "epoch": 0.2879399884593191,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996622541312103,
      "loss": 6.7534,
      "mean_token_accuracy": 0.14304690212011337,
      "num_tokens": 6284328.0,
      "step": 2495
    },
    {
      "entropy": 6.720312309265137,
      "epoch": 0.28851702250432776,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996599902369207,
      "loss": 6.6935,
      "mean_token_accuracy": 0.14403096809983254,
      "num_tokens": 6296100.0,
      "step": 2500
    },
    {
      "entropy": 6.617523860931397,
      "epoch": 0.2890940565493364,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996577187863039,
      "loss": 6.6476,
      "mean_token_accuracy": 0.15421121418476105,
      "num_tokens": 6308262.0,
      "step": 2505
    },
    {
      "entropy": 6.748142957687378,
      "epoch": 0.2896710905943451,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004996554397794364,
      "loss": 6.7182,
      "mean_token_accuracy": 0.13827703297138214,
      "num_tokens": 6319953.0,
      "step": 2510
    },
    {
      "entropy": 6.702780914306641,
      "epoch": 0.29024812463935373,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996531532163947,
      "loss": 6.6568,
      "mean_token_accuracy": 0.14931394010782242,
      "num_tokens": 6332953.0,
      "step": 2515
    },
    {
      "entropy": 6.671044635772705,
      "epoch": 0.2908251586843624,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996508590972558,
      "loss": 6.7213,
      "mean_token_accuracy": 0.1474708914756775,
      "num_tokens": 6345692.0,
      "step": 2520
    },
    {
      "entropy": 6.7257524013519285,
      "epoch": 0.291402192729371,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996485574220969,
      "loss": 6.696,
      "mean_token_accuracy": 0.1451795779168606,
      "num_tokens": 6358058.0,
      "step": 2525
    },
    {
      "entropy": 6.783554315567017,
      "epoch": 0.2919792267743797,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996462481909953,
      "loss": 6.7313,
      "mean_token_accuracy": 0.14846932888031006,
      "num_tokens": 6370469.0,
      "step": 2530
    },
    {
      "entropy": 6.787226629257202,
      "epoch": 0.29255626081938835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996439314040287,
      "loss": 6.7985,
      "mean_token_accuracy": 0.1353658489882946,
      "num_tokens": 6384333.0,
      "step": 2535
    },
    {
      "entropy": 6.7711738586425785,
      "epoch": 0.293133294864397,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499641607061275,
      "loss": 6.7377,
      "mean_token_accuracy": 0.14200848788022996,
      "num_tokens": 6397688.0,
      "step": 2540
    },
    {
      "entropy": 6.706867218017578,
      "epoch": 0.2937103289094057,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996392751628126,
      "loss": 6.699,
      "mean_token_accuracy": 0.14663641303777694,
      "num_tokens": 6410188.0,
      "step": 2545
    },
    {
      "entropy": 6.719365406036377,
      "epoch": 0.2942873629544143,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996369357087196,
      "loss": 6.7169,
      "mean_token_accuracy": 0.14374764263629913,
      "num_tokens": 6423313.0,
      "step": 2550
    },
    {
      "entropy": 6.764793729782104,
      "epoch": 0.29486439699942296,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499634588699075,
      "loss": 6.7386,
      "mean_token_accuracy": 0.1467905879020691,
      "num_tokens": 6435853.0,
      "step": 2555
    },
    {
      "entropy": 6.598630714416504,
      "epoch": 0.2954414310444316,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004996322341339575,
      "loss": 6.6725,
      "mean_token_accuracy": 0.14465839564800262,
      "num_tokens": 6448659.0,
      "step": 2560
    },
    {
      "entropy": 6.732291269302368,
      "epoch": 0.2960184650894403,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996298720134465,
      "loss": 6.7292,
      "mean_token_accuracy": 0.14818664789199829,
      "num_tokens": 6460534.0,
      "step": 2565
    },
    {
      "entropy": 6.716020393371582,
      "epoch": 0.29659549913444894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996275023376213,
      "loss": 6.6971,
      "mean_token_accuracy": 0.14435341879725455,
      "num_tokens": 6472368.0,
      "step": 2570
    },
    {
      "entropy": 6.693843650817871,
      "epoch": 0.2971725331794576,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996251251065615,
      "loss": 6.7075,
      "mean_token_accuracy": 0.14448096752166747,
      "num_tokens": 6485430.0,
      "step": 2575
    },
    {
      "entropy": 6.710673141479492,
      "epoch": 0.2977495672244662,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996227403203473,
      "loss": 6.6891,
      "mean_token_accuracy": 0.1485920026898384,
      "num_tokens": 6497303.0,
      "step": 2580
    },
    {
      "entropy": 6.699700403213501,
      "epoch": 0.2983266012694749,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996203479790589,
      "loss": 6.7567,
      "mean_token_accuracy": 0.15133860260248183,
      "num_tokens": 6510154.0,
      "step": 2585
    },
    {
      "entropy": 6.790226650238037,
      "epoch": 0.29890363531448355,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004996179480827764,
      "loss": 6.7413,
      "mean_token_accuracy": 0.143829607963562,
      "num_tokens": 6523091.0,
      "step": 2590
    },
    {
      "entropy": 6.75448055267334,
      "epoch": 0.2994806693594922,
      "grad_norm": 1.28125,
      "learning_rate": 0.000499615540631581,
      "loss": 6.6531,
      "mean_token_accuracy": 0.14935905635356903,
      "num_tokens": 6536043.0,
      "step": 2595
    },
    {
      "entropy": 6.624383497238159,
      "epoch": 0.3000577034045009,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996131256255534,
      "loss": 6.6653,
      "mean_token_accuracy": 0.14822530150413513,
      "num_tokens": 6548227.0,
      "step": 2600
    },
    {
      "entropy": 6.759153127670288,
      "epoch": 0.3006347374495095,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996107030647749,
      "loss": 6.7117,
      "mean_token_accuracy": 0.14047631323337556,
      "num_tokens": 6560204.0,
      "step": 2605
    },
    {
      "entropy": 6.684758043289184,
      "epoch": 0.30121177149451817,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004996082729493269,
      "loss": 6.726,
      "mean_token_accuracy": 0.150347737967968,
      "num_tokens": 6573235.0,
      "step": 2610
    },
    {
      "entropy": 6.650673818588257,
      "epoch": 0.3017888055395268,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996058352792913,
      "loss": 6.7416,
      "mean_token_accuracy": 0.14433368742465974,
      "num_tokens": 6585118.0,
      "step": 2615
    },
    {
      "entropy": 6.750360155105591,
      "epoch": 0.3023658395845355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049960339005475,
      "loss": 6.6339,
      "mean_token_accuracy": 0.14643110930919648,
      "num_tokens": 6596690.0,
      "step": 2620
    },
    {
      "entropy": 6.660681819915771,
      "epoch": 0.30294287362954414,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996009372757852,
      "loss": 6.6344,
      "mean_token_accuracy": 0.15728502422571183,
      "num_tokens": 6609076.0,
      "step": 2625
    },
    {
      "entropy": 6.662906169891357,
      "epoch": 0.3035199076745528,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995984769424795,
      "loss": 6.7326,
      "mean_token_accuracy": 0.14466301277279853,
      "num_tokens": 6621352.0,
      "step": 2630
    },
    {
      "entropy": 6.752713441848755,
      "epoch": 0.3040969417195615,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004995960090549155,
      "loss": 6.6363,
      "mean_token_accuracy": 0.1454766198992729,
      "num_tokens": 6633924.0,
      "step": 2635
    },
    {
      "entropy": 6.702887296676636,
      "epoch": 0.3046739757645701,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995935336131764,
      "loss": 6.7607,
      "mean_token_accuracy": 0.14415173679590226,
      "num_tokens": 6646490.0,
      "step": 2640
    },
    {
      "entropy": 6.83890929222107,
      "epoch": 0.30525100980957875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995910506173452,
      "loss": 6.7272,
      "mean_token_accuracy": 0.14090896472334863,
      "num_tokens": 6660171.0,
      "step": 2645
    },
    {
      "entropy": 6.609751653671265,
      "epoch": 0.3058280438545874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995885600675059,
      "loss": 6.7572,
      "mean_token_accuracy": 0.14278358295559884,
      "num_tokens": 6673505.0,
      "step": 2650
    },
    {
      "entropy": 6.783524751663208,
      "epoch": 0.3064050778995961,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995860619637415,
      "loss": 6.7358,
      "mean_token_accuracy": 0.14450631141662598,
      "num_tokens": 6685798.0,
      "step": 2655
    },
    {
      "entropy": 6.639528703689575,
      "epoch": 0.30698211194460473,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004995835563061367,
      "loss": 6.5888,
      "mean_token_accuracy": 0.14924851357936858,
      "num_tokens": 6699254.0,
      "step": 2660
    },
    {
      "entropy": 6.685869789123535,
      "epoch": 0.30755914598961337,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995810430947756,
      "loss": 6.6506,
      "mean_token_accuracy": 0.14859471917152406,
      "num_tokens": 6711088.0,
      "step": 2665
    },
    {
      "entropy": 6.700608539581299,
      "epoch": 0.30813618003462206,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995785223297426,
      "loss": 6.7159,
      "mean_token_accuracy": 0.14112428277730943,
      "num_tokens": 6723373.0,
      "step": 2670
    },
    {
      "entropy": 6.637734365463257,
      "epoch": 0.3087132140796307,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004995759940111225,
      "loss": 6.7234,
      "mean_token_accuracy": 0.14265928864479066,
      "num_tokens": 6737984.0,
      "step": 2675
    },
    {
      "entropy": 6.751663398742676,
      "epoch": 0.30929024812463934,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995734581390005,
      "loss": 6.6699,
      "mean_token_accuracy": 0.14623800665140152,
      "num_tokens": 6751283.0,
      "step": 2680
    },
    {
      "entropy": 6.637785387039185,
      "epoch": 0.309867282169648,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995709147134617,
      "loss": 6.6511,
      "mean_token_accuracy": 0.1425285041332245,
      "num_tokens": 6765012.0,
      "step": 2685
    },
    {
      "entropy": 6.657239103317261,
      "epoch": 0.3104443162146567,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995683637345919,
      "loss": 6.6944,
      "mean_token_accuracy": 0.14595674276351928,
      "num_tokens": 6777708.0,
      "step": 2690
    },
    {
      "entropy": 6.682321071624756,
      "epoch": 0.3110213502596653,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995658052024765,
      "loss": 6.6153,
      "mean_token_accuracy": 0.15794552862644196,
      "num_tokens": 6791325.0,
      "step": 2695
    },
    {
      "entropy": 6.583492231369019,
      "epoch": 0.31159838430467396,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995632391172019,
      "loss": 6.7509,
      "mean_token_accuracy": 0.1372663803398609,
      "num_tokens": 6805659.0,
      "step": 2700
    },
    {
      "entropy": 6.76855502128601,
      "epoch": 0.31217541834968265,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995606654788543,
      "loss": 6.6591,
      "mean_token_accuracy": 0.1441938564181328,
      "num_tokens": 6818110.0,
      "step": 2705
    },
    {
      "entropy": 6.640047025680542,
      "epoch": 0.3127524523946913,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995580842875202,
      "loss": 6.6691,
      "mean_token_accuracy": 0.14442232996225357,
      "num_tokens": 6830959.0,
      "step": 2710
    },
    {
      "entropy": 6.750000095367431,
      "epoch": 0.31332948643969993,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995554955432866,
      "loss": 6.7234,
      "mean_token_accuracy": 0.14586163461208343,
      "num_tokens": 6843318.0,
      "step": 2715
    },
    {
      "entropy": 6.675041437149048,
      "epoch": 0.3139065204847086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995528992462404,
      "loss": 6.634,
      "mean_token_accuracy": 0.14719301089644432,
      "num_tokens": 6855154.0,
      "step": 2720
    },
    {
      "entropy": 6.5928421974182125,
      "epoch": 0.31448355452971727,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499550295396469,
      "loss": 6.685,
      "mean_token_accuracy": 0.14665439277887343,
      "num_tokens": 6867428.0,
      "step": 2725
    },
    {
      "entropy": 6.754723072052002,
      "epoch": 0.3150605885747259,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995476839940598,
      "loss": 6.671,
      "mean_token_accuracy": 0.15316719859838485,
      "num_tokens": 6879989.0,
      "step": 2730
    },
    {
      "entropy": 6.6928040981292725,
      "epoch": 0.31563762261973455,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499545065039101,
      "loss": 6.6906,
      "mean_token_accuracy": 0.14096458703279496,
      "num_tokens": 6893179.0,
      "step": 2735
    },
    {
      "entropy": 6.727335596084595,
      "epoch": 0.31621465666474324,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995424385316803,
      "loss": 6.6782,
      "mean_token_accuracy": 0.14420599341392518,
      "num_tokens": 6905647.0,
      "step": 2740
    },
    {
      "entropy": 6.610552978515625,
      "epoch": 0.3167916907097519,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995398044718863,
      "loss": 6.6364,
      "mean_token_accuracy": 0.15449649840593338,
      "num_tokens": 6917361.0,
      "step": 2745
    },
    {
      "entropy": 6.71636061668396,
      "epoch": 0.3173687247547605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995371628598074,
      "loss": 6.6653,
      "mean_token_accuracy": 0.1424078017473221,
      "num_tokens": 6929515.0,
      "step": 2750
    },
    {
      "entropy": 6.707604217529297,
      "epoch": 0.31794575879976916,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995345136955328,
      "loss": 6.6391,
      "mean_token_accuracy": 0.14832933098077775,
      "num_tokens": 6942265.0,
      "step": 2755
    },
    {
      "entropy": 6.584933423995972,
      "epoch": 0.31852279284477786,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995318569791512,
      "loss": 6.687,
      "mean_token_accuracy": 0.14602783024311067,
      "num_tokens": 6955151.0,
      "step": 2760
    },
    {
      "entropy": 6.765100526809692,
      "epoch": 0.3190998268897865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995291927107522,
      "loss": 6.6589,
      "mean_token_accuracy": 0.14955082535743713,
      "num_tokens": 6968378.0,
      "step": 2765
    },
    {
      "entropy": 6.605684900283814,
      "epoch": 0.31967686093479514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004995265208904252,
      "loss": 6.6685,
      "mean_token_accuracy": 0.1471999056637287,
      "num_tokens": 6980083.0,
      "step": 2770
    },
    {
      "entropy": 6.687409925460815,
      "epoch": 0.32025389497980383,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995238415182603,
      "loss": 6.6717,
      "mean_token_accuracy": 0.14363027215003968,
      "num_tokens": 6992981.0,
      "step": 2775
    },
    {
      "entropy": 6.824726104736328,
      "epoch": 0.3208309290248125,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995211545943476,
      "loss": 6.6822,
      "mean_token_accuracy": 0.14623156040906907,
      "num_tokens": 7006115.0,
      "step": 2780
    },
    {
      "entropy": 6.589347887039184,
      "epoch": 0.3214079630698211,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995184601187772,
      "loss": 6.6416,
      "mean_token_accuracy": 0.15562028288841248,
      "num_tokens": 7019176.0,
      "step": 2785
    },
    {
      "entropy": 6.55873851776123,
      "epoch": 0.32198499711482975,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049951575809164,
      "loss": 6.5673,
      "mean_token_accuracy": 0.15908343940973282,
      "num_tokens": 7031214.0,
      "step": 2790
    },
    {
      "entropy": 6.615464973449707,
      "epoch": 0.32256203115983845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995130485130269,
      "loss": 6.6403,
      "mean_token_accuracy": 0.14976823478937148,
      "num_tokens": 7044702.0,
      "step": 2795
    },
    {
      "entropy": 6.705451583862304,
      "epoch": 0.3231390652048471,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995103313830289,
      "loss": 6.6522,
      "mean_token_accuracy": 0.14741754829883574,
      "num_tokens": 7056489.0,
      "step": 2800
    },
    {
      "entropy": 6.621556758880615,
      "epoch": 0.3237160992498557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004995076067017373,
      "loss": 6.6752,
      "mean_token_accuracy": 0.14415591955184937,
      "num_tokens": 7068972.0,
      "step": 2805
    },
    {
      "entropy": 6.72447772026062,
      "epoch": 0.3242931332948644,
      "grad_norm": 1.125,
      "learning_rate": 0.0004995048744692439,
      "loss": 6.6197,
      "mean_token_accuracy": 0.14744585007429123,
      "num_tokens": 7080894.0,
      "step": 2810
    },
    {
      "entropy": 6.598182201385498,
      "epoch": 0.32487016733987306,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995021346856405,
      "loss": 6.6696,
      "mean_token_accuracy": 0.14264965504407884,
      "num_tokens": 7094427.0,
      "step": 2815
    },
    {
      "entropy": 6.590505361557007,
      "epoch": 0.3254472013848817,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004994993873510196,
      "loss": 6.6003,
      "mean_token_accuracy": 0.15626100897789003,
      "num_tokens": 7105710.0,
      "step": 2820
    },
    {
      "entropy": 6.777636337280273,
      "epoch": 0.32602423542989034,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994966324654731,
      "loss": 6.6376,
      "mean_token_accuracy": 0.1449590690433979,
      "num_tokens": 7119042.0,
      "step": 2825
    },
    {
      "entropy": 6.564963388442993,
      "epoch": 0.32660126947489904,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499493870029094,
      "loss": 6.6343,
      "mean_token_accuracy": 0.14468320608139038,
      "num_tokens": 7132146.0,
      "step": 2830
    },
    {
      "entropy": 6.669665241241455,
      "epoch": 0.3271783035199077,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994911000419749,
      "loss": 6.6033,
      "mean_token_accuracy": 0.14561504125595093,
      "num_tokens": 7144805.0,
      "step": 2835
    },
    {
      "entropy": 6.571587038040161,
      "epoch": 0.3277553375649163,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994883225042093,
      "loss": 6.5762,
      "mean_token_accuracy": 0.15571997463703155,
      "num_tokens": 7156906.0,
      "step": 2840
    },
    {
      "entropy": 6.654489612579345,
      "epoch": 0.328332371609925,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994855374158905,
      "loss": 6.567,
      "mean_token_accuracy": 0.14885426163673401,
      "num_tokens": 7169451.0,
      "step": 2845
    },
    {
      "entropy": 6.650216817855835,
      "epoch": 0.32890940565493365,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000499482744777112,
      "loss": 6.7134,
      "mean_token_accuracy": 0.14018807634711267,
      "num_tokens": 7183211.0,
      "step": 2850
    },
    {
      "entropy": 6.727432727813721,
      "epoch": 0.3294864396999423,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499479944587968,
      "loss": 6.6583,
      "mean_token_accuracy": 0.14724580198526382,
      "num_tokens": 7196062.0,
      "step": 2855
    },
    {
      "entropy": 6.719376134872436,
      "epoch": 0.33006347374495093,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994771368485526,
      "loss": 6.7601,
      "mean_token_accuracy": 0.14483619555830957,
      "num_tokens": 7208079.0,
      "step": 2860
    },
    {
      "entropy": 6.697180366516113,
      "epoch": 0.3306405077899596,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994743215589602,
      "loss": 6.6972,
      "mean_token_accuracy": 0.14615591317415239,
      "num_tokens": 7221537.0,
      "step": 2865
    },
    {
      "entropy": 6.6564757347106935,
      "epoch": 0.33121754183496827,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994714987192854,
      "loss": 6.6652,
      "mean_token_accuracy": 0.14372813254594802,
      "num_tokens": 7234472.0,
      "step": 2870
    },
    {
      "entropy": 6.643028306961059,
      "epoch": 0.3317945758799769,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004994686683296234,
      "loss": 6.6194,
      "mean_token_accuracy": 0.15096134543418885,
      "num_tokens": 7247818.0,
      "step": 2875
    },
    {
      "entropy": 6.6859259605407715,
      "epoch": 0.3323716099249856,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004994658303900691,
      "loss": 6.7006,
      "mean_token_accuracy": 0.14851104468107224,
      "num_tokens": 7261351.0,
      "step": 2880
    },
    {
      "entropy": 6.607791996002197,
      "epoch": 0.33294864396999424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994629849007182,
      "loss": 6.568,
      "mean_token_accuracy": 0.1521388918161392,
      "num_tokens": 7274310.0,
      "step": 2885
    },
    {
      "entropy": 6.573197507858277,
      "epoch": 0.3335256780150029,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994601318616663,
      "loss": 6.6096,
      "mean_token_accuracy": 0.1469192236661911,
      "num_tokens": 7286560.0,
      "step": 2890
    },
    {
      "entropy": 6.756811904907226,
      "epoch": 0.3341027120600115,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994572712730092,
      "loss": 6.6835,
      "mean_token_accuracy": 0.14362717866897584,
      "num_tokens": 7300337.0,
      "step": 2895
    },
    {
      "entropy": 6.617220211029053,
      "epoch": 0.3346797461050202,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994544031348434,
      "loss": 6.6485,
      "mean_token_accuracy": 0.1463576778769493,
      "num_tokens": 7313436.0,
      "step": 2900
    },
    {
      "entropy": 6.632095098495483,
      "epoch": 0.33525678015002885,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994515274472653,
      "loss": 6.5149,
      "mean_token_accuracy": 0.16170531660318374,
      "num_tokens": 7325766.0,
      "step": 2905
    },
    {
      "entropy": 6.578122663497925,
      "epoch": 0.3358338141950375,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004994486442103715,
      "loss": 6.614,
      "mean_token_accuracy": 0.1468640923500061,
      "num_tokens": 7337904.0,
      "step": 2910
    },
    {
      "entropy": 6.575803184509278,
      "epoch": 0.3364108482400462,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499445753424259,
      "loss": 6.5992,
      "mean_token_accuracy": 0.14721653759479522,
      "num_tokens": 7350448.0,
      "step": 2915
    },
    {
      "entropy": 6.711412048339843,
      "epoch": 0.33698788228505483,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994428550890251,
      "loss": 6.6398,
      "mean_token_accuracy": 0.14552380070090293,
      "num_tokens": 7363859.0,
      "step": 2920
    },
    {
      "entropy": 6.663523769378662,
      "epoch": 0.33756491633006347,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994399492047672,
      "loss": 6.6548,
      "mean_token_accuracy": 0.14727362394332885,
      "num_tokens": 7376963.0,
      "step": 2925
    },
    {
      "entropy": 6.665795135498047,
      "epoch": 0.3381419503750721,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994370357715832,
      "loss": 6.6806,
      "mean_token_accuracy": 0.14258197993040084,
      "num_tokens": 7390370.0,
      "step": 2930
    },
    {
      "entropy": 6.65341739654541,
      "epoch": 0.3387189844200808,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499434114789571,
      "loss": 6.6101,
      "mean_token_accuracy": 0.14991432875394822,
      "num_tokens": 7402851.0,
      "step": 2935
    },
    {
      "entropy": 6.5872947692871096,
      "epoch": 0.33929601846508944,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994311862588286,
      "loss": 6.6406,
      "mean_token_accuracy": 0.14421918392181396,
      "num_tokens": 7415428.0,
      "step": 2940
    },
    {
      "entropy": 6.719788837432861,
      "epoch": 0.3398730525100981,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994282501794551,
      "loss": 6.6325,
      "mean_token_accuracy": 0.1465821772813797,
      "num_tokens": 7428420.0,
      "step": 2945
    },
    {
      "entropy": 6.657585000991821,
      "epoch": 0.3404500865551068,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004994253065515486,
      "loss": 6.689,
      "mean_token_accuracy": 0.14700795263051986,
      "num_tokens": 7439626.0,
      "step": 2950
    },
    {
      "entropy": 6.661691093444825,
      "epoch": 0.3410271206001154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994223553752084,
      "loss": 6.6467,
      "mean_token_accuracy": 0.14639809280633925,
      "num_tokens": 7452797.0,
      "step": 2955
    },
    {
      "entropy": 6.65012903213501,
      "epoch": 0.34160415464512406,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994193966505339,
      "loss": 6.6162,
      "mean_token_accuracy": 0.15301314145326614,
      "num_tokens": 7464311.0,
      "step": 2960
    },
    {
      "entropy": 6.575049352645874,
      "epoch": 0.3421811886901327,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994164303776244,
      "loss": 6.6131,
      "mean_token_accuracy": 0.14517637491226196,
      "num_tokens": 7475706.0,
      "step": 2965
    },
    {
      "entropy": 6.673529767990113,
      "epoch": 0.3427582227351414,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994134565565797,
      "loss": 6.7028,
      "mean_token_accuracy": 0.1451819732785225,
      "num_tokens": 7489077.0,
      "step": 2970
    },
    {
      "entropy": 6.704615211486816,
      "epoch": 0.34333525678015003,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994104751875,
      "loss": 6.651,
      "mean_token_accuracy": 0.1511762797832489,
      "num_tokens": 7501598.0,
      "step": 2975
    },
    {
      "entropy": 6.562963962554932,
      "epoch": 0.3439122908251587,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004994074862704854,
      "loss": 6.46,
      "mean_token_accuracy": 0.1568878024816513,
      "num_tokens": 7514776.0,
      "step": 2980
    },
    {
      "entropy": 6.517637109756469,
      "epoch": 0.3444893248701673,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994044898056363,
      "loss": 6.6212,
      "mean_token_accuracy": 0.14930416494607926,
      "num_tokens": 7527795.0,
      "step": 2985
    },
    {
      "entropy": 6.645235443115235,
      "epoch": 0.345066358915176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994014857930538,
      "loss": 6.5076,
      "mean_token_accuracy": 0.15968223363161088,
      "num_tokens": 7541392.0,
      "step": 2990
    },
    {
      "entropy": 6.542176866531372,
      "epoch": 0.34564339296018465,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993984742328387,
      "loss": 6.6275,
      "mean_token_accuracy": 0.14863153249025346,
      "num_tokens": 7554369.0,
      "step": 2995
    },
    {
      "entropy": 6.647294902801514,
      "epoch": 0.3462204270051933,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993954551250925,
      "loss": 6.6432,
      "mean_token_accuracy": 0.1486467257142067,
      "num_tokens": 7567528.0,
      "step": 3000
    },
    {
      "epoch": 0.3462204270051933,
      "eval_entropy": 6.603502982646388,
      "eval_loss": 6.686180591583252,
      "eval_mean_token_accuracy": 0.14944647604010136,
      "eval_num_tokens": 7567528.0,
      "eval_runtime": 22.9724,
      "eval_samples_per_second": 1224.776,
      "eval_steps_per_second": 153.097,
      "step": 3000
    },
    {
      "entropy": 6.667821407318115,
      "epoch": 0.346797461050202,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993924284699165,
      "loss": 6.5825,
      "mean_token_accuracy": 0.15041287392377853,
      "num_tokens": 7579146.0,
      "step": 3005
    },
    {
      "entropy": 6.5950047969818115,
      "epoch": 0.3473744950952106,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993893942674127,
      "loss": 6.6828,
      "mean_token_accuracy": 0.14685314893722534,
      "num_tokens": 7591902.0,
      "step": 3010
    },
    {
      "entropy": 6.754814577102661,
      "epoch": 0.34795152914021926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499386352517683,
      "loss": 6.6803,
      "mean_token_accuracy": 0.15172169506549835,
      "num_tokens": 7605565.0,
      "step": 3015
    },
    {
      "entropy": 6.726373863220215,
      "epoch": 0.3485285631852279,
      "grad_norm": 0.890625,
      "learning_rate": 0.00049938330322083,
      "loss": 6.6271,
      "mean_token_accuracy": 0.15227940529584885,
      "num_tokens": 7618380.0,
      "step": 3020
    },
    {
      "entropy": 6.556258058547973,
      "epoch": 0.3491055972302366,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993802463769558,
      "loss": 6.6459,
      "mean_token_accuracy": 0.14672804325819017,
      "num_tokens": 7632092.0,
      "step": 3025
    },
    {
      "entropy": 6.67885012626648,
      "epoch": 0.34968263127524524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993771819861636,
      "loss": 6.6585,
      "mean_token_accuracy": 0.14378119856119156,
      "num_tokens": 7644185.0,
      "step": 3030
    },
    {
      "entropy": 6.649367904663086,
      "epoch": 0.3502596653202539,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993741100485564,
      "loss": 6.5704,
      "mean_token_accuracy": 0.14890087842941285,
      "num_tokens": 7656635.0,
      "step": 3035
    },
    {
      "entropy": 6.630291414260864,
      "epoch": 0.35083669936526257,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993710305642374,
      "loss": 6.6006,
      "mean_token_accuracy": 0.14622372686862944,
      "num_tokens": 7669643.0,
      "step": 3040
    },
    {
      "entropy": 6.656918716430664,
      "epoch": 0.3514137334102712,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993679435333102,
      "loss": 6.6382,
      "mean_token_accuracy": 0.13970368951559067,
      "num_tokens": 7682489.0,
      "step": 3045
    },
    {
      "entropy": 6.555567073822021,
      "epoch": 0.35199076745527985,
      "grad_norm": 1.25,
      "learning_rate": 0.0004993648489558788,
      "loss": 6.5845,
      "mean_token_accuracy": 0.1465017504990101,
      "num_tokens": 7696670.0,
      "step": 3050
    },
    {
      "entropy": 6.7070152282714846,
      "epoch": 0.3525678015002885,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993617468320471,
      "loss": 6.6109,
      "mean_token_accuracy": 0.15563664734363555,
      "num_tokens": 7709054.0,
      "step": 3055
    },
    {
      "entropy": 6.525317859649658,
      "epoch": 0.3531448355452972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993586371619193,
      "loss": 6.6004,
      "mean_token_accuracy": 0.14743918776512147,
      "num_tokens": 7721710.0,
      "step": 3060
    },
    {
      "entropy": 6.660245132446289,
      "epoch": 0.3537218695903058,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004993555199456004,
      "loss": 6.6536,
      "mean_token_accuracy": 0.14922718554735184,
      "num_tokens": 7733689.0,
      "step": 3065
    },
    {
      "entropy": 6.594697332382202,
      "epoch": 0.35429890363531447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993523951831949,
      "loss": 6.6406,
      "mean_token_accuracy": 0.15407218933105468,
      "num_tokens": 7745735.0,
      "step": 3070
    },
    {
      "entropy": 6.747467994689941,
      "epoch": 0.35487593768032316,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993492628748081,
      "loss": 6.6545,
      "mean_token_accuracy": 0.14074160531163216,
      "num_tokens": 7757352.0,
      "step": 3075
    },
    {
      "entropy": 6.607062578201294,
      "epoch": 0.3554529717253318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993461230205453,
      "loss": 6.5873,
      "mean_token_accuracy": 0.1461851865053177,
      "num_tokens": 7768363.0,
      "step": 3080
    },
    {
      "entropy": 6.6232137203216555,
      "epoch": 0.35603000577034044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993429756205121,
      "loss": 6.7003,
      "mean_token_accuracy": 0.14090533256530763,
      "num_tokens": 7781133.0,
      "step": 3085
    },
    {
      "entropy": 6.777905750274658,
      "epoch": 0.3566070398153491,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993398206748142,
      "loss": 6.6873,
      "mean_token_accuracy": 0.14060682207345962,
      "num_tokens": 7794445.0,
      "step": 3090
    },
    {
      "entropy": 6.557980489730835,
      "epoch": 0.3571840738603578,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993366581835581,
      "loss": 6.5819,
      "mean_token_accuracy": 0.15084142833948136,
      "num_tokens": 7807601.0,
      "step": 3095
    },
    {
      "entropy": 6.540106964111328,
      "epoch": 0.3577611079053664,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993334881468498,
      "loss": 6.5576,
      "mean_token_accuracy": 0.15417141318321229,
      "num_tokens": 7819821.0,
      "step": 3100
    },
    {
      "entropy": 6.614301109313965,
      "epoch": 0.35833814195037506,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004993303105647961,
      "loss": 6.5747,
      "mean_token_accuracy": 0.14614431858062743,
      "num_tokens": 7832214.0,
      "step": 3105
    },
    {
      "entropy": 6.612269878387451,
      "epoch": 0.35891517599538375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004993271254375038,
      "loss": 6.6275,
      "mean_token_accuracy": 0.15465223789215088,
      "num_tokens": 7844566.0,
      "step": 3110
    },
    {
      "entropy": 6.661264038085937,
      "epoch": 0.3594922100403924,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00049932393276508,
      "loss": 6.6016,
      "mean_token_accuracy": 0.15167122855782508,
      "num_tokens": 7857965.0,
      "step": 3115
    },
    {
      "entropy": 6.588106966018676,
      "epoch": 0.36006924408540103,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004993207325476323,
      "loss": 6.6154,
      "mean_token_accuracy": 0.1525164559483528,
      "num_tokens": 7870902.0,
      "step": 3120
    },
    {
      "entropy": 6.747092199325562,
      "epoch": 0.36064627813040967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004993175247852681,
      "loss": 6.6855,
      "mean_token_accuracy": 0.13674851804971694,
      "num_tokens": 7883694.0,
      "step": 3125
    },
    {
      "entropy": 6.653594636917115,
      "epoch": 0.36122331217541837,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993143094780954,
      "loss": 6.5716,
      "mean_token_accuracy": 0.15035519748926163,
      "num_tokens": 7896693.0,
      "step": 3130
    },
    {
      "entropy": 6.549767684936524,
      "epoch": 0.361800346220427,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004993110866262224,
      "loss": 6.5811,
      "mean_token_accuracy": 0.14896219372749328,
      "num_tokens": 7910010.0,
      "step": 3135
    },
    {
      "entropy": 6.704126453399658,
      "epoch": 0.36237738026543564,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993078562297573,
      "loss": 6.6702,
      "mean_token_accuracy": 0.136037989705801,
      "num_tokens": 7922026.0,
      "step": 3140
    },
    {
      "entropy": 6.580215311050415,
      "epoch": 0.36295441431044434,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993046182888089,
      "loss": 6.6542,
      "mean_token_accuracy": 0.1394454210996628,
      "num_tokens": 7934633.0,
      "step": 3145
    },
    {
      "entropy": 6.594920349121094,
      "epoch": 0.363531448355453,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004993013728034861,
      "loss": 6.5489,
      "mean_token_accuracy": 0.15306085348129272,
      "num_tokens": 7947189.0,
      "step": 3150
    },
    {
      "entropy": 6.6680103778839115,
      "epoch": 0.3641084824004616,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000499298119773898,
      "loss": 6.6236,
      "mean_token_accuracy": 0.14880157262086868,
      "num_tokens": 7961283.0,
      "step": 3155
    },
    {
      "entropy": 6.523037147521973,
      "epoch": 0.36468551644547026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992948592001541,
      "loss": 6.6176,
      "mean_token_accuracy": 0.15099794641137124,
      "num_tokens": 7974589.0,
      "step": 3160
    },
    {
      "entropy": 6.691275119781494,
      "epoch": 0.36526255049047895,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499291591082364,
      "loss": 6.6758,
      "mean_token_accuracy": 0.14682248681783677,
      "num_tokens": 7987148.0,
      "step": 3165
    },
    {
      "entropy": 6.697682428359985,
      "epoch": 0.3658395845354876,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992883154206377,
      "loss": 6.6664,
      "mean_token_accuracy": 0.1428602494299412,
      "num_tokens": 8000679.0,
      "step": 3170
    },
    {
      "entropy": 6.5670891284942625,
      "epoch": 0.36641661858049623,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992850322150853,
      "loss": 6.5606,
      "mean_token_accuracy": 0.15191134065389633,
      "num_tokens": 8014011.0,
      "step": 3175
    },
    {
      "entropy": 6.69585018157959,
      "epoch": 0.36699365262550493,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992817414658172,
      "loss": 6.5802,
      "mean_token_accuracy": 0.14718159288167953,
      "num_tokens": 8026022.0,
      "step": 3180
    },
    {
      "entropy": 6.63649697303772,
      "epoch": 0.36757068667051357,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992784431729442,
      "loss": 6.6461,
      "mean_token_accuracy": 0.14684277549386024,
      "num_tokens": 8040123.0,
      "step": 3185
    },
    {
      "entropy": 6.675275611877441,
      "epoch": 0.3681477207155222,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992751373365771,
      "loss": 6.6392,
      "mean_token_accuracy": 0.15477651953697205,
      "num_tokens": 8053320.0,
      "step": 3190
    },
    {
      "entropy": 6.5776519775390625,
      "epoch": 0.36872475476053085,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992718239568273,
      "loss": 6.5876,
      "mean_token_accuracy": 0.1540097713470459,
      "num_tokens": 8066682.0,
      "step": 3195
    },
    {
      "entropy": 6.631053400039673,
      "epoch": 0.36930178880553954,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499268503033806,
      "loss": 6.5963,
      "mean_token_accuracy": 0.15034545511007308,
      "num_tokens": 8078887.0,
      "step": 3200
    },
    {
      "entropy": 6.641768169403076,
      "epoch": 0.3698788228505482,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992651745676249,
      "loss": 6.587,
      "mean_token_accuracy": 0.14526487439870833,
      "num_tokens": 8090801.0,
      "step": 3205
    },
    {
      "entropy": 6.607954406738282,
      "epoch": 0.3704558568955568,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992618385583962,
      "loss": 6.6613,
      "mean_token_accuracy": 0.14384082555770875,
      "num_tokens": 8103387.0,
      "step": 3210
    },
    {
      "entropy": 6.5886212348937985,
      "epoch": 0.3710328909405655,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992584950062319,
      "loss": 6.5145,
      "mean_token_accuracy": 0.1546319343149662,
      "num_tokens": 8115356.0,
      "step": 3215
    },
    {
      "entropy": 6.674334526062012,
      "epoch": 0.37160992498557416,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004992551439112446,
      "loss": 6.6369,
      "mean_token_accuracy": 0.14844774454832077,
      "num_tokens": 8129141.0,
      "step": 3220
    },
    {
      "entropy": 6.7329993724823,
      "epoch": 0.3721869590305828,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992517852735469,
      "loss": 6.659,
      "mean_token_accuracy": 0.15107582956552507,
      "num_tokens": 8142453.0,
      "step": 3225
    },
    {
      "entropy": 6.574329853057861,
      "epoch": 0.37276399307559144,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992484190932517,
      "loss": 6.6139,
      "mean_token_accuracy": 0.14692633971571922,
      "num_tokens": 8154361.0,
      "step": 3230
    },
    {
      "entropy": 6.7408263206481935,
      "epoch": 0.37334102712060013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992450453704723,
      "loss": 6.6697,
      "mean_token_accuracy": 0.14180598258972169,
      "num_tokens": 8168354.0,
      "step": 3235
    },
    {
      "entropy": 6.6767510890960695,
      "epoch": 0.3739180611656088,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992416641053223,
      "loss": 6.6188,
      "mean_token_accuracy": 0.15147449523210527,
      "num_tokens": 8180700.0,
      "step": 3240
    },
    {
      "entropy": 6.61651201248169,
      "epoch": 0.3744950952106174,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004992382752979153,
      "loss": 6.613,
      "mean_token_accuracy": 0.14319541156291962,
      "num_tokens": 8193987.0,
      "step": 3245
    },
    {
      "entropy": 6.638578748703003,
      "epoch": 0.3750721292556261,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992348789483651,
      "loss": 6.7071,
      "mean_token_accuracy": 0.14582299664616585,
      "num_tokens": 8207150.0,
      "step": 3250
    },
    {
      "entropy": 6.710038661956787,
      "epoch": 0.37564916330063475,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992314750567864,
      "loss": 6.6528,
      "mean_token_accuracy": 0.14282948076725005,
      "num_tokens": 8221050.0,
      "step": 3255
    },
    {
      "entropy": 6.617986631393433,
      "epoch": 0.3762261973456434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992280636232933,
      "loss": 6.5845,
      "mean_token_accuracy": 0.14726462215185165,
      "num_tokens": 8233641.0,
      "step": 3260
    },
    {
      "entropy": 6.666138219833374,
      "epoch": 0.376803231390652,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992246446480007,
      "loss": 6.5981,
      "mean_token_accuracy": 0.15233627408742906,
      "num_tokens": 8246717.0,
      "step": 3265
    },
    {
      "entropy": 6.5698161125183105,
      "epoch": 0.3773802654356607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992212181310235,
      "loss": 6.6332,
      "mean_token_accuracy": 0.14525358229875565,
      "num_tokens": 8258549.0,
      "step": 3270
    },
    {
      "entropy": 6.676126480102539,
      "epoch": 0.37795729948066936,
      "grad_norm": 1.203125,
      "learning_rate": 0.000499217784072477,
      "loss": 6.6169,
      "mean_token_accuracy": 0.15047752410173415,
      "num_tokens": 8271580.0,
      "step": 3275
    },
    {
      "entropy": 6.677121210098266,
      "epoch": 0.378534333525678,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992143424724768,
      "loss": 6.5612,
      "mean_token_accuracy": 0.15735053569078444,
      "num_tokens": 8283866.0,
      "step": 3280
    },
    {
      "entropy": 6.517127513885498,
      "epoch": 0.3791113675706867,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992108933311385,
      "loss": 6.5855,
      "mean_token_accuracy": 0.1442652702331543,
      "num_tokens": 8296141.0,
      "step": 3285
    },
    {
      "entropy": 6.704352283477784,
      "epoch": 0.37968840161569534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992074366485782,
      "loss": 6.5679,
      "mean_token_accuracy": 0.14811346232891082,
      "num_tokens": 8309028.0,
      "step": 3290
    },
    {
      "entropy": 6.6836034774780275,
      "epoch": 0.380265435660704,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499203972424912,
      "loss": 6.557,
      "mean_token_accuracy": 0.14885664880275726,
      "num_tokens": 8321268.0,
      "step": 3295
    },
    {
      "entropy": 6.48089861869812,
      "epoch": 0.3808424697057126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992005006602567,
      "loss": 6.5689,
      "mean_token_accuracy": 0.15438662618398666,
      "num_tokens": 8333518.0,
      "step": 3300
    },
    {
      "entropy": 6.604173278808593,
      "epoch": 0.3814195037507213,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991970213547289,
      "loss": 6.5919,
      "mean_token_accuracy": 0.15506176948547362,
      "num_tokens": 8345639.0,
      "step": 3305
    },
    {
      "entropy": 6.470817232131958,
      "epoch": 0.38199653779572995,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004991935345084457,
      "loss": 6.4928,
      "mean_token_accuracy": 0.15436709225177764,
      "num_tokens": 8358136.0,
      "step": 3310
    },
    {
      "entropy": 6.663488578796387,
      "epoch": 0.3825735718407386,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991900401215243,
      "loss": 6.6646,
      "mean_token_accuracy": 0.13878891617059708,
      "num_tokens": 8370480.0,
      "step": 3315
    },
    {
      "entropy": 6.650869083404541,
      "epoch": 0.3831506058857473,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991865381940822,
      "loss": 6.565,
      "mean_token_accuracy": 0.1526940122246742,
      "num_tokens": 8382894.0,
      "step": 3320
    },
    {
      "entropy": 6.598002243041992,
      "epoch": 0.3837276399307559,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991830287262374,
      "loss": 6.6647,
      "mean_token_accuracy": 0.14782873764634133,
      "num_tokens": 8395864.0,
      "step": 3325
    },
    {
      "entropy": 6.727650737762451,
      "epoch": 0.38430467397576457,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991795117181077,
      "loss": 6.6743,
      "mean_token_accuracy": 0.14051376208662986,
      "num_tokens": 8408242.0,
      "step": 3330
    },
    {
      "entropy": 6.642122983932495,
      "epoch": 0.3848817080207732,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004991759871698113,
      "loss": 6.557,
      "mean_token_accuracy": 0.1540223702788353,
      "num_tokens": 8420677.0,
      "step": 3335
    },
    {
      "entropy": 6.50219759941101,
      "epoch": 0.3854587420657819,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991724550814671,
      "loss": 6.5798,
      "mean_token_accuracy": 0.15331998690962792,
      "num_tokens": 8434505.0,
      "step": 3340
    },
    {
      "entropy": 6.6981383800506595,
      "epoch": 0.38603577611079054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991689154531937,
      "loss": 6.6211,
      "mean_token_accuracy": 0.1466370016336441,
      "num_tokens": 8447393.0,
      "step": 3345
    },
    {
      "entropy": 6.633738088607788,
      "epoch": 0.3866128101557992,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991653682851102,
      "loss": 6.515,
      "mean_token_accuracy": 0.1519007369875908,
      "num_tokens": 8459659.0,
      "step": 3350
    },
    {
      "entropy": 6.551579332351684,
      "epoch": 0.3871898442008078,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499161813577336,
      "loss": 6.6471,
      "mean_token_accuracy": 0.1430330380797386,
      "num_tokens": 8473783.0,
      "step": 3355
    },
    {
      "entropy": 6.563436841964721,
      "epoch": 0.3877668782458165,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991582513299903,
      "loss": 6.4393,
      "mean_token_accuracy": 0.14940204918384553,
      "num_tokens": 8486532.0,
      "step": 3360
    },
    {
      "entropy": 6.553758382797241,
      "epoch": 0.38834391229082516,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004991546815431932,
      "loss": 6.6361,
      "mean_token_accuracy": 0.14532858952879907,
      "num_tokens": 8499133.0,
      "step": 3365
    },
    {
      "entropy": 6.683175468444825,
      "epoch": 0.3889209463358338,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499151104217065,
      "loss": 6.6616,
      "mean_token_accuracy": 0.1406083531677723,
      "num_tokens": 8513391.0,
      "step": 3370
    },
    {
      "entropy": 6.639142322540283,
      "epoch": 0.3894979803808425,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004991475193517254,
      "loss": 6.59,
      "mean_token_accuracy": 0.14977881610393523,
      "num_tokens": 8527191.0,
      "step": 3375
    },
    {
      "entropy": 6.645763540267945,
      "epoch": 0.39007501442585113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991439269472956,
      "loss": 6.646,
      "mean_token_accuracy": 0.1420146033167839,
      "num_tokens": 8541328.0,
      "step": 3380
    },
    {
      "entropy": 6.622199296951294,
      "epoch": 0.39065204847085977,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991403270038961,
      "loss": 6.5389,
      "mean_token_accuracy": 0.14269101321697236,
      "num_tokens": 8553381.0,
      "step": 3385
    },
    {
      "entropy": 6.570790338516235,
      "epoch": 0.3912290825158684,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499136719521648,
      "loss": 6.578,
      "mean_token_accuracy": 0.1472961574792862,
      "num_tokens": 8565149.0,
      "step": 3390
    },
    {
      "entropy": 6.591652965545654,
      "epoch": 0.3918061165608771,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004991331045006726,
      "loss": 6.5872,
      "mean_token_accuracy": 0.1539955124258995,
      "num_tokens": 8576767.0,
      "step": 3395
    },
    {
      "entropy": 6.676447868347168,
      "epoch": 0.39238315060588574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991294819410916,
      "loss": 6.6427,
      "mean_token_accuracy": 0.14463468939065932,
      "num_tokens": 8589948.0,
      "step": 3400
    },
    {
      "entropy": 6.538165950775147,
      "epoch": 0.3929601846508944,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991258518430268,
      "loss": 6.4792,
      "mean_token_accuracy": 0.15250100940465927,
      "num_tokens": 8602233.0,
      "step": 3405
    },
    {
      "entropy": 6.595015525817871,
      "epoch": 0.3935372186959031,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004991222142066003,
      "loss": 6.5569,
      "mean_token_accuracy": 0.14572182223200797,
      "num_tokens": 8614938.0,
      "step": 3410
    },
    {
      "entropy": 6.681994009017944,
      "epoch": 0.3941142527409117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991185690319345,
      "loss": 6.6082,
      "mean_token_accuracy": 0.1508318766951561,
      "num_tokens": 8628203.0,
      "step": 3415
    },
    {
      "entropy": 6.562502431869507,
      "epoch": 0.39469128678592036,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499114916319152,
      "loss": 6.5623,
      "mean_token_accuracy": 0.15397722721099855,
      "num_tokens": 8640959.0,
      "step": 3420
    },
    {
      "entropy": 6.612614393234253,
      "epoch": 0.395268320830929,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991112560683755,
      "loss": 6.5933,
      "mean_token_accuracy": 0.1472369372844696,
      "num_tokens": 8656995.0,
      "step": 3425
    },
    {
      "entropy": 6.612496614456177,
      "epoch": 0.3958453548759377,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004991075882797283,
      "loss": 6.5999,
      "mean_token_accuracy": 0.14946889877319336,
      "num_tokens": 8671476.0,
      "step": 3430
    },
    {
      "entropy": 6.614114093780517,
      "epoch": 0.39642238892094633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004991039129533337,
      "loss": 6.6212,
      "mean_token_accuracy": 0.1447447918355465,
      "num_tokens": 8683667.0,
      "step": 3435
    },
    {
      "entropy": 6.663765811920166,
      "epoch": 0.396999422965955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991002300893152,
      "loss": 6.6206,
      "mean_token_accuracy": 0.1453992933034897,
      "num_tokens": 8696564.0,
      "step": 3440
    },
    {
      "entropy": 6.656809711456299,
      "epoch": 0.39757645701096367,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004990965396877969,
      "loss": 6.5592,
      "mean_token_accuracy": 0.145858995616436,
      "num_tokens": 8708512.0,
      "step": 3445
    },
    {
      "entropy": 6.593225526809692,
      "epoch": 0.3981534910559723,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990928417489027,
      "loss": 6.4353,
      "mean_token_accuracy": 0.16177331209182738,
      "num_tokens": 8721015.0,
      "step": 3450
    },
    {
      "entropy": 6.55296311378479,
      "epoch": 0.39873052510098095,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004990891362727572,
      "loss": 6.5552,
      "mean_token_accuracy": 0.1485207624733448,
      "num_tokens": 8732538.0,
      "step": 3455
    },
    {
      "entropy": 6.595230627059936,
      "epoch": 0.3993075591459896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990854232594848,
      "loss": 6.6106,
      "mean_token_accuracy": 0.141809394210577,
      "num_tokens": 8745093.0,
      "step": 3460
    },
    {
      "entropy": 6.650364685058594,
      "epoch": 0.3998845931909983,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990817027092106,
      "loss": 6.566,
      "mean_token_accuracy": 0.1506790906190872,
      "num_tokens": 8757441.0,
      "step": 3465
    },
    {
      "entropy": 6.536897468566894,
      "epoch": 0.4004616272360069,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990779746220595,
      "loss": 6.5875,
      "mean_token_accuracy": 0.15096076875925063,
      "num_tokens": 8770307.0,
      "step": 3470
    },
    {
      "entropy": 6.628611469268799,
      "epoch": 0.40103866128101556,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990742389981572,
      "loss": 6.5985,
      "mean_token_accuracy": 0.15109818130731584,
      "num_tokens": 8782575.0,
      "step": 3475
    },
    {
      "entropy": 6.702537965774536,
      "epoch": 0.40161569532602426,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499070495837629,
      "loss": 6.645,
      "mean_token_accuracy": 0.14216657131910324,
      "num_tokens": 8794962.0,
      "step": 3480
    },
    {
      "entropy": 6.64976396560669,
      "epoch": 0.4021927293710329,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990667451406012,
      "loss": 6.6286,
      "mean_token_accuracy": 0.1465877667069435,
      "num_tokens": 8807434.0,
      "step": 3485
    },
    {
      "entropy": 6.664065790176392,
      "epoch": 0.40276976341604154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990629869071995,
      "loss": 6.6134,
      "mean_token_accuracy": 0.14227251559495926,
      "num_tokens": 8820748.0,
      "step": 3490
    },
    {
      "entropy": 6.563335418701172,
      "epoch": 0.4033467974610502,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990592211375506,
      "loss": 6.5368,
      "mean_token_accuracy": 0.15409625843167304,
      "num_tokens": 8833428.0,
      "step": 3495
    },
    {
      "entropy": 6.604947519302368,
      "epoch": 0.4039238315060589,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990554478317811,
      "loss": 6.5739,
      "mean_token_accuracy": 0.14540679156780242,
      "num_tokens": 8845892.0,
      "step": 3500
    },
    {
      "entropy": 6.532162523269653,
      "epoch": 0.4045008655510675,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004990516669900179,
      "loss": 6.5188,
      "mean_token_accuracy": 0.15271870493888856,
      "num_tokens": 8857508.0,
      "step": 3505
    },
    {
      "entropy": 6.578783464431763,
      "epoch": 0.40507789959607615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990478786123882,
      "loss": 6.5523,
      "mean_token_accuracy": 0.14996312260627748,
      "num_tokens": 8868730.0,
      "step": 3510
    },
    {
      "entropy": 6.572446632385254,
      "epoch": 0.40565493364108485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990440826990193,
      "loss": 6.5346,
      "mean_token_accuracy": 0.1447156012058258,
      "num_tokens": 8882687.0,
      "step": 3515
    },
    {
      "entropy": 6.566001081466675,
      "epoch": 0.4062319676860935,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990402792500392,
      "loss": 6.5601,
      "mean_token_accuracy": 0.15300334095954896,
      "num_tokens": 8894723.0,
      "step": 3520
    },
    {
      "entropy": 6.5859610080719,
      "epoch": 0.4068090017311021,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990364682655754,
      "loss": 6.5359,
      "mean_token_accuracy": 0.15275930017232894,
      "num_tokens": 8906373.0,
      "step": 3525
    },
    {
      "entropy": 6.613490200042724,
      "epoch": 0.40738603577611077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990326497457564,
      "loss": 6.5407,
      "mean_token_accuracy": 0.1478450857102871,
      "num_tokens": 8918681.0,
      "step": 3530
    },
    {
      "entropy": 6.597956323623658,
      "epoch": 0.40796306982111946,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990288236907104,
      "loss": 6.5754,
      "mean_token_accuracy": 0.1443031385540962,
      "num_tokens": 8930232.0,
      "step": 3535
    },
    {
      "entropy": 6.596869373321534,
      "epoch": 0.4085401038661281,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990249901005661,
      "loss": 6.4973,
      "mean_token_accuracy": 0.1485630303621292,
      "num_tokens": 8942264.0,
      "step": 3540
    },
    {
      "entropy": 6.6454973220825195,
      "epoch": 0.40911713791113674,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004990211489754527,
      "loss": 6.6616,
      "mean_token_accuracy": 0.1454596146941185,
      "num_tokens": 8955679.0,
      "step": 3545
    },
    {
      "entropy": 6.612455701828003,
      "epoch": 0.40969417195614544,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990173003154993,
      "loss": 6.4888,
      "mean_token_accuracy": 0.15279372185468673,
      "num_tokens": 8968421.0,
      "step": 3550
    },
    {
      "entropy": 6.504389810562134,
      "epoch": 0.4102712060011541,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990134441208351,
      "loss": 6.5535,
      "mean_token_accuracy": 0.1467677101492882,
      "num_tokens": 8981063.0,
      "step": 3555
    },
    {
      "entropy": 6.6303013324737545,
      "epoch": 0.4108482400461627,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990095803915901,
      "loss": 6.5088,
      "mean_token_accuracy": 0.1581997647881508,
      "num_tokens": 8993740.0,
      "step": 3560
    },
    {
      "entropy": 6.543606472015381,
      "epoch": 0.41142527409117136,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990057091278942,
      "loss": 6.541,
      "mean_token_accuracy": 0.14845392405986785,
      "num_tokens": 9007135.0,
      "step": 3565
    },
    {
      "entropy": 6.554729413986206,
      "epoch": 0.41200230813618005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990018303298773,
      "loss": 6.5508,
      "mean_token_accuracy": 0.15105100274085997,
      "num_tokens": 9021596.0,
      "step": 3570
    },
    {
      "entropy": 6.514636564254761,
      "epoch": 0.4125793421811887,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004989979439976703,
      "loss": 6.4981,
      "mean_token_accuracy": 0.1580370619893074,
      "num_tokens": 9033523.0,
      "step": 3575
    },
    {
      "entropy": 6.483329582214355,
      "epoch": 0.41315637622619733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989940501314037,
      "loss": 6.4665,
      "mean_token_accuracy": 0.15388141348958015,
      "num_tokens": 9045690.0,
      "step": 3580
    },
    {
      "entropy": 6.582033348083496,
      "epoch": 0.413733410271206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989901487312084,
      "loss": 6.5432,
      "mean_token_accuracy": 0.15002324283123017,
      "num_tokens": 9057672.0,
      "step": 3585
    },
    {
      "entropy": 6.550716686248779,
      "epoch": 0.41431044431621467,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989862397972157,
      "loss": 6.5991,
      "mean_token_accuracy": 0.14664729684591293,
      "num_tokens": 9070004.0,
      "step": 3590
    },
    {
      "entropy": 6.661472129821777,
      "epoch": 0.4148874783612233,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989823233295572,
      "loss": 6.5681,
      "mean_token_accuracy": 0.15296779274940492,
      "num_tokens": 9082453.0,
      "step": 3595
    },
    {
      "entropy": 6.531029796600341,
      "epoch": 0.41546451240623195,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989783993283644,
      "loss": 6.5736,
      "mean_token_accuracy": 0.14416550397872924,
      "num_tokens": 9095839.0,
      "step": 3600
    },
    {
      "entropy": 6.591415214538574,
      "epoch": 0.41604154645124064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989744677937694,
      "loss": 6.5494,
      "mean_token_accuracy": 0.1541710987687111,
      "num_tokens": 9109476.0,
      "step": 3605
    },
    {
      "entropy": 6.541845417022705,
      "epoch": 0.4166185804962493,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989705287259046,
      "loss": 6.4685,
      "mean_token_accuracy": 0.15512136816978456,
      "num_tokens": 9122354.0,
      "step": 3610
    },
    {
      "entropy": 6.513101053237915,
      "epoch": 0.4171956145412579,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989665821249021,
      "loss": 6.4933,
      "mean_token_accuracy": 0.15538508147001268,
      "num_tokens": 9135502.0,
      "step": 3615
    },
    {
      "entropy": 6.614999198913575,
      "epoch": 0.4177726485862666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000498962627990895,
      "loss": 6.6489,
      "mean_token_accuracy": 0.14729145020246506,
      "num_tokens": 9147654.0,
      "step": 3620
    },
    {
      "entropy": 6.556717014312744,
      "epoch": 0.41834968263127525,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989586663240161,
      "loss": 6.5439,
      "mean_token_accuracy": 0.14543596506118775,
      "num_tokens": 9161907.0,
      "step": 3625
    },
    {
      "entropy": 6.6760279655456545,
      "epoch": 0.4189267166762839,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989546971243988,
      "loss": 6.5879,
      "mean_token_accuracy": 0.14227512627840042,
      "num_tokens": 9175395.0,
      "step": 3630
    },
    {
      "entropy": 6.621808767318726,
      "epoch": 0.41950375072129253,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989507203921763,
      "loss": 6.4972,
      "mean_token_accuracy": 0.15439630895853043,
      "num_tokens": 9187688.0,
      "step": 3635
    },
    {
      "entropy": 6.500979614257813,
      "epoch": 0.42008078476630123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989467361274828,
      "loss": 6.5274,
      "mean_token_accuracy": 0.15244155377149582,
      "num_tokens": 9199773.0,
      "step": 3640
    },
    {
      "entropy": 6.555854415893554,
      "epoch": 0.42065781881130987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989427443304519,
      "loss": 6.5157,
      "mean_token_accuracy": 0.15048773884773253,
      "num_tokens": 9213997.0,
      "step": 3645
    },
    {
      "entropy": 6.564808464050293,
      "epoch": 0.4212348528563185,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498938745001218,
      "loss": 6.5537,
      "mean_token_accuracy": 0.14768253713846208,
      "num_tokens": 9227187.0,
      "step": 3650
    },
    {
      "entropy": 6.5995463848114015,
      "epoch": 0.4218118869013272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989347381399158,
      "loss": 6.6187,
      "mean_token_accuracy": 0.1467508353292942,
      "num_tokens": 9240860.0,
      "step": 3655
    },
    {
      "entropy": 6.555819511413574,
      "epoch": 0.42238892094633584,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989307237466799,
      "loss": 6.5009,
      "mean_token_accuracy": 0.15100916624069213,
      "num_tokens": 9253771.0,
      "step": 3660
    },
    {
      "entropy": 6.555511999130249,
      "epoch": 0.4229659549913445,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004989267018216453,
      "loss": 6.5451,
      "mean_token_accuracy": 0.1484347626566887,
      "num_tokens": 9268380.0,
      "step": 3665
    },
    {
      "entropy": 6.649966144561768,
      "epoch": 0.4235429890363531,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004989226723649473,
      "loss": 6.5224,
      "mean_token_accuracy": 0.1520727753639221,
      "num_tokens": 9279911.0,
      "step": 3670
    },
    {
      "entropy": 6.573756361007691,
      "epoch": 0.4241200230813618,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989186353767214,
      "loss": 6.5251,
      "mean_token_accuracy": 0.15676866322755814,
      "num_tokens": 9292105.0,
      "step": 3675
    },
    {
      "entropy": 6.587281703948975,
      "epoch": 0.42469705712637046,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004989145908571035,
      "loss": 6.6282,
      "mean_token_accuracy": 0.1454088181257248,
      "num_tokens": 9305503.0,
      "step": 3680
    },
    {
      "entropy": 6.598947048187256,
      "epoch": 0.4252740911713791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989105388062295,
      "loss": 6.5304,
      "mean_token_accuracy": 0.14965798407793046,
      "num_tokens": 9317978.0,
      "step": 3685
    },
    {
      "entropy": 6.61584882736206,
      "epoch": 0.4258511252163878,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989064792242358,
      "loss": 6.6039,
      "mean_token_accuracy": 0.14742101579904557,
      "num_tokens": 9332676.0,
      "step": 3690
    },
    {
      "entropy": 6.577369356155396,
      "epoch": 0.42642815926139643,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004989024121112589,
      "loss": 6.5021,
      "mean_token_accuracy": 0.150289623439312,
      "num_tokens": 9345594.0,
      "step": 3695
    },
    {
      "entropy": 6.637693548202515,
      "epoch": 0.4270051933064051,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004988983374674356,
      "loss": 6.5817,
      "mean_token_accuracy": 0.14303911998867988,
      "num_tokens": 9358303.0,
      "step": 3700
    },
    {
      "entropy": 6.525368690490723,
      "epoch": 0.4275822273514137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988942552929029,
      "loss": 6.4649,
      "mean_token_accuracy": 0.15538589656352997,
      "num_tokens": 9370953.0,
      "step": 3705
    },
    {
      "entropy": 6.450020742416382,
      "epoch": 0.4281592613964224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988901655877981,
      "loss": 6.5505,
      "mean_token_accuracy": 0.14808087795972824,
      "num_tokens": 9384563.0,
      "step": 3710
    },
    {
      "entropy": 6.571639585494995,
      "epoch": 0.42873629544143105,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004988860683522589,
      "loss": 6.4914,
      "mean_token_accuracy": 0.15418324396014213,
      "num_tokens": 9396401.0,
      "step": 3715
    },
    {
      "entropy": 6.549563884735107,
      "epoch": 0.4293133294864397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498881963586423,
      "loss": 6.5066,
      "mean_token_accuracy": 0.15194836109876633,
      "num_tokens": 9408520.0,
      "step": 3720
    },
    {
      "entropy": 6.4020223140716555,
      "epoch": 0.4298903635314484,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988778512904282,
      "loss": 6.4085,
      "mean_token_accuracy": 0.1552750959992409,
      "num_tokens": 9421431.0,
      "step": 3725
    },
    {
      "entropy": 6.5411797046661375,
      "epoch": 0.430467397576457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988737314644134,
      "loss": 6.5322,
      "mean_token_accuracy": 0.15222309082746505,
      "num_tokens": 9432566.0,
      "step": 3730
    },
    {
      "entropy": 6.651175785064697,
      "epoch": 0.43104443162146566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988696041085168,
      "loss": 6.6239,
      "mean_token_accuracy": 0.14554398953914643,
      "num_tokens": 9444885.0,
      "step": 3735
    },
    {
      "entropy": 6.622100782394409,
      "epoch": 0.4316214656664743,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004988654692228772,
      "loss": 6.6173,
      "mean_token_accuracy": 0.14461512714624405,
      "num_tokens": 9457823.0,
      "step": 3740
    },
    {
      "entropy": 6.5988099575042725,
      "epoch": 0.432198499711483,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004988613268076335,
      "loss": 6.467,
      "mean_token_accuracy": 0.1583518996834755,
      "num_tokens": 9470799.0,
      "step": 3745
    },
    {
      "entropy": 6.589520645141602,
      "epoch": 0.43277553375649164,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004988571768629257,
      "loss": 6.5479,
      "mean_token_accuracy": 0.15336680859327317,
      "num_tokens": 9484175.0,
      "step": 3750
    },
    {
      "entropy": 6.627070426940918,
      "epoch": 0.4333525678015003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004988530193888927,
      "loss": 6.5747,
      "mean_token_accuracy": 0.14501228407025338,
      "num_tokens": 9497809.0,
      "step": 3755
    },
    {
      "entropy": 6.5905872821807865,
      "epoch": 0.4339296018465089,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988488543856747,
      "loss": 6.5109,
      "mean_token_accuracy": 0.14952886253595352,
      "num_tokens": 9509755.0,
      "step": 3760
    },
    {
      "entropy": 6.532172775268554,
      "epoch": 0.4345066358915176,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988446818534115,
      "loss": 6.5331,
      "mean_token_accuracy": 0.14465863406658172,
      "num_tokens": 9523801.0,
      "step": 3765
    },
    {
      "entropy": 6.547245168685913,
      "epoch": 0.43508366993652625,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004988405017922438,
      "loss": 6.5018,
      "mean_token_accuracy": 0.1546433985233307,
      "num_tokens": 9535846.0,
      "step": 3770
    },
    {
      "entropy": 6.636412525177002,
      "epoch": 0.4356607039815349,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498836314202312,
      "loss": 6.5842,
      "mean_token_accuracy": 0.14722789376974105,
      "num_tokens": 9548194.0,
      "step": 3775
    },
    {
      "entropy": 6.50278000831604,
      "epoch": 0.4362377380265436,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988321190837568,
      "loss": 6.499,
      "mean_token_accuracy": 0.1481459118425846,
      "num_tokens": 9561069.0,
      "step": 3780
    },
    {
      "entropy": 6.522726202011109,
      "epoch": 0.4368147720715522,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988279164367196,
      "loss": 6.4527,
      "mean_token_accuracy": 0.15349378883838655,
      "num_tokens": 9573778.0,
      "step": 3785
    },
    {
      "entropy": 6.55922565460205,
      "epoch": 0.43739180611656087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004988237062613415,
      "loss": 6.4631,
      "mean_token_accuracy": 0.1588495686650276,
      "num_tokens": 9586616.0,
      "step": 3790
    },
    {
      "entropy": 6.531683540344238,
      "epoch": 0.4379688401615695,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004988194885577644,
      "loss": 6.5371,
      "mean_token_accuracy": 0.15208741575479506,
      "num_tokens": 9597943.0,
      "step": 3795
    },
    {
      "entropy": 6.567034959793091,
      "epoch": 0.4385458742065782,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988152633261299,
      "loss": 6.4738,
      "mean_token_accuracy": 0.15672459453344345,
      "num_tokens": 9609861.0,
      "step": 3800
    },
    {
      "entropy": 6.538378524780273,
      "epoch": 0.43912290825158684,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049881103056658,
      "loss": 6.5315,
      "mean_token_accuracy": 0.1557203412055969,
      "num_tokens": 9623307.0,
      "step": 3805
    },
    {
      "entropy": 6.5774413585662845,
      "epoch": 0.4396999422965955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988067902792575,
      "loss": 6.548,
      "mean_token_accuracy": 0.14982074946165086,
      "num_tokens": 9635827.0,
      "step": 3810
    },
    {
      "entropy": 6.617994737625122,
      "epoch": 0.4402769763416042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988025424643047,
      "loss": 6.5941,
      "mean_token_accuracy": 0.1468472957611084,
      "num_tokens": 9649422.0,
      "step": 3815
    },
    {
      "entropy": 6.6371983051300045,
      "epoch": 0.4408540103866128,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987982871218645,
      "loss": 6.5795,
      "mean_token_accuracy": 0.15525637120008468,
      "num_tokens": 9661542.0,
      "step": 3820
    },
    {
      "entropy": 6.573576974868774,
      "epoch": 0.44143104443162146,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987940242520802,
      "loss": 6.563,
      "mean_token_accuracy": 0.14920065850019454,
      "num_tokens": 9674684.0,
      "step": 3825
    },
    {
      "entropy": 6.644643783569336,
      "epoch": 0.4420080784766301,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498789753855095,
      "loss": 6.5816,
      "mean_token_accuracy": 0.14375625625252725,
      "num_tokens": 9687522.0,
      "step": 3830
    },
    {
      "entropy": 6.556054496765137,
      "epoch": 0.4425851125216388,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987854759310526,
      "loss": 6.5199,
      "mean_token_accuracy": 0.15176642686128616,
      "num_tokens": 9699518.0,
      "step": 3835
    },
    {
      "entropy": 6.550537872314453,
      "epoch": 0.44316214656664743,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987811904800968,
      "loss": 6.5914,
      "mean_token_accuracy": 0.15061496943235397,
      "num_tokens": 9711918.0,
      "step": 3840
    },
    {
      "entropy": 6.618970155715942,
      "epoch": 0.44373918061165607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987768975023719,
      "loss": 6.4637,
      "mean_token_accuracy": 0.1534193515777588,
      "num_tokens": 9725643.0,
      "step": 3845
    },
    {
      "entropy": 6.508399486541748,
      "epoch": 0.44431621465666477,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004987725969980222,
      "loss": 6.4791,
      "mean_token_accuracy": 0.15461074262857438,
      "num_tokens": 9738871.0,
      "step": 3850
    },
    {
      "entropy": 6.529386281967163,
      "epoch": 0.4448932487016734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987682889671923,
      "loss": 6.4722,
      "mean_token_accuracy": 0.15544941425323486,
      "num_tokens": 9751166.0,
      "step": 3855
    },
    {
      "entropy": 6.528097200393677,
      "epoch": 0.44547028274668204,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004987639734100272,
      "loss": 6.4395,
      "mean_token_accuracy": 0.1568680375814438,
      "num_tokens": 9765165.0,
      "step": 3860
    },
    {
      "entropy": 6.498267030715942,
      "epoch": 0.4460473167916907,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987596503266721,
      "loss": 6.485,
      "mean_token_accuracy": 0.1533951610326767,
      "num_tokens": 9776625.0,
      "step": 3865
    },
    {
      "entropy": 6.5177834033966064,
      "epoch": 0.4466243508366994,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987553197172722,
      "loss": 6.4855,
      "mean_token_accuracy": 0.1559309795498848,
      "num_tokens": 9789689.0,
      "step": 3870
    },
    {
      "entropy": 6.627424955368042,
      "epoch": 0.447201384881708,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987509815819732,
      "loss": 6.6022,
      "mean_token_accuracy": 0.14655733555555345,
      "num_tokens": 9801124.0,
      "step": 3875
    },
    {
      "entropy": 6.550175523757934,
      "epoch": 0.44777841892671666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004987466359209213,
      "loss": 6.5342,
      "mean_token_accuracy": 0.15006100833415986,
      "num_tokens": 9814264.0,
      "step": 3880
    },
    {
      "entropy": 6.550042581558228,
      "epoch": 0.44835545297172535,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987422827342622,
      "loss": 6.4655,
      "mean_token_accuracy": 0.15616977363824844,
      "num_tokens": 9826149.0,
      "step": 3885
    },
    {
      "entropy": 6.557639408111572,
      "epoch": 0.448932487016734,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987379220221426,
      "loss": 6.5503,
      "mean_token_accuracy": 0.14510683268308638,
      "num_tokens": 9838485.0,
      "step": 3890
    },
    {
      "entropy": 6.5344490051269535,
      "epoch": 0.44950952106174263,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987335537847092,
      "loss": 6.485,
      "mean_token_accuracy": 0.15009704530239104,
      "num_tokens": 9851371.0,
      "step": 3895
    },
    {
      "entropy": 6.512049579620362,
      "epoch": 0.4500865551067513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004987291780221088,
      "loss": 6.5257,
      "mean_token_accuracy": 0.14940400570631027,
      "num_tokens": 9864371.0,
      "step": 3900
    },
    {
      "entropy": 6.560714340209961,
      "epoch": 0.45066358915175997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987247947344887,
      "loss": 6.5233,
      "mean_token_accuracy": 0.1517380103468895,
      "num_tokens": 9876086.0,
      "step": 3905
    },
    {
      "entropy": 6.528545999526978,
      "epoch": 0.4512406231967686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987204039219962,
      "loss": 6.5254,
      "mean_token_accuracy": 0.1479131981730461,
      "num_tokens": 9888850.0,
      "step": 3910
    },
    {
      "entropy": 6.582427883148194,
      "epoch": 0.45181765724177725,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004987160055847791,
      "loss": 6.5666,
      "mean_token_accuracy": 0.15192154794931412,
      "num_tokens": 9901467.0,
      "step": 3915
    },
    {
      "entropy": 6.58316740989685,
      "epoch": 0.45239469128678594,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987115997229852,
      "loss": 6.5235,
      "mean_token_accuracy": 0.1524689018726349,
      "num_tokens": 9914431.0,
      "step": 3920
    },
    {
      "entropy": 6.614464235305786,
      "epoch": 0.4529717253317946,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987071863367629,
      "loss": 6.5604,
      "mean_token_accuracy": 0.15112543925642968,
      "num_tokens": 9926863.0,
      "step": 3925
    },
    {
      "entropy": 6.565920209884643,
      "epoch": 0.4535487593768032,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004987027654262604,
      "loss": 6.5035,
      "mean_token_accuracy": 0.15478066951036454,
      "num_tokens": 9939064.0,
      "step": 3930
    },
    {
      "entropy": 6.430483436584472,
      "epoch": 0.45412579342181186,
      "grad_norm": 1.125,
      "learning_rate": 0.0004986983369916264,
      "loss": 6.4946,
      "mean_token_accuracy": 0.15007816404104232,
      "num_tokens": 9950897.0,
      "step": 3935
    },
    {
      "entropy": 6.655666780471802,
      "epoch": 0.45470282746682056,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004986939010330102,
      "loss": 6.5449,
      "mean_token_accuracy": 0.1454168014228344,
      "num_tokens": 9964388.0,
      "step": 3940
    },
    {
      "entropy": 6.409772491455078,
      "epoch": 0.4552798615118292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986894575505606,
      "loss": 6.382,
      "mean_token_accuracy": 0.16272770017385482,
      "num_tokens": 9977363.0,
      "step": 3945
    },
    {
      "entropy": 6.591909456253052,
      "epoch": 0.45585689555683784,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986850065444272,
      "loss": 6.4427,
      "mean_token_accuracy": 0.15807752758264543,
      "num_tokens": 9989691.0,
      "step": 3950
    },
    {
      "entropy": 6.590207910537719,
      "epoch": 0.45643392960184653,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986805480147598,
      "loss": 6.4787,
      "mean_token_accuracy": 0.14620172828435898,
      "num_tokens": 10002177.0,
      "step": 3955
    },
    {
      "entropy": 6.495525979995728,
      "epoch": 0.4570109636468552,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004986760819617082,
      "loss": 6.5341,
      "mean_token_accuracy": 0.14582275450229645,
      "num_tokens": 10014844.0,
      "step": 3960
    },
    {
      "entropy": 6.659038019180298,
      "epoch": 0.4575879976918638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986716083854228,
      "loss": 6.5687,
      "mean_token_accuracy": 0.1485973134636879,
      "num_tokens": 10028351.0,
      "step": 3965
    },
    {
      "entropy": 6.487327241897583,
      "epoch": 0.45816503173687245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986671272860538,
      "loss": 6.4832,
      "mean_token_accuracy": 0.14729729741811753,
      "num_tokens": 10041318.0,
      "step": 3970
    },
    {
      "entropy": 6.540215110778808,
      "epoch": 0.45874206578188115,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004986626386637521,
      "loss": 6.4811,
      "mean_token_accuracy": 0.14569370001554488,
      "num_tokens": 10054985.0,
      "step": 3975
    },
    {
      "entropy": 6.571915769577027,
      "epoch": 0.4593190998268898,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004986581425186688,
      "loss": 6.5598,
      "mean_token_accuracy": 0.1512555092573166,
      "num_tokens": 10067528.0,
      "step": 3980
    },
    {
      "entropy": 6.542682886123657,
      "epoch": 0.4598961338718984,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986536388509548,
      "loss": 6.4675,
      "mean_token_accuracy": 0.15116373002529143,
      "num_tokens": 10080210.0,
      "step": 3985
    },
    {
      "entropy": 6.51995530128479,
      "epoch": 0.4604731679169071,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004986491276607618,
      "loss": 6.5335,
      "mean_token_accuracy": 0.15148406326770783,
      "num_tokens": 10092040.0,
      "step": 3990
    },
    {
      "entropy": 6.583942508697509,
      "epoch": 0.46105020196191576,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004986446089482416,
      "loss": 6.4508,
      "mean_token_accuracy": 0.15163778215646745,
      "num_tokens": 10104923.0,
      "step": 3995
    },
    {
      "entropy": 6.4749840259552,
      "epoch": 0.4616272360069244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986400827135459,
      "loss": 6.4746,
      "mean_token_accuracy": 0.15648574531078338,
      "num_tokens": 10118963.0,
      "step": 4000
    },
    {
      "entropy": 6.527820158004761,
      "epoch": 0.46220427005193304,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004986355489568272,
      "loss": 6.4285,
      "mean_token_accuracy": 0.15958915799856185,
      "num_tokens": 10131457.0,
      "step": 4005
    },
    {
      "entropy": 6.568796968460083,
      "epoch": 0.46278130409694174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986310076782379,
      "loss": 6.4985,
      "mean_token_accuracy": 0.1522387072443962,
      "num_tokens": 10143796.0,
      "step": 4010
    },
    {
      "entropy": 6.494457387924195,
      "epoch": 0.4633583381419504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986264588779307,
      "loss": 6.5045,
      "mean_token_accuracy": 0.15380825847387314,
      "num_tokens": 10156723.0,
      "step": 4015
    },
    {
      "entropy": 6.5013172149658205,
      "epoch": 0.463935372186959,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986219025560586,
      "loss": 6.5023,
      "mean_token_accuracy": 0.14863586351275443,
      "num_tokens": 10168701.0,
      "step": 4020
    },
    {
      "entropy": 6.557641696929932,
      "epoch": 0.4645124062319677,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498617338712775,
      "loss": 6.4602,
      "mean_token_accuracy": 0.1560764953494072,
      "num_tokens": 10182308.0,
      "step": 4025
    },
    {
      "entropy": 6.510044240951538,
      "epoch": 0.46508944027697635,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986127673482332,
      "loss": 6.4539,
      "mean_token_accuracy": 0.15890756398439407,
      "num_tokens": 10194237.0,
      "step": 4030
    },
    {
      "entropy": 6.52144136428833,
      "epoch": 0.465666474321985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986081884625871,
      "loss": 6.5215,
      "mean_token_accuracy": 0.14942877292633056,
      "num_tokens": 10208570.0,
      "step": 4035
    },
    {
      "entropy": 6.601692390441895,
      "epoch": 0.46624350836699363,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986036020559906,
      "loss": 6.4814,
      "mean_token_accuracy": 0.15481637567281722,
      "num_tokens": 10221079.0,
      "step": 4040
    },
    {
      "entropy": 6.356483602523804,
      "epoch": 0.4668205424120023,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498599008128598,
      "loss": 6.4155,
      "mean_token_accuracy": 0.15378812998533248,
      "num_tokens": 10233473.0,
      "step": 4045
    },
    {
      "entropy": 6.593820810317993,
      "epoch": 0.46739757645701097,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985944066805639,
      "loss": 6.5525,
      "mean_token_accuracy": 0.14852241575717925,
      "num_tokens": 10247840.0,
      "step": 4050
    },
    {
      "entropy": 6.525872278213501,
      "epoch": 0.4679746105020196,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000498589797712043,
      "loss": 6.473,
      "mean_token_accuracy": 0.16045304834842683,
      "num_tokens": 10261471.0,
      "step": 4055
    },
    {
      "entropy": 6.504099178314209,
      "epoch": 0.4685516445470283,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004985851812231903,
      "loss": 6.4736,
      "mean_token_accuracy": 0.15337898284196855,
      "num_tokens": 10274386.0,
      "step": 4060
    },
    {
      "entropy": 6.592873764038086,
      "epoch": 0.46912867859203694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985805572141611,
      "loss": 6.5064,
      "mean_token_accuracy": 0.15148324966430665,
      "num_tokens": 10286405.0,
      "step": 4065
    },
    {
      "entropy": 6.614938163757325,
      "epoch": 0.4697057126370456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985759256851108,
      "loss": 6.5546,
      "mean_token_accuracy": 0.14901065975427627,
      "num_tokens": 10300084.0,
      "step": 4070
    },
    {
      "entropy": 6.57010850906372,
      "epoch": 0.4702827466820542,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004985712866361954,
      "loss": 6.5524,
      "mean_token_accuracy": 0.15720511972904205,
      "num_tokens": 10312558.0,
      "step": 4075
    },
    {
      "entropy": 6.493395519256592,
      "epoch": 0.4708597807270629,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004985666400675709,
      "loss": 6.4604,
      "mean_token_accuracy": 0.1560864806175232,
      "num_tokens": 10324185.0,
      "step": 4080
    },
    {
      "entropy": 6.488433313369751,
      "epoch": 0.47143681477207156,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985619859793934,
      "loss": 6.4833,
      "mean_token_accuracy": 0.1477129116654396,
      "num_tokens": 10338634.0,
      "step": 4085
    },
    {
      "entropy": 6.7163480758667,
      "epoch": 0.4720138488170802,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985573243718195,
      "loss": 6.5503,
      "mean_token_accuracy": 0.14475530236959458,
      "num_tokens": 10351272.0,
      "step": 4090
    },
    {
      "entropy": 6.561359500885009,
      "epoch": 0.4725908828620889,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985526552450061,
      "loss": 6.5081,
      "mean_token_accuracy": 0.1437559977173805,
      "num_tokens": 10362894.0,
      "step": 4095
    },
    {
      "entropy": 6.548456525802612,
      "epoch": 0.47316791690709753,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985479785991103,
      "loss": 6.5842,
      "mean_token_accuracy": 0.14605074226856232,
      "num_tokens": 10376120.0,
      "step": 4100
    },
    {
      "entropy": 6.55332236289978,
      "epoch": 0.47374495095210617,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004985432944342892,
      "loss": 6.5026,
      "mean_token_accuracy": 0.153361913561821,
      "num_tokens": 10387986.0,
      "step": 4105
    },
    {
      "entropy": 6.548603773117065,
      "epoch": 0.4743219849971148,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004985386027507003,
      "loss": 6.5225,
      "mean_token_accuracy": 0.1517445206642151,
      "num_tokens": 10400920.0,
      "step": 4110
    },
    {
      "entropy": 6.661869049072266,
      "epoch": 0.4748990190421235,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985339035485018,
      "loss": 6.525,
      "mean_token_accuracy": 0.14937230050563813,
      "num_tokens": 10413835.0,
      "step": 4115
    },
    {
      "entropy": 6.533235025405884,
      "epoch": 0.47547605308713214,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985291968278513,
      "loss": 6.5224,
      "mean_token_accuracy": 0.14905970692634582,
      "num_tokens": 10426959.0,
      "step": 4120
    },
    {
      "entropy": 6.536430978775025,
      "epoch": 0.4760530871321408,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985244825889074,
      "loss": 6.5637,
      "mean_token_accuracy": 0.15390570759773253,
      "num_tokens": 10440319.0,
      "step": 4125
    },
    {
      "entropy": 6.541896820068359,
      "epoch": 0.4766301211771495,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004985197608318284,
      "loss": 6.5325,
      "mean_token_accuracy": 0.15225679874420167,
      "num_tokens": 10453715.0,
      "step": 4130
    },
    {
      "entropy": 6.594496631622315,
      "epoch": 0.4772071552221581,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985150315567735,
      "loss": 6.5164,
      "mean_token_accuracy": 0.14864421635866165,
      "num_tokens": 10465574.0,
      "step": 4135
    },
    {
      "entropy": 6.506676197052002,
      "epoch": 0.47778418926716676,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004985102947639013,
      "loss": 6.5007,
      "mean_token_accuracy": 0.15153668075799942,
      "num_tokens": 10478105.0,
      "step": 4140
    },
    {
      "entropy": 6.552076911926269,
      "epoch": 0.4783612233121754,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004985055504533715,
      "loss": 6.5134,
      "mean_token_accuracy": 0.15088197439908982,
      "num_tokens": 10490723.0,
      "step": 4145
    },
    {
      "entropy": 6.602122354507446,
      "epoch": 0.4789382573571841,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985007986253435,
      "loss": 6.5254,
      "mean_token_accuracy": 0.14911809265613557,
      "num_tokens": 10503155.0,
      "step": 4150
    },
    {
      "entropy": 6.554518699645996,
      "epoch": 0.47951529140219273,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004984960392799772,
      "loss": 6.5106,
      "mean_token_accuracy": 0.1461018964648247,
      "num_tokens": 10515077.0,
      "step": 4155
    },
    {
      "entropy": 6.592270565032959,
      "epoch": 0.4800923254472014,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004984912724174326,
      "loss": 6.4926,
      "mean_token_accuracy": 0.1490560546517372,
      "num_tokens": 10528445.0,
      "step": 4160
    },
    {
      "entropy": 6.542512369155884,
      "epoch": 0.48066935949221,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984864980378701,
      "loss": 6.518,
      "mean_token_accuracy": 0.15329633057117462,
      "num_tokens": 10541877.0,
      "step": 4165
    },
    {
      "entropy": 6.511092805862427,
      "epoch": 0.4812463935372187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984817161414502,
      "loss": 6.4941,
      "mean_token_accuracy": 0.15064932703971862,
      "num_tokens": 10553527.0,
      "step": 4170
    },
    {
      "entropy": 6.550889539718628,
      "epoch": 0.48182342758222735,
      "grad_norm": 1.125,
      "learning_rate": 0.0004984769267283338,
      "loss": 6.5247,
      "mean_token_accuracy": 0.14874509274959563,
      "num_tokens": 10566492.0,
      "step": 4175
    },
    {
      "entropy": 6.496022319793701,
      "epoch": 0.482400461627236,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004984721297986819,
      "loss": 6.458,
      "mean_token_accuracy": 0.15352972596883774,
      "num_tokens": 10578962.0,
      "step": 4180
    },
    {
      "entropy": 6.561640214920044,
      "epoch": 0.4829774956722447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004984673253526561,
      "loss": 6.4676,
      "mean_token_accuracy": 0.15236313343048097,
      "num_tokens": 10591254.0,
      "step": 4185
    },
    {
      "entropy": 6.533859395980835,
      "epoch": 0.4835545297172533,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004984625133904176,
      "loss": 6.5014,
      "mean_token_accuracy": 0.15217659398913383,
      "num_tokens": 10604961.0,
      "step": 4190
    },
    {
      "entropy": 6.529065370559692,
      "epoch": 0.48413156376226196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984576939121286,
      "loss": 6.4615,
      "mean_token_accuracy": 0.15530140995979308,
      "num_tokens": 10617419.0,
      "step": 4195
    },
    {
      "entropy": 6.506486940383911,
      "epoch": 0.4847085978072706,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004984528669179511,
      "loss": 6.531,
      "mean_token_accuracy": 0.14775909036397933,
      "num_tokens": 10629721.0,
      "step": 4200
    },
    {
      "entropy": 6.585139322280884,
      "epoch": 0.4852856318522793,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004984480324080472,
      "loss": 6.5338,
      "mean_token_accuracy": 0.14746304452419282,
      "num_tokens": 10643869.0,
      "step": 4205
    },
    {
      "entropy": 6.593662929534912,
      "epoch": 0.48586266589728794,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00049844319038258,
      "loss": 6.5939,
      "mean_token_accuracy": 0.14522895812988282,
      "num_tokens": 10658418.0,
      "step": 4210
    },
    {
      "entropy": 6.564258623123169,
      "epoch": 0.4864396999422966,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004984383408417119,
      "loss": 6.4647,
      "mean_token_accuracy": 0.15308941155672073,
      "num_tokens": 10671105.0,
      "step": 4215
    },
    {
      "entropy": 6.582191467285156,
      "epoch": 0.4870167339873053,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984334837856064,
      "loss": 6.5699,
      "mean_token_accuracy": 0.1451310083270073,
      "num_tokens": 10683862.0,
      "step": 4220
    },
    {
      "entropy": 6.55675687789917,
      "epoch": 0.4875937680323139,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004984286192144263,
      "loss": 6.4655,
      "mean_token_accuracy": 0.15599070638418197,
      "num_tokens": 10696559.0,
      "step": 4225
    },
    {
      "entropy": 6.599395895004273,
      "epoch": 0.48817080207732255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984237471283359,
      "loss": 6.5383,
      "mean_token_accuracy": 0.15275808423757553,
      "num_tokens": 10710227.0,
      "step": 4230
    },
    {
      "entropy": 6.507157993316651,
      "epoch": 0.4887478361223312,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004984188675274986,
      "loss": 6.5131,
      "mean_token_accuracy": 0.14976295828819275,
      "num_tokens": 10722761.0,
      "step": 4235
    },
    {
      "entropy": 6.592884159088134,
      "epoch": 0.4893248701673399,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984139804120787,
      "loss": 6.4778,
      "mean_token_accuracy": 0.1465604603290558,
      "num_tokens": 10734126.0,
      "step": 4240
    },
    {
      "entropy": 6.563351154327393,
      "epoch": 0.4899019042123485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984090857822406,
      "loss": 6.5365,
      "mean_token_accuracy": 0.14797307029366494,
      "num_tokens": 10747570.0,
      "step": 4245
    },
    {
      "entropy": 6.63136625289917,
      "epoch": 0.49047893825735717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984041836381488,
      "loss": 6.6419,
      "mean_token_accuracy": 0.14504213631153107,
      "num_tokens": 10762975.0,
      "step": 4250
    },
    {
      "entropy": 6.632890224456787,
      "epoch": 0.49105597230236586,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004983992739799682,
      "loss": 6.4833,
      "mean_token_accuracy": 0.15367807745933532,
      "num_tokens": 10774520.0,
      "step": 4255
    },
    {
      "entropy": 6.475261068344116,
      "epoch": 0.4916330063473745,
      "grad_norm": 1.125,
      "learning_rate": 0.0004983943568078642,
      "loss": 6.4071,
      "mean_token_accuracy": 0.15381105840206147,
      "num_tokens": 10786250.0,
      "step": 4260
    },
    {
      "entropy": 6.515493011474609,
      "epoch": 0.49221004039238314,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983894321220017,
      "loss": 6.5271,
      "mean_token_accuracy": 0.14624571949243545,
      "num_tokens": 10799831.0,
      "step": 4265
    },
    {
      "entropy": 6.583249425888061,
      "epoch": 0.4927870744373918,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983844999225468,
      "loss": 6.4694,
      "mean_token_accuracy": 0.15290576964616776,
      "num_tokens": 10812803.0,
      "step": 4270
    },
    {
      "entropy": 6.512081813812256,
      "epoch": 0.4933641084824005,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983795602096651,
      "loss": 6.4783,
      "mean_token_accuracy": 0.15587753057479858,
      "num_tokens": 10825150.0,
      "step": 4275
    },
    {
      "entropy": 6.589651012420655,
      "epoch": 0.4939411425274091,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498374612983523,
      "loss": 6.5828,
      "mean_token_accuracy": 0.14737853184342384,
      "num_tokens": 10838337.0,
      "step": 4280
    },
    {
      "entropy": 6.635598182678223,
      "epoch": 0.49451817657241776,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004983696582442866,
      "loss": 6.4461,
      "mean_token_accuracy": 0.16733498871326447,
      "num_tokens": 10852288.0,
      "step": 4285
    },
    {
      "entropy": 6.519201326370239,
      "epoch": 0.49509521061742645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004983646959921227,
      "loss": 6.5311,
      "mean_token_accuracy": 0.14743692576885223,
      "num_tokens": 10864189.0,
      "step": 4290
    },
    {
      "entropy": 6.6310670375823975,
      "epoch": 0.4956722446624351,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983597262271984,
      "loss": 6.5307,
      "mean_token_accuracy": 0.14658827036619188,
      "num_tokens": 10877410.0,
      "step": 4295
    },
    {
      "entropy": 6.480897855758667,
      "epoch": 0.49624927870744373,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004983547489496804,
      "loss": 6.5108,
      "mean_token_accuracy": 0.14898531660437583,
      "num_tokens": 10889757.0,
      "step": 4300
    },
    {
      "entropy": 6.511052417755127,
      "epoch": 0.49682631275245237,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004983497641597365,
      "loss": 6.399,
      "mean_token_accuracy": 0.1597002699971199,
      "num_tokens": 10901636.0,
      "step": 4305
    },
    {
      "entropy": 6.496632814407349,
      "epoch": 0.49740334679746107,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983447718575342,
      "loss": 6.4418,
      "mean_token_accuracy": 0.15892936736345292,
      "num_tokens": 10914063.0,
      "step": 4310
    },
    {
      "entropy": 6.581086921691894,
      "epoch": 0.4979803808424697,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983397720432414,
      "loss": 6.4744,
      "mean_token_accuracy": 0.15386273562908173,
      "num_tokens": 10927730.0,
      "step": 4315
    },
    {
      "entropy": 6.537801122665405,
      "epoch": 0.49855741488747835,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983347647170264,
      "loss": 6.6001,
      "mean_token_accuracy": 0.14655838459730147,
      "num_tokens": 10942440.0,
      "step": 4320
    },
    {
      "entropy": 6.567225456237793,
      "epoch": 0.49913444893248704,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004983297498790574,
      "loss": 6.4809,
      "mean_token_accuracy": 0.14959282279014588,
      "num_tokens": 10955801.0,
      "step": 4325
    },
    {
      "entropy": 6.585222673416138,
      "epoch": 0.4997114829774957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004983247275295034,
      "loss": 6.4775,
      "mean_token_accuracy": 0.15232663005590438,
      "num_tokens": 10967918.0,
      "step": 4330
    },
    {
      "entropy": 6.481248474121093,
      "epoch": 0.5002885170225043,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983196976685329,
      "loss": 6.4152,
      "mean_token_accuracy": 0.15650064051151275,
      "num_tokens": 10979652.0,
      "step": 4335
    },
    {
      "entropy": 6.49305682182312,
      "epoch": 0.500865551067513,
      "grad_norm": 2.359375,
      "learning_rate": 0.0004983146602963155,
      "loss": 6.3465,
      "mean_token_accuracy": 0.16713591665029526,
      "num_tokens": 10991433.0,
      "step": 4340
    },
    {
      "entropy": 6.41333909034729,
      "epoch": 0.5014425851125216,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004983096154130203,
      "loss": 6.5509,
      "mean_token_accuracy": 0.15196776390075684,
      "num_tokens": 11003976.0,
      "step": 4345
    },
    {
      "entropy": 6.585015773773193,
      "epoch": 0.5020196191575302,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004983045630188171,
      "loss": 6.524,
      "mean_token_accuracy": 0.1477636769413948,
      "num_tokens": 11018848.0,
      "step": 4350
    },
    {
      "entropy": 6.6534788608551025,
      "epoch": 0.502596653202539,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982995031138759,
      "loss": 6.5037,
      "mean_token_accuracy": 0.1479504019021988,
      "num_tokens": 11030803.0,
      "step": 4355
    },
    {
      "entropy": 6.515288352966309,
      "epoch": 0.5031736872475476,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982944356983666,
      "loss": 6.5648,
      "mean_token_accuracy": 0.14597940295934678,
      "num_tokens": 11044688.0,
      "step": 4360
    },
    {
      "entropy": 6.573613023757934,
      "epoch": 0.5037507212925563,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982893607724602,
      "loss": 6.5372,
      "mean_token_accuracy": 0.1448674187064171,
      "num_tokens": 11057026.0,
      "step": 4365
    },
    {
      "entropy": 6.661010122299194,
      "epoch": 0.5043277553375649,
      "grad_norm": 1.703125,
      "learning_rate": 0.000498284278336327,
      "loss": 6.5456,
      "mean_token_accuracy": 0.14358318001031875,
      "num_tokens": 11070407.0,
      "step": 4370
    },
    {
      "entropy": 6.476870441436768,
      "epoch": 0.5049047893825735,
      "grad_norm": 1.25,
      "learning_rate": 0.0004982791883901379,
      "loss": 6.4595,
      "mean_token_accuracy": 0.15498376861214638,
      "num_tokens": 11082655.0,
      "step": 4375
    },
    {
      "entropy": 6.414759540557862,
      "epoch": 0.5054818234275822,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982740909340643,
      "loss": 6.4234,
      "mean_token_accuracy": 0.16286834478378295,
      "num_tokens": 11095948.0,
      "step": 4380
    },
    {
      "entropy": 6.6195331573486325,
      "epoch": 0.5060588574725908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982689859682775,
      "loss": 6.467,
      "mean_token_accuracy": 0.14826145470142366,
      "num_tokens": 11108105.0,
      "step": 4385
    },
    {
      "entropy": 6.554189395904541,
      "epoch": 0.5066358915175996,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004982638734929492,
      "loss": 6.5182,
      "mean_token_accuracy": 0.14397745132446288,
      "num_tokens": 11120322.0,
      "step": 4390
    },
    {
      "entropy": 6.4076619148254395,
      "epoch": 0.5072129255626082,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982587535082515,
      "loss": 6.3114,
      "mean_token_accuracy": 0.16224624663591386,
      "num_tokens": 11133903.0,
      "step": 4395
    },
    {
      "entropy": 6.466379690170288,
      "epoch": 0.5077899596076169,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004982536260143565,
      "loss": 6.3768,
      "mean_token_accuracy": 0.16100105494260789,
      "num_tokens": 11145607.0,
      "step": 4400
    },
    {
      "entropy": 6.51287145614624,
      "epoch": 0.5083669936526255,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004982484910114367,
      "loss": 6.4834,
      "mean_token_accuracy": 0.15186810493469238,
      "num_tokens": 11157332.0,
      "step": 4405
    },
    {
      "entropy": 6.590687894821167,
      "epoch": 0.5089440276976341,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982433484996648,
      "loss": 6.5847,
      "mean_token_accuracy": 0.15006298422813416,
      "num_tokens": 11169404.0,
      "step": 4410
    },
    {
      "entropy": 6.554287099838257,
      "epoch": 0.5095210617426428,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004982381984792138,
      "loss": 6.4567,
      "mean_token_accuracy": 0.15406338274478912,
      "num_tokens": 11182506.0,
      "step": 4415
    },
    {
      "entropy": 6.521548748016357,
      "epoch": 0.5100980957876514,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982330409502568,
      "loss": 6.4238,
      "mean_token_accuracy": 0.15082940310239792,
      "num_tokens": 11195507.0,
      "step": 4420
    },
    {
      "entropy": 6.535209894180298,
      "epoch": 0.5106751298326602,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004982278759129674,
      "loss": 6.4329,
      "mean_token_accuracy": 0.15719768702983855,
      "num_tokens": 11207219.0,
      "step": 4425
    },
    {
      "entropy": 6.438648223876953,
      "epoch": 0.5112521638776688,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004982227033675194,
      "loss": 6.4106,
      "mean_token_accuracy": 0.15815673768520355,
      "num_tokens": 11219301.0,
      "step": 4430
    },
    {
      "entropy": 6.460758399963379,
      "epoch": 0.5118291979226774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004982175233140865,
      "loss": 6.422,
      "mean_token_accuracy": 0.15656259655952454,
      "num_tokens": 11230957.0,
      "step": 4435
    },
    {
      "entropy": 6.560869407653809,
      "epoch": 0.5124062319676861,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982123357528431,
      "loss": 6.4481,
      "mean_token_accuracy": 0.1526852384209633,
      "num_tokens": 11242813.0,
      "step": 4440
    },
    {
      "entropy": 6.507971668243409,
      "epoch": 0.5129832660126947,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004982071406839636,
      "loss": 6.463,
      "mean_token_accuracy": 0.15058354064822196,
      "num_tokens": 11255193.0,
      "step": 4445
    },
    {
      "entropy": 6.388872909545898,
      "epoch": 0.5135603000577034,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982019381076229,
      "loss": 6.4273,
      "mean_token_accuracy": 0.15397539511322975,
      "num_tokens": 11267988.0,
      "step": 4450
    },
    {
      "entropy": 6.536984300613403,
      "epoch": 0.514137334102712,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981967280239958,
      "loss": 6.5089,
      "mean_token_accuracy": 0.14631436094641687,
      "num_tokens": 11280456.0,
      "step": 4455
    },
    {
      "entropy": 6.541901874542236,
      "epoch": 0.5147143681477208,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004981915104332578,
      "loss": 6.5337,
      "mean_token_accuracy": 0.14940202683210374,
      "num_tokens": 11293304.0,
      "step": 4460
    },
    {
      "entropy": 6.4461493492126465,
      "epoch": 0.5152914021927294,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498186285335584,
      "loss": 6.4935,
      "mean_token_accuracy": 0.14736132621765136,
      "num_tokens": 11305936.0,
      "step": 4465
    },
    {
      "entropy": 6.674417877197266,
      "epoch": 0.515868436237738,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004981810527311505,
      "loss": 6.546,
      "mean_token_accuracy": 0.14426559805870057,
      "num_tokens": 11319991.0,
      "step": 4470
    },
    {
      "entropy": 6.577980184555054,
      "epoch": 0.5164454702827467,
      "grad_norm": 1.25,
      "learning_rate": 0.000498175812620133,
      "loss": 6.4438,
      "mean_token_accuracy": 0.15654727965593337,
      "num_tokens": 11332572.0,
      "step": 4475
    },
    {
      "entropy": 6.419868278503418,
      "epoch": 0.5170225043277553,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981705650027081,
      "loss": 6.4572,
      "mean_token_accuracy": 0.15874334275722504,
      "num_tokens": 11344435.0,
      "step": 4480
    },
    {
      "entropy": 6.5055430889129635,
      "epoch": 0.517599538372764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000498165309879052,
      "loss": 6.3757,
      "mean_token_accuracy": 0.15824664235115052,
      "num_tokens": 11356078.0,
      "step": 4485
    },
    {
      "entropy": 6.469321823120117,
      "epoch": 0.5181765724177726,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981600472493415,
      "loss": 6.4263,
      "mean_token_accuracy": 0.1546856477856636,
      "num_tokens": 11368933.0,
      "step": 4490
    },
    {
      "entropy": 6.574912452697754,
      "epoch": 0.5187536064627813,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981547771137537,
      "loss": 6.5351,
      "mean_token_accuracy": 0.1477719157934189,
      "num_tokens": 11380810.0,
      "step": 4495
    },
    {
      "entropy": 6.48729829788208,
      "epoch": 0.51933064050779,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004981494994724658,
      "loss": 6.4445,
      "mean_token_accuracy": 0.1526978924870491,
      "num_tokens": 11392447.0,
      "step": 4500
    },
    {
      "entropy": 6.454307603836059,
      "epoch": 0.5199076745527986,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004981442143256554,
      "loss": 6.4863,
      "mean_token_accuracy": 0.15106553584337234,
      "num_tokens": 11405237.0,
      "step": 4505
    },
    {
      "entropy": 6.603475046157837,
      "epoch": 0.5204847085978073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981389216735001,
      "loss": 6.6036,
      "mean_token_accuracy": 0.14028041437268257,
      "num_tokens": 11418951.0,
      "step": 4510
    },
    {
      "entropy": 6.578861427307129,
      "epoch": 0.5210617426428159,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498133621516178,
      "loss": 6.492,
      "mean_token_accuracy": 0.15183505415916443,
      "num_tokens": 11432587.0,
      "step": 4515
    },
    {
      "entropy": 6.549359607696533,
      "epoch": 0.5216387766878245,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004981283138538675,
      "loss": 6.5326,
      "mean_token_accuracy": 0.14482680857181549,
      "num_tokens": 11445572.0,
      "step": 4520
    },
    {
      "entropy": 6.562268686294556,
      "epoch": 0.5222158107328332,
      "grad_norm": 1.078125,
      "learning_rate": 0.000498122998686747,
      "loss": 6.4821,
      "mean_token_accuracy": 0.14933309257030486,
      "num_tokens": 11457469.0,
      "step": 4525
    },
    {
      "entropy": 6.4967255115509035,
      "epoch": 0.5227928447778419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981176760149951,
      "loss": 6.5183,
      "mean_token_accuracy": 0.15408243611454964,
      "num_tokens": 11470306.0,
      "step": 4530
    },
    {
      "entropy": 6.562189292907715,
      "epoch": 0.5233698788228506,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004981123458387911,
      "loss": 6.5062,
      "mean_token_accuracy": 0.1558932587504387,
      "num_tokens": 11483839.0,
      "step": 4535
    },
    {
      "entropy": 6.578125905990601,
      "epoch": 0.5239469128678592,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004981070081583142,
      "loss": 6.5025,
      "mean_token_accuracy": 0.14841696321964265,
      "num_tokens": 11495703.0,
      "step": 4540
    },
    {
      "entropy": 6.603797674179077,
      "epoch": 0.5245239469128679,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498101662973744,
      "loss": 6.5831,
      "mean_token_accuracy": 0.14712464138865472,
      "num_tokens": 11508061.0,
      "step": 4545
    },
    {
      "entropy": 6.508516645431518,
      "epoch": 0.5251009809578765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049809631028526,
      "loss": 6.4236,
      "mean_token_accuracy": 0.15662758201360702,
      "num_tokens": 11520135.0,
      "step": 4550
    },
    {
      "entropy": 6.499169921875,
      "epoch": 0.5256780150028851,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980909500930424,
      "loss": 6.4971,
      "mean_token_accuracy": 0.14866289794445037,
      "num_tokens": 11532255.0,
      "step": 4555
    },
    {
      "entropy": 6.516340065002441,
      "epoch": 0.5262550490478938,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004980855823972717,
      "loss": 6.4104,
      "mean_token_accuracy": 0.15316666960716246,
      "num_tokens": 11544596.0,
      "step": 4560
    },
    {
      "entropy": 6.5355082035064695,
      "epoch": 0.5268320830929025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000498080207198128,
      "loss": 6.5571,
      "mean_token_accuracy": 0.1471469670534134,
      "num_tokens": 11556758.0,
      "step": 4565
    },
    {
      "entropy": 6.5845283508300785,
      "epoch": 0.5274091171379112,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004980748244957925,
      "loss": 6.5381,
      "mean_token_accuracy": 0.14486694410443307,
      "num_tokens": 11570070.0,
      "step": 4570
    },
    {
      "entropy": 6.519506359100342,
      "epoch": 0.5279861511829198,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004980694342904461,
      "loss": 6.427,
      "mean_token_accuracy": 0.15766648799180985,
      "num_tokens": 11582566.0,
      "step": 4575
    },
    {
      "entropy": 6.420183849334717,
      "epoch": 0.5285631852279284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980640365822701,
      "loss": 6.3809,
      "mean_token_accuracy": 0.1610960304737091,
      "num_tokens": 11595238.0,
      "step": 4580
    },
    {
      "entropy": 6.540434312820435,
      "epoch": 0.5291402192729371,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498058631371446,
      "loss": 6.5134,
      "mean_token_accuracy": 0.15686920285224915,
      "num_tokens": 11608646.0,
      "step": 4585
    },
    {
      "entropy": 6.461825847625732,
      "epoch": 0.5297172533179457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004980532186581555,
      "loss": 6.4396,
      "mean_token_accuracy": 0.15351363271474838,
      "num_tokens": 11619610.0,
      "step": 4590
    },
    {
      "entropy": 6.42177448272705,
      "epoch": 0.5302942873629544,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498047798442581,
      "loss": 6.3527,
      "mean_token_accuracy": 0.16407090574502944,
      "num_tokens": 11631691.0,
      "step": 4595
    },
    {
      "entropy": 6.516407203674317,
      "epoch": 0.5308713214079631,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980423707249044,
      "loss": 6.3881,
      "mean_token_accuracy": 0.1592670902609825,
      "num_tokens": 11644330.0,
      "step": 4600
    },
    {
      "entropy": 6.531644344329834,
      "epoch": 0.5314483554529718,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980369355053086,
      "loss": 6.5237,
      "mean_token_accuracy": 0.14757917076349258,
      "num_tokens": 11656826.0,
      "step": 4605
    },
    {
      "entropy": 6.451550531387329,
      "epoch": 0.5320253894979804,
      "grad_norm": 1.25,
      "learning_rate": 0.0004980314927839763,
      "loss": 6.4796,
      "mean_token_accuracy": 0.1471990704536438,
      "num_tokens": 11668669.0,
      "step": 4610
    },
    {
      "entropy": 6.606909942626953,
      "epoch": 0.532602423542989,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004980260425610903,
      "loss": 6.4498,
      "mean_token_accuracy": 0.14908241629600524,
      "num_tokens": 11681237.0,
      "step": 4615
    },
    {
      "entropy": 6.498003244400024,
      "epoch": 0.5331794575879977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980205848368343,
      "loss": 6.4956,
      "mean_token_accuracy": 0.15172789618372917,
      "num_tokens": 11693953.0,
      "step": 4620
    },
    {
      "entropy": 6.501087999343872,
      "epoch": 0.5337564916330063,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004980151196113917,
      "loss": 6.4413,
      "mean_token_accuracy": 0.1522899702191353,
      "num_tokens": 11705364.0,
      "step": 4625
    },
    {
      "entropy": 6.514171934127807,
      "epoch": 0.534333525678015,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004980096468849464,
      "loss": 6.455,
      "mean_token_accuracy": 0.15325311124324797,
      "num_tokens": 11717664.0,
      "step": 4630
    },
    {
      "entropy": 6.466876649856568,
      "epoch": 0.5349105597230237,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004980041666576823,
      "loss": 6.456,
      "mean_token_accuracy": 0.14555764496326445,
      "num_tokens": 11731143.0,
      "step": 4635
    },
    {
      "entropy": 6.550040197372437,
      "epoch": 0.5354875937680323,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004979986789297837,
      "loss": 6.4115,
      "mean_token_accuracy": 0.15628498494625093,
      "num_tokens": 11743904.0,
      "step": 4640
    },
    {
      "entropy": 6.330129718780517,
      "epoch": 0.536064627813041,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004979931837014355,
      "loss": 6.2898,
      "mean_token_accuracy": 0.15889219790697098,
      "num_tokens": 11756731.0,
      "step": 4645
    },
    {
      "entropy": 6.439878225326538,
      "epoch": 0.5366416618580496,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004979876809728223,
      "loss": 6.4106,
      "mean_token_accuracy": 0.1573275938630104,
      "num_tokens": 11768597.0,
      "step": 4650
    },
    {
      "entropy": 6.4509971141815186,
      "epoch": 0.5372186959030583,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004979821707441292,
      "loss": 6.3422,
      "mean_token_accuracy": 0.16196626275777817,
      "num_tokens": 11781464.0,
      "step": 4655
    },
    {
      "entropy": 6.384463500976563,
      "epoch": 0.5377957299480669,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004979766530155416,
      "loss": 6.3178,
      "mean_token_accuracy": 0.16592992544174195,
      "num_tokens": 11794215.0,
      "step": 4660
    },
    {
      "entropy": 6.5064746856689455,
      "epoch": 0.5383727639930755,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004979711277872449,
      "loss": 6.5442,
      "mean_token_accuracy": 0.14707691445946694,
      "num_tokens": 11807456.0,
      "step": 4665
    },
    {
      "entropy": 6.544828224182129,
      "epoch": 0.5389497980380843,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004979655950594252,
      "loss": 6.4476,
      "mean_token_accuracy": 0.15236640498042106,
      "num_tokens": 11819730.0,
      "step": 4670
    },
    {
      "entropy": 6.453517055511474,
      "epoch": 0.5395268320830929,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979600548322684,
      "loss": 6.432,
      "mean_token_accuracy": 0.16160822808742523,
      "num_tokens": 11831892.0,
      "step": 4675
    },
    {
      "entropy": 6.519505739212036,
      "epoch": 0.5401038661281016,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000497954507105961,
      "loss": 6.4578,
      "mean_token_accuracy": 0.15195320397615433,
      "num_tokens": 11844114.0,
      "step": 4680
    },
    {
      "entropy": 6.573498392105103,
      "epoch": 0.5406809001731102,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004979489518806893,
      "loss": 6.4187,
      "mean_token_accuracy": 0.1524272158741951,
      "num_tokens": 11855832.0,
      "step": 4685
    },
    {
      "entropy": 6.521721220016479,
      "epoch": 0.5412579342181189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979433891566405,
      "loss": 6.5266,
      "mean_token_accuracy": 0.15105092376470566,
      "num_tokens": 11868837.0,
      "step": 4690
    },
    {
      "entropy": 6.530892992019654,
      "epoch": 0.5418349682631275,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004979378189340015,
      "loss": 6.4183,
      "mean_token_accuracy": 0.16228097975254058,
      "num_tokens": 11880608.0,
      "step": 4695
    },
    {
      "entropy": 6.527881860733032,
      "epoch": 0.5424120023081361,
      "grad_norm": 1.125,
      "learning_rate": 0.0004979322412129597,
      "loss": 6.4508,
      "mean_token_accuracy": 0.15379999428987504,
      "num_tokens": 11892652.0,
      "step": 4700
    },
    {
      "entropy": 6.513641881942749,
      "epoch": 0.5429890363531449,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979266559937028,
      "loss": 6.3945,
      "mean_token_accuracy": 0.15369104892015456,
      "num_tokens": 11904572.0,
      "step": 4705
    },
    {
      "entropy": 6.591770267486572,
      "epoch": 0.5435660703981535,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979210632764185,
      "loss": 6.5345,
      "mean_token_accuracy": 0.14894211888313294,
      "num_tokens": 11917794.0,
      "step": 4710
    },
    {
      "entropy": 6.52169017791748,
      "epoch": 0.5441431044431622,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004979154630612949,
      "loss": 6.4262,
      "mean_token_accuracy": 0.15118273198604584,
      "num_tokens": 11930403.0,
      "step": 4715
    },
    {
      "entropy": 6.5592405796051025,
      "epoch": 0.5447201384881708,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004979098553485205,
      "loss": 6.5375,
      "mean_token_accuracy": 0.1481676384806633,
      "num_tokens": 11943974.0,
      "step": 4720
    },
    {
      "entropy": 6.52493953704834,
      "epoch": 0.5452971725331794,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979042401382838,
      "loss": 6.5349,
      "mean_token_accuracy": 0.14051474183797835,
      "num_tokens": 11956887.0,
      "step": 4725
    },
    {
      "entropy": 6.567168235778809,
      "epoch": 0.5458742065781881,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978986174307737,
      "loss": 6.3789,
      "mean_token_accuracy": 0.16181773245334624,
      "num_tokens": 11969504.0,
      "step": 4730
    },
    {
      "entropy": 6.418928146362305,
      "epoch": 0.5464512406231967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978929872261794,
      "loss": 6.4336,
      "mean_token_accuracy": 0.15474231988191606,
      "num_tokens": 11983432.0,
      "step": 4735
    },
    {
      "entropy": 6.59085521697998,
      "epoch": 0.5470282746682055,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004978873495246901,
      "loss": 6.4298,
      "mean_token_accuracy": 0.16287856549024582,
      "num_tokens": 11995915.0,
      "step": 4740
    },
    {
      "entropy": 6.460890197753907,
      "epoch": 0.5476053087132141,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004978817043264955,
      "loss": 6.4353,
      "mean_token_accuracy": 0.15744587630033494,
      "num_tokens": 12008395.0,
      "step": 4745
    },
    {
      "entropy": 6.377586364746094,
      "epoch": 0.5481823427582228,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004978760516317856,
      "loss": 6.3956,
      "mean_token_accuracy": 0.1520675890147686,
      "num_tokens": 12021179.0,
      "step": 4750
    },
    {
      "entropy": 6.540531587600708,
      "epoch": 0.5487593768032314,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004978703914407503,
      "loss": 6.489,
      "mean_token_accuracy": 0.14723778814077376,
      "num_tokens": 12033798.0,
      "step": 4755
    },
    {
      "entropy": 6.561176300048828,
      "epoch": 0.54933641084824,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004978647237535802,
      "loss": 6.4887,
      "mean_token_accuracy": 0.1505315586924553,
      "num_tokens": 12045720.0,
      "step": 4760
    },
    {
      "entropy": 6.543166399002075,
      "epoch": 0.5499134448932487,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004978590485704657,
      "loss": 6.4096,
      "mean_token_accuracy": 0.15576629489660263,
      "num_tokens": 12058616.0,
      "step": 4765
    },
    {
      "entropy": 6.486966133117676,
      "epoch": 0.5504904789382573,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004978533658915979,
      "loss": 6.5079,
      "mean_token_accuracy": 0.15112854689359664,
      "num_tokens": 12070173.0,
      "step": 4770
    },
    {
      "entropy": 6.481332635879516,
      "epoch": 0.5510675129832661,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004978476757171678,
      "loss": 6.4059,
      "mean_token_accuracy": 0.16395777463912964,
      "num_tokens": 12083499.0,
      "step": 4775
    },
    {
      "entropy": 6.472747802734375,
      "epoch": 0.5516445470282747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004978419780473668,
      "loss": 6.4507,
      "mean_token_accuracy": 0.1544622302055359,
      "num_tokens": 12095773.0,
      "step": 4780
    },
    {
      "entropy": 6.4680047035217285,
      "epoch": 0.5522215810732833,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004978362728823865,
      "loss": 6.4666,
      "mean_token_accuracy": 0.15453778207302094,
      "num_tokens": 12107640.0,
      "step": 4785
    },
    {
      "entropy": 6.480125045776367,
      "epoch": 0.552798615118292,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004978305602224189,
      "loss": 6.4045,
      "mean_token_accuracy": 0.1534944549202919,
      "num_tokens": 12120433.0,
      "step": 4790
    },
    {
      "entropy": 6.440623235702515,
      "epoch": 0.5533756491633006,
      "grad_norm": 1.25,
      "learning_rate": 0.0004978248400676562,
      "loss": 6.3806,
      "mean_token_accuracy": 0.15867497473955156,
      "num_tokens": 12133326.0,
      "step": 4795
    },
    {
      "entropy": 6.586127471923828,
      "epoch": 0.5539526832083093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004978191124182905,
      "loss": 6.4862,
      "mean_token_accuracy": 0.149084635078907,
      "num_tokens": 12145119.0,
      "step": 4800
    },
    {
      "entropy": 6.5216453075408936,
      "epoch": 0.5545297172533179,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004978133772745149,
      "loss": 6.4498,
      "mean_token_accuracy": 0.14746844321489333,
      "num_tokens": 12157414.0,
      "step": 4805
    },
    {
      "entropy": 6.436687326431274,
      "epoch": 0.5551067512983267,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978076346365218,
      "loss": 6.316,
      "mean_token_accuracy": 0.15791162550449372,
      "num_tokens": 12170226.0,
      "step": 4810
    },
    {
      "entropy": 6.456769037246704,
      "epoch": 0.5556837853433353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004978018845045047,
      "loss": 6.38,
      "mean_token_accuracy": 0.1574267938733101,
      "num_tokens": 12182804.0,
      "step": 4815
    },
    {
      "entropy": 6.512805700302124,
      "epoch": 0.5562608193883439,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004977961268786568,
      "loss": 6.4683,
      "mean_token_accuracy": 0.15100056529045106,
      "num_tokens": 12194795.0,
      "step": 4820
    },
    {
      "entropy": 6.4469624042510985,
      "epoch": 0.5568378534333526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977903617591719,
      "loss": 6.4403,
      "mean_token_accuracy": 0.1635904222726822,
      "num_tokens": 12207880.0,
      "step": 4825
    },
    {
      "entropy": 6.49699854850769,
      "epoch": 0.5574148874783612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004977845891462439,
      "loss": 6.5177,
      "mean_token_accuracy": 0.1520783096551895,
      "num_tokens": 12219434.0,
      "step": 4830
    },
    {
      "entropy": 6.497847318649292,
      "epoch": 0.5579919215233698,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977788090400668,
      "loss": 6.4267,
      "mean_token_accuracy": 0.16363779753446578,
      "num_tokens": 12232775.0,
      "step": 4835
    },
    {
      "entropy": 6.469997453689575,
      "epoch": 0.5585689555683785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004977730214408352,
      "loss": 6.4055,
      "mean_token_accuracy": 0.15352552086114885,
      "num_tokens": 12245306.0,
      "step": 4840
    },
    {
      "entropy": 6.411099481582641,
      "epoch": 0.5591459896133872,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004977672263487435,
      "loss": 6.3635,
      "mean_token_accuracy": 0.1560710147023201,
      "num_tokens": 12257553.0,
      "step": 4845
    },
    {
      "entropy": 6.45292592048645,
      "epoch": 0.5597230236583959,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000497761423763987,
      "loss": 6.4251,
      "mean_token_accuracy": 0.15596397668123246,
      "num_tokens": 12269626.0,
      "step": 4850
    },
    {
      "entropy": 6.528067350387573,
      "epoch": 0.5603000577034045,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004977556136867606,
      "loss": 6.4507,
      "mean_token_accuracy": 0.15139178037643433,
      "num_tokens": 12282180.0,
      "step": 4855
    },
    {
      "entropy": 6.565957403182983,
      "epoch": 0.5608770917484132,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004977497961172598,
      "loss": 6.4972,
      "mean_token_accuracy": 0.15649604499340058,
      "num_tokens": 12296315.0,
      "step": 4860
    },
    {
      "entropy": 6.459849786758423,
      "epoch": 0.5614541257934218,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977439710556802,
      "loss": 6.4183,
      "mean_token_accuracy": 0.14864314943552018,
      "num_tokens": 12308800.0,
      "step": 4865
    },
    {
      "entropy": 6.478931283950805,
      "epoch": 0.5620311598384304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977381385022179,
      "loss": 6.4801,
      "mean_token_accuracy": 0.1533154159784317,
      "num_tokens": 12321717.0,
      "step": 4870
    },
    {
      "entropy": 6.566097736358643,
      "epoch": 0.5626081938834391,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004977322984570688,
      "loss": 6.5094,
      "mean_token_accuracy": 0.1451910525560379,
      "num_tokens": 12335446.0,
      "step": 4875
    },
    {
      "entropy": 6.505554676055908,
      "epoch": 0.5631852279284478,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004977264509204296,
      "loss": 6.4127,
      "mean_token_accuracy": 0.15165958628058435,
      "num_tokens": 12348677.0,
      "step": 4880
    },
    {
      "entropy": 6.485540676116943,
      "epoch": 0.5637622619734565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004977205958924967,
      "loss": 6.4415,
      "mean_token_accuracy": 0.1551760196685791,
      "num_tokens": 12360934.0,
      "step": 4885
    },
    {
      "entropy": 6.537460088729858,
      "epoch": 0.5643392960184651,
      "grad_norm": 1.25,
      "learning_rate": 0.0004977147333734673,
      "loss": 6.5441,
      "mean_token_accuracy": 0.14865117520093918,
      "num_tokens": 12373538.0,
      "step": 4890
    },
    {
      "entropy": 6.429516363143921,
      "epoch": 0.5649163300634737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977088633635385,
      "loss": 6.4281,
      "mean_token_accuracy": 0.1513530433177948,
      "num_tokens": 12385291.0,
      "step": 4895
    },
    {
      "entropy": 6.613570261001587,
      "epoch": 0.5654933641084824,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004977029858629076,
      "loss": 6.485,
      "mean_token_accuracy": 0.14706155732274057,
      "num_tokens": 12398324.0,
      "step": 4900
    },
    {
      "entropy": 6.503150606155396,
      "epoch": 0.566070398153491,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004976971008717726,
      "loss": 6.4339,
      "mean_token_accuracy": 0.1525312729179859,
      "num_tokens": 12409807.0,
      "step": 4905
    },
    {
      "entropy": 6.515785837173462,
      "epoch": 0.5666474321984997,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976912083903309,
      "loss": 6.5215,
      "mean_token_accuracy": 0.14461371153593064,
      "num_tokens": 12422658.0,
      "step": 4910
    },
    {
      "entropy": 6.550714159011841,
      "epoch": 0.5672244662435084,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976853084187814,
      "loss": 6.4638,
      "mean_token_accuracy": 0.14637623578310013,
      "num_tokens": 12435912.0,
      "step": 4915
    },
    {
      "entropy": 6.453034782409668,
      "epoch": 0.5678015002885171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976794009573219,
      "loss": 6.4488,
      "mean_token_accuracy": 0.15766249001026153,
      "num_tokens": 12448382.0,
      "step": 4920
    },
    {
      "entropy": 6.508889102935791,
      "epoch": 0.5683785343335257,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004976734860061515,
      "loss": 6.5688,
      "mean_token_accuracy": 0.15048769414424895,
      "num_tokens": 12461355.0,
      "step": 4925
    },
    {
      "entropy": 6.566471338272095,
      "epoch": 0.5689555683785343,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004976675635654688,
      "loss": 6.4012,
      "mean_token_accuracy": 0.15381542593240738,
      "num_tokens": 12473931.0,
      "step": 4930
    },
    {
      "entropy": 6.4354527473449705,
      "epoch": 0.569532602423543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004976616336354733,
      "loss": 6.4381,
      "mean_token_accuracy": 0.15482519343495368,
      "num_tokens": 12486130.0,
      "step": 4935
    },
    {
      "entropy": 6.471525001525879,
      "epoch": 0.5701096364685516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976556962163645,
      "loss": 6.3731,
      "mean_token_accuracy": 0.156865394115448,
      "num_tokens": 12499336.0,
      "step": 4940
    },
    {
      "entropy": 6.539810228347778,
      "epoch": 0.5706866705135603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004976497513083419,
      "loss": 6.3757,
      "mean_token_accuracy": 0.15898544192314149,
      "num_tokens": 12512875.0,
      "step": 4945
    },
    {
      "entropy": 6.477344799041748,
      "epoch": 0.571263704558569,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004976437989116055,
      "loss": 6.487,
      "mean_token_accuracy": 0.15137718468904496,
      "num_tokens": 12524985.0,
      "step": 4950
    },
    {
      "entropy": 6.505421304702759,
      "epoch": 0.5718407386035776,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004976378390263554,
      "loss": 6.4346,
      "mean_token_accuracy": 0.15926533192396164,
      "num_tokens": 12538244.0,
      "step": 4955
    },
    {
      "entropy": 6.559779691696167,
      "epoch": 0.5724177726485863,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004976318716527924,
      "loss": 6.4404,
      "mean_token_accuracy": 0.15664353370666503,
      "num_tokens": 12550769.0,
      "step": 4960
    },
    {
      "entropy": 6.451388883590698,
      "epoch": 0.5729948066935949,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976258967911168,
      "loss": 6.465,
      "mean_token_accuracy": 0.1511801674962044,
      "num_tokens": 12563294.0,
      "step": 4965
    },
    {
      "entropy": 6.450545358657837,
      "epoch": 0.5735718407386036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976199144415298,
      "loss": 6.4484,
      "mean_token_accuracy": 0.14877015054225923,
      "num_tokens": 12575857.0,
      "step": 4970
    },
    {
      "entropy": 6.539900922775269,
      "epoch": 0.5741488747836122,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004976139246042325,
      "loss": 6.4198,
      "mean_token_accuracy": 0.15611875653266907,
      "num_tokens": 12589881.0,
      "step": 4975
    },
    {
      "entropy": 6.510217142105103,
      "epoch": 0.5747259088286208,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004976079272794265,
      "loss": 6.4881,
      "mean_token_accuracy": 0.14711457639932632,
      "num_tokens": 12604100.0,
      "step": 4980
    },
    {
      "entropy": 6.591409635543823,
      "epoch": 0.5753029428736296,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004976019224673134,
      "loss": 6.5016,
      "mean_token_accuracy": 0.1494516521692276,
      "num_tokens": 12616643.0,
      "step": 4985
    },
    {
      "entropy": 6.485559701919556,
      "epoch": 0.5758799769186382,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004975959101680953,
      "loss": 6.4145,
      "mean_token_accuracy": 0.15998050570487976,
      "num_tokens": 12627676.0,
      "step": 4990
    },
    {
      "entropy": 6.3982460498809814,
      "epoch": 0.5764570109636469,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004975898903819742,
      "loss": 6.3992,
      "mean_token_accuracy": 0.16120668202638627,
      "num_tokens": 12639352.0,
      "step": 4995
    },
    {
      "entropy": 6.444949531555176,
      "epoch": 0.5770340450086555,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975838631091527,
      "loss": 6.3632,
      "mean_token_accuracy": 0.15658557564020156,
      "num_tokens": 12652102.0,
      "step": 5000
    },
    {
      "entropy": 6.556867361068726,
      "epoch": 0.5776110790536642,
      "grad_norm": 1.375,
      "learning_rate": 0.0004975778283498336,
      "loss": 6.4777,
      "mean_token_accuracy": 0.1522747591137886,
      "num_tokens": 12665084.0,
      "step": 5005
    },
    {
      "entropy": 6.399194192886353,
      "epoch": 0.5781881130986728,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004975717861042198,
      "loss": 6.3905,
      "mean_token_accuracy": 0.1488622397184372,
      "num_tokens": 12677512.0,
      "step": 5010
    },
    {
      "entropy": 6.519528579711914,
      "epoch": 0.5787651471436814,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004975657363725146,
      "loss": 6.4929,
      "mean_token_accuracy": 0.14468461647629738,
      "num_tokens": 12690488.0,
      "step": 5015
    },
    {
      "entropy": 6.569351863861084,
      "epoch": 0.5793421811886902,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004975596791549213,
      "loss": 6.4386,
      "mean_token_accuracy": 0.15072959661483765,
      "num_tokens": 12703830.0,
      "step": 5020
    },
    {
      "entropy": 6.528122615814209,
      "epoch": 0.5799192152336988,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004975536144516437,
      "loss": 6.4533,
      "mean_token_accuracy": 0.15344501584768294,
      "num_tokens": 12716020.0,
      "step": 5025
    },
    {
      "entropy": 6.51330943107605,
      "epoch": 0.5804962492787075,
      "grad_norm": 1.265625,
      "learning_rate": 0.000497547542262886,
      "loss": 6.518,
      "mean_token_accuracy": 0.14512748420238494,
      "num_tokens": 12728989.0,
      "step": 5030
    },
    {
      "entropy": 6.550060606002807,
      "epoch": 0.5810732833237161,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004975414625888521,
      "loss": 6.3809,
      "mean_token_accuracy": 0.15464969575405121,
      "num_tokens": 12741090.0,
      "step": 5035
    },
    {
      "entropy": 6.5245343208312985,
      "epoch": 0.5816503173687247,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004975353754297468,
      "loss": 6.422,
      "mean_token_accuracy": 0.15102563947439193,
      "num_tokens": 12754176.0,
      "step": 5040
    },
    {
      "entropy": 6.372476387023926,
      "epoch": 0.5822273514137334,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004975292807857745,
      "loss": 6.406,
      "mean_token_accuracy": 0.1529885560274124,
      "num_tokens": 12766149.0,
      "step": 5045
    },
    {
      "entropy": 6.4757085800170895,
      "epoch": 0.582804385458742,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004975231786571406,
      "loss": 6.4473,
      "mean_token_accuracy": 0.15353928953409196,
      "num_tokens": 12778337.0,
      "step": 5050
    },
    {
      "entropy": 6.508154201507568,
      "epoch": 0.5833814195037508,
      "grad_norm": 1.125,
      "learning_rate": 0.0004975170690440499,
      "loss": 6.3845,
      "mean_token_accuracy": 0.15635424107313156,
      "num_tokens": 12790146.0,
      "step": 5055
    },
    {
      "entropy": 6.451329183578491,
      "epoch": 0.5839584535487594,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004975109519467083,
      "loss": 6.3748,
      "mean_token_accuracy": 0.1544017016887665,
      "num_tokens": 12802821.0,
      "step": 5060
    },
    {
      "entropy": 6.4523937702178955,
      "epoch": 0.584535487593768,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975048273653212,
      "loss": 6.5114,
      "mean_token_accuracy": 0.15367977023124696,
      "num_tokens": 12814409.0,
      "step": 5065
    },
    {
      "entropy": 6.532897138595581,
      "epoch": 0.5851125216387767,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974986953000948,
      "loss": 6.4037,
      "mean_token_accuracy": 0.14837779253721237,
      "num_tokens": 12827398.0,
      "step": 5070
    },
    {
      "entropy": 6.420920324325562,
      "epoch": 0.5856895556837853,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004974925557512354,
      "loss": 6.3946,
      "mean_token_accuracy": 0.15595378279685973,
      "num_tokens": 12840378.0,
      "step": 5075
    },
    {
      "entropy": 6.459633493423462,
      "epoch": 0.586266589728794,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004974864087189493,
      "loss": 6.4854,
      "mean_token_accuracy": 0.15719999819993974,
      "num_tokens": 12853493.0,
      "step": 5080
    },
    {
      "entropy": 6.489672040939331,
      "epoch": 0.5868436237738026,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004974802542034434,
      "loss": 6.3489,
      "mean_token_accuracy": 0.16045349091291428,
      "num_tokens": 12865052.0,
      "step": 5085
    },
    {
      "entropy": 6.51693902015686,
      "epoch": 0.5874206578188114,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004974740922049246,
      "loss": 6.4822,
      "mean_token_accuracy": 0.14741032123565673,
      "num_tokens": 12878095.0,
      "step": 5090
    },
    {
      "entropy": 6.476896572113037,
      "epoch": 0.58799769186382,
      "grad_norm": 1.125,
      "learning_rate": 0.0004974679227236004,
      "loss": 6.3246,
      "mean_token_accuracy": 0.162299744784832,
      "num_tokens": 12890855.0,
      "step": 5095
    },
    {
      "entropy": 6.500436401367187,
      "epoch": 0.5885747259088286,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004974617457596779,
      "loss": 6.4031,
      "mean_token_accuracy": 0.1562817022204399,
      "num_tokens": 12903833.0,
      "step": 5100
    },
    {
      "entropy": 6.537255668640137,
      "epoch": 0.5891517599538373,
      "grad_norm": 1.25,
      "learning_rate": 0.0004974555613133652,
      "loss": 6.5106,
      "mean_token_accuracy": 0.1479603409767151,
      "num_tokens": 12917579.0,
      "step": 5105
    },
    {
      "entropy": 6.4827710628509525,
      "epoch": 0.5897287939988459,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004974493693848702,
      "loss": 6.479,
      "mean_token_accuracy": 0.15590202063322067,
      "num_tokens": 12929718.0,
      "step": 5110
    },
    {
      "entropy": 6.338919973373413,
      "epoch": 0.5903058280438546,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974431699744011,
      "loss": 6.3377,
      "mean_token_accuracy": 0.16054340749979018,
      "num_tokens": 12943031.0,
      "step": 5115
    },
    {
      "entropy": 6.465764427185059,
      "epoch": 0.5908828620888632,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004974369630821665,
      "loss": 6.4345,
      "mean_token_accuracy": 0.16000231057405473,
      "num_tokens": 12954843.0,
      "step": 5120
    },
    {
      "entropy": 6.479459381103515,
      "epoch": 0.5914598961338718,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974307487083752,
      "loss": 6.4406,
      "mean_token_accuracy": 0.15593716353178025,
      "num_tokens": 12966915.0,
      "step": 5125
    },
    {
      "entropy": 6.507039642333984,
      "epoch": 0.5920369301788806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004974245268532361,
      "loss": 6.4844,
      "mean_token_accuracy": 0.14569110572338104,
      "num_tokens": 12979941.0,
      "step": 5130
    },
    {
      "entropy": 6.544470405578613,
      "epoch": 0.5926139642238892,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004974182975169585,
      "loss": 6.3857,
      "mean_token_accuracy": 0.15324565023183823,
      "num_tokens": 12992836.0,
      "step": 5135
    },
    {
      "entropy": 6.545553588867188,
      "epoch": 0.5931909982688979,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497412060699752,
      "loss": 6.4179,
      "mean_token_accuracy": 0.1531425304710865,
      "num_tokens": 13005246.0,
      "step": 5140
    },
    {
      "entropy": 6.393571567535401,
      "epoch": 0.5937680323139065,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004974058164018263,
      "loss": 6.3902,
      "mean_token_accuracy": 0.15206687152385712,
      "num_tokens": 13016842.0,
      "step": 5145
    },
    {
      "entropy": 6.4142437934875485,
      "epoch": 0.5943450663589152,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004973995646233914,
      "loss": 6.3766,
      "mean_token_accuracy": 0.15784988552331924,
      "num_tokens": 13028388.0,
      "step": 5150
    },
    {
      "entropy": 6.508879947662353,
      "epoch": 0.5949221004039238,
      "grad_norm": 1.125,
      "learning_rate": 0.0004973933053646576,
      "loss": 6.4231,
      "mean_token_accuracy": 0.1518938884139061,
      "num_tokens": 13041895.0,
      "step": 5155
    },
    {
      "entropy": 6.523378658294678,
      "epoch": 0.5954991344489324,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004973870386258355,
      "loss": 6.4875,
      "mean_token_accuracy": 0.1547953560948372,
      "num_tokens": 13055984.0,
      "step": 5160
    },
    {
      "entropy": 6.517923307418823,
      "epoch": 0.5960761684939412,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004973807644071357,
      "loss": 6.4272,
      "mean_token_accuracy": 0.15377702489495276,
      "num_tokens": 13068735.0,
      "step": 5165
    },
    {
      "entropy": 6.57601752281189,
      "epoch": 0.5966532025389498,
      "grad_norm": 1.125,
      "learning_rate": 0.0004973744827087695,
      "loss": 6.5136,
      "mean_token_accuracy": 0.1491558223962784,
      "num_tokens": 13081564.0,
      "step": 5170
    },
    {
      "entropy": 6.510840225219726,
      "epoch": 0.5972302365839585,
      "grad_norm": 1.40625,
      "learning_rate": 0.000497368193530948,
      "loss": 6.43,
      "mean_token_accuracy": 0.1526069462299347,
      "num_tokens": 13094146.0,
      "step": 5175
    },
    {
      "entropy": 6.429110145568847,
      "epoch": 0.5978072706289671,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004973618968738828,
      "loss": 6.3644,
      "mean_token_accuracy": 0.15681997686624527,
      "num_tokens": 13106753.0,
      "step": 5180
    },
    {
      "entropy": 6.384876012802124,
      "epoch": 0.5983843046739757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973555927377856,
      "loss": 6.3462,
      "mean_token_accuracy": 0.16190443187952042,
      "num_tokens": 13118645.0,
      "step": 5185
    },
    {
      "entropy": 6.50286865234375,
      "epoch": 0.5989613387189844,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004973492811228685,
      "loss": 6.4626,
      "mean_token_accuracy": 0.14891703352332114,
      "num_tokens": 13132250.0,
      "step": 5190
    },
    {
      "entropy": 6.502938985824585,
      "epoch": 0.599538372763993,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004973429620293438,
      "loss": 6.4612,
      "mean_token_accuracy": 0.14949870407581328,
      "num_tokens": 13145069.0,
      "step": 5195
    },
    {
      "entropy": 6.515492725372314,
      "epoch": 0.6001154068090018,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973366354574239,
      "loss": 6.4103,
      "mean_token_accuracy": 0.15858462899923326,
      "num_tokens": 13158244.0,
      "step": 5200
    },
    {
      "entropy": 6.51719102859497,
      "epoch": 0.6006924408540104,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004973303014073219,
      "loss": 6.4766,
      "mean_token_accuracy": 0.14795382171869279,
      "num_tokens": 13171810.0,
      "step": 5205
    },
    {
      "entropy": 6.426719427108765,
      "epoch": 0.601269474899019,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004973239598792504,
      "loss": 6.3122,
      "mean_token_accuracy": 0.15710260570049286,
      "num_tokens": 13184448.0,
      "step": 5210
    },
    {
      "entropy": 6.503760433197021,
      "epoch": 0.6018465089440277,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004973176108734232,
      "loss": 6.4881,
      "mean_token_accuracy": 0.1439894437789917,
      "num_tokens": 13198350.0,
      "step": 5215
    },
    {
      "entropy": 6.521775341033935,
      "epoch": 0.6024235429890363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004973112543900535,
      "loss": 6.4138,
      "mean_token_accuracy": 0.15527199506759642,
      "num_tokens": 13210660.0,
      "step": 5220
    },
    {
      "entropy": 6.409607553482056,
      "epoch": 0.603000577034045,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004973048904293551,
      "loss": 6.3827,
      "mean_token_accuracy": 0.16117679625749587,
      "num_tokens": 13222991.0,
      "step": 5225
    },
    {
      "entropy": 6.459820032119751,
      "epoch": 0.6035776110790536,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004972985189915422,
      "loss": 6.4393,
      "mean_token_accuracy": 0.14954399913549424,
      "num_tokens": 13237693.0,
      "step": 5230
    },
    {
      "entropy": 6.474009656906128,
      "epoch": 0.6041546451240624,
      "grad_norm": 1.65625,
      "learning_rate": 0.000497292140076829,
      "loss": 6.4548,
      "mean_token_accuracy": 0.15322939604520797,
      "num_tokens": 13249843.0,
      "step": 5235
    },
    {
      "entropy": 6.510292482376099,
      "epoch": 0.604731679169071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972857536854301,
      "loss": 6.4215,
      "mean_token_accuracy": 0.15622295439243317,
      "num_tokens": 13262185.0,
      "step": 5240
    },
    {
      "entropy": 6.454311656951904,
      "epoch": 0.6053087132140796,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004972793598175603,
      "loss": 6.4642,
      "mean_token_accuracy": 0.1478947676718235,
      "num_tokens": 13275344.0,
      "step": 5245
    },
    {
      "entropy": 6.513757181167603,
      "epoch": 0.6058857472590883,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004972729584734347,
      "loss": 6.3843,
      "mean_token_accuracy": 0.1522686466574669,
      "num_tokens": 13287741.0,
      "step": 5250
    },
    {
      "entropy": 6.494974374771118,
      "epoch": 0.6064627813040969,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972665496532685,
      "loss": 6.3797,
      "mean_token_accuracy": 0.15595458894968034,
      "num_tokens": 13299918.0,
      "step": 5255
    },
    {
      "entropy": 6.378756904602051,
      "epoch": 0.6070398153491056,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004972601333572774,
      "loss": 6.3684,
      "mean_token_accuracy": 0.15408076345920563,
      "num_tokens": 13312765.0,
      "step": 5260
    },
    {
      "entropy": 6.469124364852905,
      "epoch": 0.6076168493941142,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004972537095856769,
      "loss": 6.3743,
      "mean_token_accuracy": 0.1507202297449112,
      "num_tokens": 13325324.0,
      "step": 5265
    },
    {
      "entropy": 6.516126489639282,
      "epoch": 0.608193883439123,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004972472783386834,
      "loss": 6.4594,
      "mean_token_accuracy": 0.14877294152975082,
      "num_tokens": 13337578.0,
      "step": 5270
    },
    {
      "entropy": 6.530485391616821,
      "epoch": 0.6087709174841316,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004972408396165132,
      "loss": 6.423,
      "mean_token_accuracy": 0.15622132271528244,
      "num_tokens": 13349218.0,
      "step": 5275
    },
    {
      "entropy": 6.458002948760987,
      "epoch": 0.6093479515291402,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004972343934193826,
      "loss": 6.3933,
      "mean_token_accuracy": 0.1541997194290161,
      "num_tokens": 13362221.0,
      "step": 5280
    },
    {
      "entropy": 6.478120613098144,
      "epoch": 0.6099249855741489,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004972279397475086,
      "loss": 6.384,
      "mean_token_accuracy": 0.15952515602111816,
      "num_tokens": 13374147.0,
      "step": 5285
    },
    {
      "entropy": 6.396870422363281,
      "epoch": 0.6105020196191575,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972214786011083,
      "loss": 6.3315,
      "mean_token_accuracy": 0.15817706137895585,
      "num_tokens": 13386077.0,
      "step": 5290
    },
    {
      "entropy": 6.551906967163086,
      "epoch": 0.6110790536641661,
      "grad_norm": 1.296875,
      "learning_rate": 0.000497215009980399,
      "loss": 6.4253,
      "mean_token_accuracy": 0.15904004573822023,
      "num_tokens": 13398442.0,
      "step": 5295
    },
    {
      "entropy": 6.448291397094726,
      "epoch": 0.6116560877091748,
      "grad_norm": 1.28125,
      "learning_rate": 0.000497208533885598,
      "loss": 6.4321,
      "mean_token_accuracy": 0.1560212790966034,
      "num_tokens": 13411249.0,
      "step": 5300
    },
    {
      "entropy": 6.414372968673706,
      "epoch": 0.6122331217541835,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004972020503169235,
      "loss": 6.4052,
      "mean_token_accuracy": 0.16007892787456512,
      "num_tokens": 13424518.0,
      "step": 5305
    },
    {
      "entropy": 6.490907001495361,
      "epoch": 0.6128101557991922,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004971955592745934,
      "loss": 6.418,
      "mean_token_accuracy": 0.14899912327528,
      "num_tokens": 13438530.0,
      "step": 5310
    },
    {
      "entropy": 6.42531452178955,
      "epoch": 0.6133871898442008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497189060758826,
      "loss": 6.3676,
      "mean_token_accuracy": 0.16263744086027146,
      "num_tokens": 13451154.0,
      "step": 5315
    },
    {
      "entropy": 6.432960033416748,
      "epoch": 0.6139642238892095,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00049718255476984,
      "loss": 6.3967,
      "mean_token_accuracy": 0.161380535364151,
      "num_tokens": 13461938.0,
      "step": 5320
    },
    {
      "entropy": 6.436057424545288,
      "epoch": 0.6145412579342181,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004971760413078539,
      "loss": 6.3713,
      "mean_token_accuracy": 0.15706276893615723,
      "num_tokens": 13474423.0,
      "step": 5325
    },
    {
      "entropy": 6.489777135848999,
      "epoch": 0.6151182919792267,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004971695203730872,
      "loss": 6.3786,
      "mean_token_accuracy": 0.16456449776887894,
      "num_tokens": 13486882.0,
      "step": 5330
    },
    {
      "entropy": 6.458373928070069,
      "epoch": 0.6156953260242354,
      "grad_norm": 1.203125,
      "learning_rate": 0.000497162991965759,
      "loss": 6.3343,
      "mean_token_accuracy": 0.15441225916147233,
      "num_tokens": 13499692.0,
      "step": 5335
    },
    {
      "entropy": 6.4966339588165285,
      "epoch": 0.6162723600692441,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971564560860889,
      "loss": 6.4064,
      "mean_token_accuracy": 0.15078266859054565,
      "num_tokens": 13512995.0,
      "step": 5340
    },
    {
      "entropy": 6.369403266906739,
      "epoch": 0.6168493941142528,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971499127342968,
      "loss": 6.3278,
      "mean_token_accuracy": 0.16098241358995438,
      "num_tokens": 13525804.0,
      "step": 5345
    },
    {
      "entropy": 6.437539720535279,
      "epoch": 0.6174264281592614,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004971433619106027,
      "loss": 6.4364,
      "mean_token_accuracy": 0.1512218251824379,
      "num_tokens": 13537807.0,
      "step": 5350
    },
    {
      "entropy": 6.584434080123901,
      "epoch": 0.61800346220427,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000497136803615227,
      "loss": 6.4724,
      "mean_token_accuracy": 0.14654513746500014,
      "num_tokens": 13550765.0,
      "step": 5355
    },
    {
      "entropy": 6.452790403366089,
      "epoch": 0.6185804962492787,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004971302378483902,
      "loss": 6.4246,
      "mean_token_accuracy": 0.1566222384572029,
      "num_tokens": 13565084.0,
      "step": 5360
    },
    {
      "entropy": 6.529478931427002,
      "epoch": 0.6191575302942873,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004971236646103132,
      "loss": 6.4197,
      "mean_token_accuracy": 0.15318245440721512,
      "num_tokens": 13578851.0,
      "step": 5365
    },
    {
      "entropy": 6.477718639373779,
      "epoch": 0.619734564339296,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004971170839012171,
      "loss": 6.3086,
      "mean_token_accuracy": 0.16245864033699037,
      "num_tokens": 13589984.0,
      "step": 5370
    },
    {
      "entropy": 6.382431983947754,
      "epoch": 0.6203115983843047,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004971104957213233,
      "loss": 6.3887,
      "mean_token_accuracy": 0.15082724690437316,
      "num_tokens": 13602899.0,
      "step": 5375
    },
    {
      "entropy": 6.47854175567627,
      "epoch": 0.6208886324293134,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004971039000708531,
      "loss": 6.5548,
      "mean_token_accuracy": 0.1424169920384884,
      "num_tokens": 13617672.0,
      "step": 5380
    },
    {
      "entropy": 6.5490500926971436,
      "epoch": 0.621465666474322,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004970972969500286,
      "loss": 6.4049,
      "mean_token_accuracy": 0.15098279267549514,
      "num_tokens": 13630029.0,
      "step": 5385
    },
    {
      "entropy": 6.42647614479065,
      "epoch": 0.6220427005193306,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000497090686359072,
      "loss": 6.3839,
      "mean_token_accuracy": 0.15539321005344392,
      "num_tokens": 13643091.0,
      "step": 5390
    },
    {
      "entropy": 6.385086917877198,
      "epoch": 0.6226197345643393,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004970840682982054,
      "loss": 6.2634,
      "mean_token_accuracy": 0.16022852510213853,
      "num_tokens": 13654790.0,
      "step": 5395
    },
    {
      "entropy": 6.455251741409302,
      "epoch": 0.6231967686093479,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004970774427676514,
      "loss": 6.3606,
      "mean_token_accuracy": 0.1599068224430084,
      "num_tokens": 13667425.0,
      "step": 5400
    },
    {
      "entropy": 6.516546249389648,
      "epoch": 0.6237738026543566,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000497070809767633,
      "loss": 6.4386,
      "mean_token_accuracy": 0.15648358017206193,
      "num_tokens": 13680199.0,
      "step": 5405
    },
    {
      "entropy": 6.4648191928863525,
      "epoch": 0.6243508366993653,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004970641692983731,
      "loss": 6.438,
      "mean_token_accuracy": 0.15207870900630951,
      "num_tokens": 13693613.0,
      "step": 5410
    },
    {
      "entropy": 6.374730634689331,
      "epoch": 0.624927870744374,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004970575213600954,
      "loss": 6.4049,
      "mean_token_accuracy": 0.15325464606285094,
      "num_tokens": 13706809.0,
      "step": 5415
    },
    {
      "entropy": 6.499360609054565,
      "epoch": 0.6255049047893826,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004970508659530231,
      "loss": 6.3474,
      "mean_token_accuracy": 0.15512218326330185,
      "num_tokens": 13718976.0,
      "step": 5420
    },
    {
      "entropy": 6.463284254074097,
      "epoch": 0.6260819388343912,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970442030773802,
      "loss": 6.354,
      "mean_token_accuracy": 0.15274603366851808,
      "num_tokens": 13731037.0,
      "step": 5425
    },
    {
      "entropy": 6.401796483993531,
      "epoch": 0.6266589728793999,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004970375327333909,
      "loss": 6.412,
      "mean_token_accuracy": 0.1536305785179138,
      "num_tokens": 13743816.0,
      "step": 5430
    },
    {
      "entropy": 6.407013082504273,
      "epoch": 0.6272360069244085,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970308549212796,
      "loss": 6.3983,
      "mean_token_accuracy": 0.15897883772850036,
      "num_tokens": 13757165.0,
      "step": 5435
    },
    {
      "entropy": 6.470421314239502,
      "epoch": 0.6278130409694171,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004970241696412705,
      "loss": 6.3747,
      "mean_token_accuracy": 0.15822548270225525,
      "num_tokens": 13769982.0,
      "step": 5440
    },
    {
      "entropy": 6.47586350440979,
      "epoch": 0.6283900750144259,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000497017476893589,
      "loss": 6.3833,
      "mean_token_accuracy": 0.16209243834018708,
      "num_tokens": 13782922.0,
      "step": 5445
    },
    {
      "entropy": 6.44397201538086,
      "epoch": 0.6289671090594345,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004970107766784598,
      "loss": 6.3988,
      "mean_token_accuracy": 0.15553117990493776,
      "num_tokens": 13795683.0,
      "step": 5450
    },
    {
      "entropy": 6.577775239944458,
      "epoch": 0.6295441431044432,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004970040689961084,
      "loss": 6.4834,
      "mean_token_accuracy": 0.15517319813370706,
      "num_tokens": 13808332.0,
      "step": 5455
    },
    {
      "entropy": 6.529779148101807,
      "epoch": 0.6301211771494518,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004969973538467605,
      "loss": 6.4313,
      "mean_token_accuracy": 0.1575072228908539,
      "num_tokens": 13820949.0,
      "step": 5460
    },
    {
      "entropy": 6.391829109191894,
      "epoch": 0.6306982111944605,
      "grad_norm": 1.5,
      "learning_rate": 0.0004969906312306419,
      "loss": 6.4199,
      "mean_token_accuracy": 0.15567118525505066,
      "num_tokens": 13833623.0,
      "step": 5465
    },
    {
      "entropy": 6.449386215209961,
      "epoch": 0.6312752452394691,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004969839011479786,
      "loss": 6.3437,
      "mean_token_accuracy": 0.16317939907312393,
      "num_tokens": 13844368.0,
      "step": 5470
    },
    {
      "entropy": 6.489554214477539,
      "epoch": 0.6318522792844777,
      "grad_norm": 2.828125,
      "learning_rate": 0.000496977163598997,
      "loss": 6.4002,
      "mean_token_accuracy": 0.16044973731040954,
      "num_tokens": 13857261.0,
      "step": 5475
    },
    {
      "entropy": 6.304276037216186,
      "epoch": 0.6324293133294865,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969704185839239,
      "loss": 6.2707,
      "mean_token_accuracy": 0.15524207204580306,
      "num_tokens": 13869438.0,
      "step": 5480
    },
    {
      "entropy": 6.527767086029053,
      "epoch": 0.6330063473744951,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004969636661029859,
      "loss": 6.4693,
      "mean_token_accuracy": 0.14667147621512414,
      "num_tokens": 13883353.0,
      "step": 5485
    },
    {
      "entropy": 6.453281211853027,
      "epoch": 0.6335833814195038,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004969569061564103,
      "loss": 6.4342,
      "mean_token_accuracy": 0.16208689510822297,
      "num_tokens": 13896950.0,
      "step": 5490
    },
    {
      "entropy": 6.5273314952850345,
      "epoch": 0.6341604154645124,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004969501387444245,
      "loss": 6.409,
      "mean_token_accuracy": 0.15331158488988877,
      "num_tokens": 13910447.0,
      "step": 5495
    },
    {
      "entropy": 6.448615264892578,
      "epoch": 0.634737449509521,
      "grad_norm": 1.125,
      "learning_rate": 0.0004969433638672559,
      "loss": 6.4131,
      "mean_token_accuracy": 0.1641377866268158,
      "num_tokens": 13922894.0,
      "step": 5500
    },
    {
      "entropy": 6.470922517776489,
      "epoch": 0.6353144835545297,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969365815251325,
      "loss": 6.4349,
      "mean_token_accuracy": 0.15074554085731506,
      "num_tokens": 13936558.0,
      "step": 5505
    },
    {
      "entropy": 6.49364275932312,
      "epoch": 0.6358915175995383,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004969297917182825,
      "loss": 6.3486,
      "mean_token_accuracy": 0.15768291652202607,
      "num_tokens": 13949005.0,
      "step": 5510
    },
    {
      "entropy": 6.509783792495727,
      "epoch": 0.6364685516445471,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969229944469342,
      "loss": 6.4245,
      "mean_token_accuracy": 0.14853326678276063,
      "num_tokens": 13962285.0,
      "step": 5515
    },
    {
      "entropy": 6.48953537940979,
      "epoch": 0.6370455856895557,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004969161897113162,
      "loss": 6.4058,
      "mean_token_accuracy": 0.15401887595653535,
      "num_tokens": 13976037.0,
      "step": 5520
    },
    {
      "entropy": 6.460662126541138,
      "epoch": 0.6376226197345644,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004969093775116574,
      "loss": 6.3599,
      "mean_token_accuracy": 0.15469125062227249,
      "num_tokens": 13989028.0,
      "step": 5525
    },
    {
      "entropy": 6.354991722106933,
      "epoch": 0.638199653779573,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004969025578481869,
      "loss": 6.3485,
      "mean_token_accuracy": 0.15838611572980882,
      "num_tokens": 14001814.0,
      "step": 5530
    },
    {
      "entropy": 6.5595146179199215,
      "epoch": 0.6387766878245816,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496895730721134,
      "loss": 6.4544,
      "mean_token_accuracy": 0.14454537704586984,
      "num_tokens": 14014305.0,
      "step": 5535
    },
    {
      "entropy": 6.438262844085694,
      "epoch": 0.6393537218695903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004968888961307286,
      "loss": 6.3267,
      "mean_token_accuracy": 0.16070610731840135,
      "num_tokens": 14027475.0,
      "step": 5540
    },
    {
      "entropy": 6.447474813461303,
      "epoch": 0.6399307559145989,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004968820540772003,
      "loss": 6.3943,
      "mean_token_accuracy": 0.15843632072210312,
      "num_tokens": 14041031.0,
      "step": 5545
    },
    {
      "entropy": 6.407632684707641,
      "epoch": 0.6405077899596077,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004968752045607794,
      "loss": 6.345,
      "mean_token_accuracy": 0.15532173216342926,
      "num_tokens": 14054138.0,
      "step": 5550
    },
    {
      "entropy": 6.498112392425537,
      "epoch": 0.6410848240046163,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004968683475816961,
      "loss": 6.3824,
      "mean_token_accuracy": 0.15433506965637206,
      "num_tokens": 14066708.0,
      "step": 5555
    },
    {
      "entropy": 6.413477897644043,
      "epoch": 0.641661858049625,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004968614831401811,
      "loss": 6.4372,
      "mean_token_accuracy": 0.14834733828902244,
      "num_tokens": 14080251.0,
      "step": 5560
    },
    {
      "entropy": 6.533289051055908,
      "epoch": 0.6422388920946336,
      "grad_norm": 1.25,
      "learning_rate": 0.0004968546112364654,
      "loss": 6.4059,
      "mean_token_accuracy": 0.15228572264313697,
      "num_tokens": 14092622.0,
      "step": 5565
    },
    {
      "entropy": 6.488448190689087,
      "epoch": 0.6428159261396422,
      "grad_norm": 1.09375,
      "learning_rate": 0.00049684773187078,
      "loss": 6.4901,
      "mean_token_accuracy": 0.14898357838392257,
      "num_tokens": 14107033.0,
      "step": 5570
    },
    {
      "entropy": 6.364928388595581,
      "epoch": 0.6433929601846509,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004968408450433565,
      "loss": 6.2017,
      "mean_token_accuracy": 0.15701743066310883,
      "num_tokens": 14119264.0,
      "step": 5575
    },
    {
      "entropy": 6.429807901382446,
      "epoch": 0.6439699942296595,
      "grad_norm": 1.375,
      "learning_rate": 0.0004968339507544263,
      "loss": 6.3988,
      "mean_token_accuracy": 0.1542530760169029,
      "num_tokens": 14131360.0,
      "step": 5580
    },
    {
      "entropy": 6.443619251251221,
      "epoch": 0.6445470282746683,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004968270490042212,
      "loss": 6.3842,
      "mean_token_accuracy": 0.1619213879108429,
      "num_tokens": 14143236.0,
      "step": 5585
    },
    {
      "entropy": 6.433837938308716,
      "epoch": 0.6451240623196769,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004968201397929737,
      "loss": 6.4,
      "mean_token_accuracy": 0.1572832927107811,
      "num_tokens": 14156470.0,
      "step": 5590
    },
    {
      "entropy": 6.525268507003784,
      "epoch": 0.6457010963646855,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004968132231209158,
      "loss": 6.4646,
      "mean_token_accuracy": 0.1507378861308098,
      "num_tokens": 14169359.0,
      "step": 5595
    },
    {
      "entropy": 6.43701810836792,
      "epoch": 0.6462781304096942,
      "grad_norm": 1.125,
      "learning_rate": 0.0004968062989882803,
      "loss": 6.4709,
      "mean_token_accuracy": 0.15128603354096412,
      "num_tokens": 14181980.0,
      "step": 5600
    },
    {
      "entropy": 6.559728527069092,
      "epoch": 0.6468551644547028,
      "grad_norm": 1.125,
      "learning_rate": 0.0004967993673953003,
      "loss": 6.433,
      "mean_token_accuracy": 0.15603036880493165,
      "num_tokens": 14195456.0,
      "step": 5605
    },
    {
      "entropy": 6.483079195022583,
      "epoch": 0.6474321984997115,
      "grad_norm": 1.25,
      "learning_rate": 0.0004967924283422087,
      "loss": 6.3258,
      "mean_token_accuracy": 0.16091519445180893,
      "num_tokens": 14208269.0,
      "step": 5610
    },
    {
      "entropy": 6.397562265396118,
      "epoch": 0.6480092325447201,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000496785481829239,
      "loss": 6.4322,
      "mean_token_accuracy": 0.15433711111545562,
      "num_tokens": 14221474.0,
      "step": 5615
    },
    {
      "entropy": 6.533019590377807,
      "epoch": 0.6485862665897288,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004967785278566245,
      "loss": 6.4061,
      "mean_token_accuracy": 0.1547924056649208,
      "num_tokens": 14234185.0,
      "step": 5620
    },
    {
      "entropy": 6.418440961837769,
      "epoch": 0.6491633006347375,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004967715664245997,
      "loss": 6.3589,
      "mean_token_accuracy": 0.1545765370130539,
      "num_tokens": 14246853.0,
      "step": 5625
    },
    {
      "entropy": 6.416825199127198,
      "epoch": 0.6497403346797461,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004967645975333983,
      "loss": 6.4719,
      "mean_token_accuracy": 0.15569017976522445,
      "num_tokens": 14260047.0,
      "step": 5630
    },
    {
      "entropy": 6.428913116455078,
      "epoch": 0.6503173687247548,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004967576211832548,
      "loss": 6.335,
      "mean_token_accuracy": 0.16324738264083863,
      "num_tokens": 14273478.0,
      "step": 5635
    },
    {
      "entropy": 6.456370687484741,
      "epoch": 0.6508944027697634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967506373744039,
      "loss": 6.401,
      "mean_token_accuracy": 0.15274422019720077,
      "num_tokens": 14286085.0,
      "step": 5640
    },
    {
      "entropy": 6.430538129806519,
      "epoch": 0.651471436814772,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004967436461070805,
      "loss": 6.3692,
      "mean_token_accuracy": 0.1592407152056694,
      "num_tokens": 14299472.0,
      "step": 5645
    },
    {
      "entropy": 6.490675067901611,
      "epoch": 0.6520484708597807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004967366473815196,
      "loss": 6.336,
      "mean_token_accuracy": 0.15632698833942413,
      "num_tokens": 14312624.0,
      "step": 5650
    },
    {
      "entropy": 6.445674705505371,
      "epoch": 0.6526255049047894,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004967296411979568,
      "loss": 6.3308,
      "mean_token_accuracy": 0.16287260502576828,
      "num_tokens": 14325418.0,
      "step": 5655
    },
    {
      "entropy": 6.4445984840393065,
      "epoch": 0.6532025389497981,
      "grad_norm": 1.5,
      "learning_rate": 0.0004967226275566275,
      "loss": 6.3672,
      "mean_token_accuracy": 0.1629082068800926,
      "num_tokens": 14337360.0,
      "step": 5660
    },
    {
      "entropy": 6.445386791229248,
      "epoch": 0.6537795729948067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000496715606457768,
      "loss": 6.3572,
      "mean_token_accuracy": 0.15481363832950593,
      "num_tokens": 14349119.0,
      "step": 5665
    },
    {
      "entropy": 6.4672932624816895,
      "epoch": 0.6543566070398154,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004967085779016142,
      "loss": 6.3848,
      "mean_token_accuracy": 0.16024565547704697,
      "num_tokens": 14361760.0,
      "step": 5670
    },
    {
      "entropy": 6.423304557800293,
      "epoch": 0.654933641084824,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004967015418884022,
      "loss": 6.3657,
      "mean_token_accuracy": 0.15302406400442123,
      "num_tokens": 14374725.0,
      "step": 5675
    },
    {
      "entropy": 6.4012918949127195,
      "epoch": 0.6555106751298326,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004966944984183692,
      "loss": 6.3318,
      "mean_token_accuracy": 0.155929097533226,
      "num_tokens": 14387267.0,
      "step": 5680
    },
    {
      "entropy": 6.516444730758667,
      "epoch": 0.6560877091748413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004966874474917518,
      "loss": 6.3887,
      "mean_token_accuracy": 0.156803160905838,
      "num_tokens": 14400531.0,
      "step": 5685
    },
    {
      "entropy": 6.489614677429199,
      "epoch": 0.65666474321985,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004966803891087873,
      "loss": 6.4069,
      "mean_token_accuracy": 0.1479627177119255,
      "num_tokens": 14412507.0,
      "step": 5690
    },
    {
      "entropy": 6.422414636611938,
      "epoch": 0.6572417772648587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000496673323269713,
      "loss": 6.3767,
      "mean_token_accuracy": 0.1557417944073677,
      "num_tokens": 14424934.0,
      "step": 5695
    },
    {
      "entropy": 6.513529872894287,
      "epoch": 0.6578188113098673,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004966662499747666,
      "loss": 6.4852,
      "mean_token_accuracy": 0.14428578540682793,
      "num_tokens": 14438875.0,
      "step": 5700
    },
    {
      "entropy": 6.384660053253174,
      "epoch": 0.6583958453548759,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004966591692241859,
      "loss": 6.3007,
      "mean_token_accuracy": 0.1601344585418701,
      "num_tokens": 14451089.0,
      "step": 5705
    },
    {
      "entropy": 6.498195934295654,
      "epoch": 0.6589728793998846,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004966520810182092,
      "loss": 6.4322,
      "mean_token_accuracy": 0.15660066306591033,
      "num_tokens": 14463557.0,
      "step": 5710
    },
    {
      "entropy": 6.441149425506592,
      "epoch": 0.6595499134448932,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004966449853570749,
      "loss": 6.3287,
      "mean_token_accuracy": 0.15349343568086624,
      "num_tokens": 14475865.0,
      "step": 5715
    },
    {
      "entropy": 6.5019457817077635,
      "epoch": 0.6601269474899019,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004966378822410215,
      "loss": 6.4257,
      "mean_token_accuracy": 0.1521094784140587,
      "num_tokens": 14487534.0,
      "step": 5720
    },
    {
      "entropy": 6.501988220214844,
      "epoch": 0.6607039815349106,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004966307716702881,
      "loss": 6.3778,
      "mean_token_accuracy": 0.15803005397319794,
      "num_tokens": 14500006.0,
      "step": 5725
    },
    {
      "entropy": 6.485233545303345,
      "epoch": 0.6612810155799193,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004966236536451138,
      "loss": 6.5159,
      "mean_token_accuracy": 0.1492684841156006,
      "num_tokens": 14512844.0,
      "step": 5730
    },
    {
      "entropy": 6.489582204818726,
      "epoch": 0.6618580496249279,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496616528165738,
      "loss": 6.4387,
      "mean_token_accuracy": 0.14985530823469162,
      "num_tokens": 14525555.0,
      "step": 5735
    },
    {
      "entropy": 6.457883834838867,
      "epoch": 0.6624350836699365,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004966093952324003,
      "loss": 6.4158,
      "mean_token_accuracy": 0.14997079819440842,
      "num_tokens": 14537861.0,
      "step": 5740
    },
    {
      "entropy": 6.472036552429199,
      "epoch": 0.6630121177149452,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004966022548453406,
      "loss": 6.3897,
      "mean_token_accuracy": 0.1547412559390068,
      "num_tokens": 14550116.0,
      "step": 5745
    },
    {
      "entropy": 6.530148077011108,
      "epoch": 0.6635891517599538,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004965951070047993,
      "loss": 6.3645,
      "mean_token_accuracy": 0.1600572571158409,
      "num_tokens": 14561841.0,
      "step": 5750
    },
    {
      "entropy": 6.504292011260986,
      "epoch": 0.6641661858049625,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004965879517110166,
      "loss": 6.4188,
      "mean_token_accuracy": 0.15461863577365875,
      "num_tokens": 14574889.0,
      "step": 5755
    },
    {
      "entropy": 6.3978352546691895,
      "epoch": 0.6647432198499712,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004965807889642333,
      "loss": 6.4055,
      "mean_token_accuracy": 0.16118682324886321,
      "num_tokens": 14587032.0,
      "step": 5760
    },
    {
      "entropy": 6.445254135131836,
      "epoch": 0.6653202538949798,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00049657361876469,
      "loss": 6.3496,
      "mean_token_accuracy": 0.15695588439702987,
      "num_tokens": 14599654.0,
      "step": 5765
    },
    {
      "entropy": 6.441715860366822,
      "epoch": 0.6658972879399885,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004965664411126282,
      "loss": 6.2881,
      "mean_token_accuracy": 0.1609531193971634,
      "num_tokens": 14611740.0,
      "step": 5770
    },
    {
      "entropy": 6.444937467575073,
      "epoch": 0.6664743219849971,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004965592560082894,
      "loss": 6.3812,
      "mean_token_accuracy": 0.15895504653453826,
      "num_tokens": 14625038.0,
      "step": 5775
    },
    {
      "entropy": 6.482732677459717,
      "epoch": 0.6670513560300058,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004965520634519149,
      "loss": 6.4483,
      "mean_token_accuracy": 0.1515585035085678,
      "num_tokens": 14638318.0,
      "step": 5780
    },
    {
      "entropy": 6.4918629169464115,
      "epoch": 0.6676283900750144,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004965448634437468,
      "loss": 6.4173,
      "mean_token_accuracy": 0.15017091631889343,
      "num_tokens": 14651519.0,
      "step": 5785
    },
    {
      "entropy": 6.548173141479492,
      "epoch": 0.668205424120023,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965376559840272,
      "loss": 6.3733,
      "mean_token_accuracy": 0.15527796149253845,
      "num_tokens": 14662994.0,
      "step": 5790
    },
    {
      "entropy": 6.433021783828735,
      "epoch": 0.6687824581650318,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004965304410729986,
      "loss": 6.354,
      "mean_token_accuracy": 0.15886851251125336,
      "num_tokens": 14675488.0,
      "step": 5795
    },
    {
      "entropy": 6.427807283401489,
      "epoch": 0.6693594922100404,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004965232187109037,
      "loss": 6.2822,
      "mean_token_accuracy": 0.1672770693898201,
      "num_tokens": 14688512.0,
      "step": 5800
    },
    {
      "entropy": 6.499064922332764,
      "epoch": 0.6699365262550491,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004965159888979854,
      "loss": 6.3532,
      "mean_token_accuracy": 0.15719753950834275,
      "num_tokens": 14700740.0,
      "step": 5805
    },
    {
      "entropy": 6.412437057495117,
      "epoch": 0.6705135603000577,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004965087516344869,
      "loss": 6.3985,
      "mean_token_accuracy": 0.1540493041276932,
      "num_tokens": 14713769.0,
      "step": 5810
    },
    {
      "entropy": 6.440626096725464,
      "epoch": 0.6710905943450663,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004965015069206515,
      "loss": 6.2844,
      "mean_token_accuracy": 0.16277436017990113,
      "num_tokens": 14725710.0,
      "step": 5815
    },
    {
      "entropy": 6.488617420196533,
      "epoch": 0.671667628390075,
      "grad_norm": 1.234375,
      "learning_rate": 0.000496494254756723,
      "loss": 6.2693,
      "mean_token_accuracy": 0.16651962399482728,
      "num_tokens": 14738504.0,
      "step": 5820
    },
    {
      "entropy": 6.411191701889038,
      "epoch": 0.6722446624350836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004964869951429451,
      "loss": 6.3907,
      "mean_token_accuracy": 0.1528874784708023,
      "num_tokens": 14750344.0,
      "step": 5825
    },
    {
      "entropy": 6.4745879650115965,
      "epoch": 0.6728216964800924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004964797280795622,
      "loss": 6.4371,
      "mean_token_accuracy": 0.15475648939609526,
      "num_tokens": 14764316.0,
      "step": 5830
    },
    {
      "entropy": 6.491079187393188,
      "epoch": 0.673398730525101,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004964724535668188,
      "loss": 6.3532,
      "mean_token_accuracy": 0.154931178689003,
      "num_tokens": 14776404.0,
      "step": 5835
    },
    {
      "entropy": 6.527394390106201,
      "epoch": 0.6739757645701097,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004964651716049593,
      "loss": 6.4032,
      "mean_token_accuracy": 0.15441878139972687,
      "num_tokens": 14788843.0,
      "step": 5840
    },
    {
      "entropy": 6.389228677749633,
      "epoch": 0.6745527986151183,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004964578821942288,
      "loss": 6.3468,
      "mean_token_accuracy": 0.15969842821359634,
      "num_tokens": 14802662.0,
      "step": 5845
    },
    {
      "entropy": 6.394847726821899,
      "epoch": 0.6751298326601269,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004964505853348724,
      "loss": 6.397,
      "mean_token_accuracy": 0.14911102056503295,
      "num_tokens": 14814354.0,
      "step": 5850
    },
    {
      "entropy": 6.425918245315552,
      "epoch": 0.6757068667051356,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004964432810271356,
      "loss": 6.2954,
      "mean_token_accuracy": 0.16615222096443177,
      "num_tokens": 14826404.0,
      "step": 5855
    },
    {
      "entropy": 6.446470594406128,
      "epoch": 0.6762839007501442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004964359692712641,
      "loss": 6.366,
      "mean_token_accuracy": 0.15671302229166031,
      "num_tokens": 14840046.0,
      "step": 5860
    },
    {
      "entropy": 6.457760667800903,
      "epoch": 0.676860934795153,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004964286500675037,
      "loss": 6.3953,
      "mean_token_accuracy": 0.1482774406671524,
      "num_tokens": 14853566.0,
      "step": 5865
    },
    {
      "entropy": 6.436824560165405,
      "epoch": 0.6774379688401616,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004964213234161007,
      "loss": 6.3278,
      "mean_token_accuracy": 0.15916989594697953,
      "num_tokens": 14866538.0,
      "step": 5870
    },
    {
      "entropy": 6.392537069320679,
      "epoch": 0.6780150028851702,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004964139893173014,
      "loss": 6.2944,
      "mean_token_accuracy": 0.16514576077461243,
      "num_tokens": 14879237.0,
      "step": 5875
    },
    {
      "entropy": 6.483046245574951,
      "epoch": 0.6785920369301789,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004964066477713525,
      "loss": 6.4018,
      "mean_token_accuracy": 0.15573488026857377,
      "num_tokens": 14891295.0,
      "step": 5880
    },
    {
      "entropy": 6.415683555603027,
      "epoch": 0.6791690709751875,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963992987785011,
      "loss": 6.3904,
      "mean_token_accuracy": 0.1575924351811409,
      "num_tokens": 14902927.0,
      "step": 5885
    },
    {
      "entropy": 6.421587514877319,
      "epoch": 0.6797461050201962,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004963919423389942,
      "loss": 6.375,
      "mean_token_accuracy": 0.15122182816267013,
      "num_tokens": 14915203.0,
      "step": 5890
    },
    {
      "entropy": 6.392851495742798,
      "epoch": 0.6803231390652048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963845784530794,
      "loss": 6.2337,
      "mean_token_accuracy": 0.16648308783769608,
      "num_tokens": 14927428.0,
      "step": 5895
    },
    {
      "entropy": 6.382313442230225,
      "epoch": 0.6809001731102136,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000496377207121004,
      "loss": 6.3774,
      "mean_token_accuracy": 0.1617804765701294,
      "num_tokens": 14939979.0,
      "step": 5900
    },
    {
      "entropy": 6.505252695083618,
      "epoch": 0.6814772071552222,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004963698283430164,
      "loss": 6.3816,
      "mean_token_accuracy": 0.15156475305557252,
      "num_tokens": 14952292.0,
      "step": 5905
    },
    {
      "entropy": 6.482873964309692,
      "epoch": 0.6820542412002308,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004963624421193646,
      "loss": 6.386,
      "mean_token_accuracy": 0.1532788097858429,
      "num_tokens": 14964942.0,
      "step": 5910
    },
    {
      "entropy": 6.450105810165406,
      "epoch": 0.6826312752452395,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000496355048450297,
      "loss": 6.4974,
      "mean_token_accuracy": 0.14616498947143555,
      "num_tokens": 14979196.0,
      "step": 5915
    },
    {
      "entropy": 6.507519197463989,
      "epoch": 0.6832083092902481,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004963476473360623,
      "loss": 6.3601,
      "mean_token_accuracy": 0.1571754828095436,
      "num_tokens": 14991162.0,
      "step": 5920
    },
    {
      "entropy": 6.466042995452881,
      "epoch": 0.6837853433352568,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004963402387769094,
      "loss": 6.3804,
      "mean_token_accuracy": 0.1588766470551491,
      "num_tokens": 15003207.0,
      "step": 5925
    },
    {
      "entropy": 6.41431040763855,
      "epoch": 0.6843623773802654,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004963328227730876,
      "loss": 6.4254,
      "mean_token_accuracy": 0.14741162210702896,
      "num_tokens": 15016633.0,
      "step": 5930
    },
    {
      "entropy": 6.442387247085572,
      "epoch": 0.684939411425274,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004963253993248461,
      "loss": 6.343,
      "mean_token_accuracy": 0.1599098488688469,
      "num_tokens": 15029256.0,
      "step": 5935
    },
    {
      "entropy": 6.425724887847901,
      "epoch": 0.6855164454702828,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004963179684324349,
      "loss": 6.3,
      "mean_token_accuracy": 0.15923123359680175,
      "num_tokens": 15041162.0,
      "step": 5940
    },
    {
      "entropy": 6.402348136901855,
      "epoch": 0.6860934795152914,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004963105300961036,
      "loss": 6.3577,
      "mean_token_accuracy": 0.16127298772335052,
      "num_tokens": 15054122.0,
      "step": 5945
    },
    {
      "entropy": 6.403144645690918,
      "epoch": 0.6866705135603001,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004963030843161026,
      "loss": 6.3328,
      "mean_token_accuracy": 0.15635056793689728,
      "num_tokens": 15066123.0,
      "step": 5950
    },
    {
      "entropy": 6.3505641460418705,
      "epoch": 0.6872475476053087,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004962956310926823,
      "loss": 6.2803,
      "mean_token_accuracy": 0.16309783309698106,
      "num_tokens": 15078192.0,
      "step": 5955
    },
    {
      "entropy": 6.487194633483886,
      "epoch": 0.6878245816503173,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004962881704260934,
      "loss": 6.3194,
      "mean_token_accuracy": 0.16090887039899826,
      "num_tokens": 15090227.0,
      "step": 5960
    },
    {
      "entropy": 6.460311222076416,
      "epoch": 0.688401615695326,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004962807023165868,
      "loss": 6.4389,
      "mean_token_accuracy": 0.159395045787096,
      "num_tokens": 15104093.0,
      "step": 5965
    },
    {
      "entropy": 6.396960067749023,
      "epoch": 0.6889786497403346,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962732267644138,
      "loss": 6.329,
      "mean_token_accuracy": 0.15944764763116837,
      "num_tokens": 15118064.0,
      "step": 5970
    },
    {
      "entropy": 6.463270425796509,
      "epoch": 0.6895556837853434,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004962657437698254,
      "loss": 6.2442,
      "mean_token_accuracy": 0.1626880243420601,
      "num_tokens": 15131686.0,
      "step": 5975
    },
    {
      "entropy": 6.419650077819824,
      "epoch": 0.690132717830352,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004962582533330739,
      "loss": 6.4456,
      "mean_token_accuracy": 0.1501251846551895,
      "num_tokens": 15143920.0,
      "step": 5980
    },
    {
      "entropy": 6.49657621383667,
      "epoch": 0.6907097518753607,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962507554544109,
      "loss": 6.3803,
      "mean_token_accuracy": 0.15870121121406555,
      "num_tokens": 15156989.0,
      "step": 5985
    },
    {
      "entropy": 6.5632383823394775,
      "epoch": 0.6912867859203693,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004962432501340886,
      "loss": 6.4622,
      "mean_token_accuracy": 0.14517625272274018,
      "num_tokens": 15169068.0,
      "step": 5990
    },
    {
      "entropy": 6.417279434204102,
      "epoch": 0.6918638199653779,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004962357373723596,
      "loss": 6.4305,
      "mean_token_accuracy": 0.1551764652132988,
      "num_tokens": 15182081.0,
      "step": 5995
    },
    {
      "entropy": 6.562622547149658,
      "epoch": 0.6924408540103866,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004962282171694763,
      "loss": 6.4065,
      "mean_token_accuracy": 0.1580890581011772,
      "num_tokens": 15194474.0,
      "step": 6000
    },
    {
      "epoch": 0.6924408540103866,
      "eval_entropy": 6.400503651743963,
      "eval_loss": 6.432427883148193,
      "eval_mean_token_accuracy": 0.15697687899999432,
      "eval_num_tokens": 15194474.0,
      "eval_runtime": 22.6959,
      "eval_samples_per_second": 1239.697,
      "eval_steps_per_second": 154.962,
      "step": 6000
    },
    {
      "entropy": 6.498694944381714,
      "epoch": 0.6930178880553952,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004962206895256919,
      "loss": 6.368,
      "mean_token_accuracy": 0.15929650962352754,
      "num_tokens": 15206371.0,
      "step": 6005
    },
    {
      "entropy": 6.367174577713013,
      "epoch": 0.693594922100404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004962131544412595,
      "loss": 6.3868,
      "mean_token_accuracy": 0.15386435389518738,
      "num_tokens": 15218913.0,
      "step": 6010
    },
    {
      "entropy": 6.497800350189209,
      "epoch": 0.6941719561454126,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004962056119164325,
      "loss": 6.3709,
      "mean_token_accuracy": 0.16197654753923416,
      "num_tokens": 15231647.0,
      "step": 6015
    },
    {
      "entropy": 6.4635841846466064,
      "epoch": 0.6947489901904212,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961980619514646,
      "loss": 6.3708,
      "mean_token_accuracy": 0.15795217901468278,
      "num_tokens": 15243764.0,
      "step": 6020
    },
    {
      "entropy": 6.443425607681275,
      "epoch": 0.6953260242354299,
      "grad_norm": 1.5,
      "learning_rate": 0.0004961905045466098,
      "loss": 6.4417,
      "mean_token_accuracy": 0.14838726967573165,
      "num_tokens": 15256165.0,
      "step": 6025
    },
    {
      "entropy": 6.429970169067383,
      "epoch": 0.6959030582804385,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004961829397021222,
      "loss": 6.29,
      "mean_token_accuracy": 0.16175810992717743,
      "num_tokens": 15269169.0,
      "step": 6030
    },
    {
      "entropy": 6.467025279998779,
      "epoch": 0.6964800923254472,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004961753674182564,
      "loss": 6.349,
      "mean_token_accuracy": 0.1544795885682106,
      "num_tokens": 15281530.0,
      "step": 6035
    },
    {
      "entropy": 6.419221019744873,
      "epoch": 0.6970571263704558,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004961677876952669,
      "loss": 6.3357,
      "mean_token_accuracy": 0.15590357780456543,
      "num_tokens": 15295578.0,
      "step": 6040
    },
    {
      "entropy": 6.406968784332276,
      "epoch": 0.6976341604154646,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004961602005334087,
      "loss": 6.397,
      "mean_token_accuracy": 0.15791086107492447,
      "num_tokens": 15308258.0,
      "step": 6045
    },
    {
      "entropy": 6.478301000595093,
      "epoch": 0.6982111944604732,
      "grad_norm": 1.15625,
      "learning_rate": 0.000496152605932937,
      "loss": 6.4014,
      "mean_token_accuracy": 0.15565335899591445,
      "num_tokens": 15320370.0,
      "step": 6050
    },
    {
      "entropy": 6.512187814712524,
      "epoch": 0.6987882285054818,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004961450038941074,
      "loss": 6.4014,
      "mean_token_accuracy": 0.1576465040445328,
      "num_tokens": 15332350.0,
      "step": 6055
    },
    {
      "entropy": 6.39268012046814,
      "epoch": 0.6993652625504905,
      "grad_norm": 1.75,
      "learning_rate": 0.0004961373944171754,
      "loss": 6.3002,
      "mean_token_accuracy": 0.16454465687274933,
      "num_tokens": 15343424.0,
      "step": 6060
    },
    {
      "entropy": 6.427528715133667,
      "epoch": 0.6999422965954991,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004961297775023968,
      "loss": 6.3385,
      "mean_token_accuracy": 0.1592035174369812,
      "num_tokens": 15356126.0,
      "step": 6065
    },
    {
      "entropy": 6.345870637893677,
      "epoch": 0.7005193306405078,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000496122153150028,
      "loss": 6.2546,
      "mean_token_accuracy": 0.16224194318056107,
      "num_tokens": 15369253.0,
      "step": 6070
    },
    {
      "entropy": 6.557456541061401,
      "epoch": 0.7010963646855164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004961145213603255,
      "loss": 6.4026,
      "mean_token_accuracy": 0.152187579870224,
      "num_tokens": 15382679.0,
      "step": 6075
    },
    {
      "entropy": 6.414884233474732,
      "epoch": 0.7016733987305251,
      "grad_norm": 1.484375,
      "learning_rate": 0.000496106882133546,
      "loss": 6.3941,
      "mean_token_accuracy": 0.15569975972175598,
      "num_tokens": 15396052.0,
      "step": 6080
    },
    {
      "entropy": 6.4202300071716305,
      "epoch": 0.7022504327755338,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004960992354699463,
      "loss": 6.444,
      "mean_token_accuracy": 0.15025331526994706,
      "num_tokens": 15409249.0,
      "step": 6085
    },
    {
      "entropy": 6.48999605178833,
      "epoch": 0.7028274668205424,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004960915813697836,
      "loss": 6.3182,
      "mean_token_accuracy": 0.15329759567975998,
      "num_tokens": 15421060.0,
      "step": 6090
    },
    {
      "entropy": 6.440291547775269,
      "epoch": 0.7034045008655511,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960839198333154,
      "loss": 6.3987,
      "mean_token_accuracy": 0.15348696261644362,
      "num_tokens": 15433083.0,
      "step": 6095
    },
    {
      "entropy": 6.496385335922241,
      "epoch": 0.7039815349105597,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004960762508607994,
      "loss": 6.2319,
      "mean_token_accuracy": 0.16515322625637055,
      "num_tokens": 15446664.0,
      "step": 6100
    },
    {
      "entropy": 6.437723684310913,
      "epoch": 0.7045585689555683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004960685744524934,
      "loss": 6.3833,
      "mean_token_accuracy": 0.15098819732666016,
      "num_tokens": 15458609.0,
      "step": 6105
    },
    {
      "entropy": 6.418120431900024,
      "epoch": 0.705135603000577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004960608906086558,
      "loss": 6.2931,
      "mean_token_accuracy": 0.15668573081493378,
      "num_tokens": 15470091.0,
      "step": 6110
    },
    {
      "entropy": 6.441114664077759,
      "epoch": 0.7057126370455857,
      "grad_norm": 1.234375,
      "learning_rate": 0.000496053199329545,
      "loss": 6.3913,
      "mean_token_accuracy": 0.15550165474414826,
      "num_tokens": 15483557.0,
      "step": 6115
    },
    {
      "entropy": 6.462738084793091,
      "epoch": 0.7062896710905944,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004960455006154196,
      "loss": 6.3639,
      "mean_token_accuracy": 0.15281105488538743,
      "num_tokens": 15497437.0,
      "step": 6120
    },
    {
      "entropy": 6.461950254440308,
      "epoch": 0.706866705135603,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004960377944665386,
      "loss": 6.3832,
      "mean_token_accuracy": 0.15458154678344727,
      "num_tokens": 15509942.0,
      "step": 6125
    },
    {
      "entropy": 6.415175914764404,
      "epoch": 0.7074437391806117,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004960300808831611,
      "loss": 6.3258,
      "mean_token_accuracy": 0.1606331631541252,
      "num_tokens": 15522866.0,
      "step": 6130
    },
    {
      "entropy": 6.437211227416992,
      "epoch": 0.7080207732256203,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004960223598655467,
      "loss": 6.3391,
      "mean_token_accuracy": 0.15498406440019608,
      "num_tokens": 15534731.0,
      "step": 6135
    },
    {
      "entropy": 6.390599346160888,
      "epoch": 0.7085978072706289,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000496014631413955,
      "loss": 6.3549,
      "mean_token_accuracy": 0.15641144812107086,
      "num_tokens": 15546401.0,
      "step": 6140
    },
    {
      "entropy": 6.464565753936768,
      "epoch": 0.7091748413156376,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004960068955286459,
      "loss": 6.3453,
      "mean_token_accuracy": 0.1545490711927414,
      "num_tokens": 15559532.0,
      "step": 6145
    },
    {
      "entropy": 6.450887155532837,
      "epoch": 0.7097518753606463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004959991522098797,
      "loss": 6.3283,
      "mean_token_accuracy": 0.1544230192899704,
      "num_tokens": 15572426.0,
      "step": 6150
    },
    {
      "entropy": 6.5202630996704105,
      "epoch": 0.710328909405655,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004959914014579168,
      "loss": 6.3851,
      "mean_token_accuracy": 0.15095667839050292,
      "num_tokens": 15584422.0,
      "step": 6155
    },
    {
      "entropy": 6.4888464450836185,
      "epoch": 0.7109059434506636,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004959836432730178,
      "loss": 6.3886,
      "mean_token_accuracy": 0.1545424059033394,
      "num_tokens": 15595538.0,
      "step": 6160
    },
    {
      "entropy": 6.371583795547485,
      "epoch": 0.7114829774956722,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004959758776554438,
      "loss": 6.3184,
      "mean_token_accuracy": 0.15887101888656616,
      "num_tokens": 15608315.0,
      "step": 6165
    },
    {
      "entropy": 6.514468574523926,
      "epoch": 0.7120600115406809,
      "grad_norm": 1.421875,
      "learning_rate": 0.000495968104605456,
      "loss": 6.4199,
      "mean_token_accuracy": 0.15411719381809236,
      "num_tokens": 15620181.0,
      "step": 6170
    },
    {
      "entropy": 6.394854784011841,
      "epoch": 0.7126370455856895,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004959603241233157,
      "loss": 6.2669,
      "mean_token_accuracy": 0.16061466783285142,
      "num_tokens": 15632100.0,
      "step": 6175
    },
    {
      "entropy": 6.404596328735352,
      "epoch": 0.7132140796306982,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959525362092846,
      "loss": 6.2783,
      "mean_token_accuracy": 0.1624158054590225,
      "num_tokens": 15644185.0,
      "step": 6180
    },
    {
      "entropy": 6.3875651359558105,
      "epoch": 0.7137911136757069,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004959447408636247,
      "loss": 6.3236,
      "mean_token_accuracy": 0.16416954696178437,
      "num_tokens": 15657761.0,
      "step": 6185
    },
    {
      "entropy": 6.419309377670288,
      "epoch": 0.7143681477207156,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004959369380865983,
      "loss": 6.3595,
      "mean_token_accuracy": 0.15183151364326478,
      "num_tokens": 15671135.0,
      "step": 6190
    },
    {
      "entropy": 6.441372919082641,
      "epoch": 0.7149451817657242,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004959291278784676,
      "loss": 6.3455,
      "mean_token_accuracy": 0.15111906975507736,
      "num_tokens": 15683286.0,
      "step": 6195
    },
    {
      "entropy": 6.3563837051391605,
      "epoch": 0.7155222158107328,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004959213102394954,
      "loss": 6.2707,
      "mean_token_accuracy": 0.167560413479805,
      "num_tokens": 15697535.0,
      "step": 6200
    },
    {
      "entropy": 6.3913243293762205,
      "epoch": 0.7160992498557415,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004959134851699449,
      "loss": 6.3598,
      "mean_token_accuracy": 0.15648961514234544,
      "num_tokens": 15708748.0,
      "step": 6205
    },
    {
      "entropy": 6.477129936218262,
      "epoch": 0.7166762839007501,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004959056526700788,
      "loss": 6.3456,
      "mean_token_accuracy": 0.15493126660585405,
      "num_tokens": 15721720.0,
      "step": 6210
    },
    {
      "entropy": 6.409880685806274,
      "epoch": 0.7172533179457588,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004958978127401609,
      "loss": 6.2937,
      "mean_token_accuracy": 0.15915343463420867,
      "num_tokens": 15734854.0,
      "step": 6215
    },
    {
      "entropy": 6.413718938827515,
      "epoch": 0.7178303519907675,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004958899653804548,
      "loss": 6.3539,
      "mean_token_accuracy": 0.1601460188627243,
      "num_tokens": 15747575.0,
      "step": 6220
    },
    {
      "entropy": 6.381495141983033,
      "epoch": 0.7184073860357761,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004958821105912246,
      "loss": 6.3453,
      "mean_token_accuracy": 0.16180343478918074,
      "num_tokens": 15760917.0,
      "step": 6225
    },
    {
      "entropy": 6.55801100730896,
      "epoch": 0.7189844200807848,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000495874248372734,
      "loss": 6.4266,
      "mean_token_accuracy": 0.14924235120415688,
      "num_tokens": 15774045.0,
      "step": 6230
    },
    {
      "entropy": 6.491186237335205,
      "epoch": 0.7195614541257934,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000495866378725248,
      "loss": 6.3379,
      "mean_token_accuracy": 0.15854604691267013,
      "num_tokens": 15786006.0,
      "step": 6235
    },
    {
      "entropy": 6.407785272598266,
      "epoch": 0.7201384881708021,
      "grad_norm": 1.265625,
      "learning_rate": 0.000495858501649031,
      "loss": 6.3539,
      "mean_token_accuracy": 0.15561899542808533,
      "num_tokens": 15798078.0,
      "step": 6240
    },
    {
      "entropy": 6.3833420753479,
      "epoch": 0.7207155222158107,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000495850617144348,
      "loss": 6.3406,
      "mean_token_accuracy": 0.15743837952613832,
      "num_tokens": 15810281.0,
      "step": 6245
    },
    {
      "entropy": 6.432937860488892,
      "epoch": 0.7212925562608193,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004958427252114643,
      "loss": 6.4155,
      "mean_token_accuracy": 0.1525063246488571,
      "num_tokens": 15822593.0,
      "step": 6250
    },
    {
      "entropy": 6.489366436004639,
      "epoch": 0.7218695903058281,
      "grad_norm": 1.078125,
      "learning_rate": 0.000495834825850645,
      "loss": 6.3534,
      "mean_token_accuracy": 0.15766695141792297,
      "num_tokens": 15835181.0,
      "step": 6255
    },
    {
      "entropy": 6.476584815979004,
      "epoch": 0.7224466243508367,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004958269190621562,
      "loss": 6.3561,
      "mean_token_accuracy": 0.1545557916164398,
      "num_tokens": 15847679.0,
      "step": 6260
    },
    {
      "entropy": 6.45306282043457,
      "epoch": 0.7230236583958454,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004958190048462637,
      "loss": 6.3914,
      "mean_token_accuracy": 0.15713143199682236,
      "num_tokens": 15860209.0,
      "step": 6265
    },
    {
      "entropy": 6.485035228729248,
      "epoch": 0.723600692440854,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004958110832032336,
      "loss": 6.3716,
      "mean_token_accuracy": 0.1563394472002983,
      "num_tokens": 15873165.0,
      "step": 6270
    },
    {
      "entropy": 6.390137529373169,
      "epoch": 0.7241777264858626,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958031541333322,
      "loss": 6.3932,
      "mean_token_accuracy": 0.15398206412792206,
      "num_tokens": 15885201.0,
      "step": 6275
    },
    {
      "entropy": 6.590787267684936,
      "epoch": 0.7247547605308713,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004957952176368267,
      "loss": 6.4638,
      "mean_token_accuracy": 0.1497330904006958,
      "num_tokens": 15899071.0,
      "step": 6280
    },
    {
      "entropy": 6.381853342056274,
      "epoch": 0.7253317945758799,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004957872737139836,
      "loss": 6.3386,
      "mean_token_accuracy": 0.15825416520237923,
      "num_tokens": 15912274.0,
      "step": 6285
    },
    {
      "entropy": 6.435012531280518,
      "epoch": 0.7259088286208887,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004957793223650702,
      "loss": 6.3326,
      "mean_token_accuracy": 0.16138885617256166,
      "num_tokens": 15925302.0,
      "step": 6290
    },
    {
      "entropy": 6.3799999237060545,
      "epoch": 0.7264858626658973,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000495771363590354,
      "loss": 6.392,
      "mean_token_accuracy": 0.1516269102692604,
      "num_tokens": 15938226.0,
      "step": 6295
    },
    {
      "entropy": 6.394499444961548,
      "epoch": 0.727062896710906,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004957633973901027,
      "loss": 6.384,
      "mean_token_accuracy": 0.15402197688817978,
      "num_tokens": 15951136.0,
      "step": 6300
    },
    {
      "entropy": 6.564555644989014,
      "epoch": 0.7276399307559146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004957554237645841,
      "loss": 6.3203,
      "mean_token_accuracy": 0.15522871762514115,
      "num_tokens": 15962820.0,
      "step": 6305
    },
    {
      "entropy": 6.393157958984375,
      "epoch": 0.7282169648009232,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004957474427140665,
      "loss": 6.3002,
      "mean_token_accuracy": 0.1554463192820549,
      "num_tokens": 15974871.0,
      "step": 6310
    },
    {
      "entropy": 6.330068969726563,
      "epoch": 0.7287939988459319,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004957394542388182,
      "loss": 6.2902,
      "mean_token_accuracy": 0.1576317049562931,
      "num_tokens": 15986916.0,
      "step": 6315
    },
    {
      "entropy": 6.503025722503662,
      "epoch": 0.7293710328909405,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004957314583391082,
      "loss": 6.4285,
      "mean_token_accuracy": 0.1535441353917122,
      "num_tokens": 15999735.0,
      "step": 6320
    },
    {
      "entropy": 6.462880992889405,
      "epoch": 0.7299480669359493,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004957234550152052,
      "loss": 6.3256,
      "mean_token_accuracy": 0.15937893241643905,
      "num_tokens": 16013137.0,
      "step": 6325
    },
    {
      "entropy": 6.423413181304932,
      "epoch": 0.7305251009809579,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004957154442673784,
      "loss": 6.3429,
      "mean_token_accuracy": 0.15815951824188232,
      "num_tokens": 16026750.0,
      "step": 6330
    },
    {
      "entropy": 6.467150354385376,
      "epoch": 0.7311021350259665,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957074260958972,
      "loss": 6.44,
      "mean_token_accuracy": 0.1503904193639755,
      "num_tokens": 16039028.0,
      "step": 6335
    },
    {
      "entropy": 6.422294855117798,
      "epoch": 0.7316791690709752,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004956994005010315,
      "loss": 6.3013,
      "mean_token_accuracy": 0.16040119528770447,
      "num_tokens": 16051349.0,
      "step": 6340
    },
    {
      "entropy": 6.459820508956909,
      "epoch": 0.7322562031159838,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000495691367483051,
      "loss": 6.3379,
      "mean_token_accuracy": 0.15461637526750566,
      "num_tokens": 16064015.0,
      "step": 6345
    },
    {
      "entropy": 6.490715360641479,
      "epoch": 0.7328332371609925,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004956833270422259,
      "loss": 6.3473,
      "mean_token_accuracy": 0.15606531947851182,
      "num_tokens": 16075571.0,
      "step": 6350
    },
    {
      "entropy": 6.381386137008667,
      "epoch": 0.7334102712060011,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004956752791788269,
      "loss": 6.365,
      "mean_token_accuracy": 0.16088762283325195,
      "num_tokens": 16088176.0,
      "step": 6355
    },
    {
      "entropy": 6.450045108795166,
      "epoch": 0.7339873052510099,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004956672238931244,
      "loss": 6.3975,
      "mean_token_accuracy": 0.15093077719211578,
      "num_tokens": 16099933.0,
      "step": 6360
    },
    {
      "entropy": 6.377976322174073,
      "epoch": 0.7345643392960185,
      "grad_norm": 1.125,
      "learning_rate": 0.0004956591611853894,
      "loss": 6.2235,
      "mean_token_accuracy": 0.16211253106594087,
      "num_tokens": 16113200.0,
      "step": 6365
    },
    {
      "entropy": 6.424292087554932,
      "epoch": 0.7351413733410271,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004956510910558931,
      "loss": 6.3801,
      "mean_token_accuracy": 0.16113436222076416,
      "num_tokens": 16125532.0,
      "step": 6370
    },
    {
      "entropy": 6.452923822402954,
      "epoch": 0.7357184073860358,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004956430135049071,
      "loss": 6.3211,
      "mean_token_accuracy": 0.15369930267333984,
      "num_tokens": 16137197.0,
      "step": 6375
    },
    {
      "entropy": 6.468174600601197,
      "epoch": 0.7362954414310444,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004956349285327028,
      "loss": 6.387,
      "mean_token_accuracy": 0.16085880845785142,
      "num_tokens": 16150481.0,
      "step": 6380
    },
    {
      "entropy": 6.512467575073242,
      "epoch": 0.7368724754760531,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956268361395523,
      "loss": 6.3866,
      "mean_token_accuracy": 0.16043339669704437,
      "num_tokens": 16162395.0,
      "step": 6385
    },
    {
      "entropy": 6.474280881881714,
      "epoch": 0.7374495095210617,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004956187363257278,
      "loss": 6.3575,
      "mean_token_accuracy": 0.15275195389986038,
      "num_tokens": 16174550.0,
      "step": 6390
    },
    {
      "entropy": 6.465876197814941,
      "epoch": 0.7380265435660704,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004956106290915017,
      "loss": 6.3673,
      "mean_token_accuracy": 0.1536040261387825,
      "num_tokens": 16187571.0,
      "step": 6395
    },
    {
      "entropy": 6.422832822799682,
      "epoch": 0.7386035776110791,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956025144371467,
      "loss": 6.3985,
      "mean_token_accuracy": 0.15287847071886063,
      "num_tokens": 16199684.0,
      "step": 6400
    },
    {
      "entropy": 6.385442638397217,
      "epoch": 0.7391806116560877,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004955943923629356,
      "loss": 6.3548,
      "mean_token_accuracy": 0.1548342451453209,
      "num_tokens": 16211935.0,
      "step": 6405
    },
    {
      "entropy": 6.44033145904541,
      "epoch": 0.7397576457010964,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004955862628691417,
      "loss": 6.3459,
      "mean_token_accuracy": 0.1508117288351059,
      "num_tokens": 16224346.0,
      "step": 6410
    },
    {
      "entropy": 6.487747716903686,
      "epoch": 0.740334679746105,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004955781259560386,
      "loss": 6.3897,
      "mean_token_accuracy": 0.14921767711639405,
      "num_tokens": 16238137.0,
      "step": 6415
    },
    {
      "entropy": 6.391067790985107,
      "epoch": 0.7409117137911136,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004955699816238995,
      "loss": 6.3537,
      "mean_token_accuracy": 0.15329435765743255,
      "num_tokens": 16250777.0,
      "step": 6420
    },
    {
      "entropy": 6.545683860778809,
      "epoch": 0.7414887478361223,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004955618298729988,
      "loss": 6.365,
      "mean_token_accuracy": 0.15731996297836304,
      "num_tokens": 16262440.0,
      "step": 6425
    },
    {
      "entropy": 6.4182055473327635,
      "epoch": 0.742065781881131,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955536707036105,
      "loss": 6.3149,
      "mean_token_accuracy": 0.15865564495325088,
      "num_tokens": 16275227.0,
      "step": 6430
    },
    {
      "entropy": 6.390710258483887,
      "epoch": 0.7426428159261397,
      "grad_norm": 1.234375,
      "learning_rate": 0.000495545504116009,
      "loss": 6.3697,
      "mean_token_accuracy": 0.16148028522729874,
      "num_tokens": 16287246.0,
      "step": 6435
    },
    {
      "entropy": 6.4247143268585205,
      "epoch": 0.7432198499711483,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004955373301104691,
      "loss": 6.25,
      "mean_token_accuracy": 0.16333142668008804,
      "num_tokens": 16298779.0,
      "step": 6440
    },
    {
      "entropy": 6.32188081741333,
      "epoch": 0.743796884016157,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004955291486872657,
      "loss": 6.2682,
      "mean_token_accuracy": 0.16090297400951387,
      "num_tokens": 16310965.0,
      "step": 6445
    },
    {
      "entropy": 6.425542116165161,
      "epoch": 0.7443739180611656,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004955209598466738,
      "loss": 6.2903,
      "mean_token_accuracy": 0.15480603873729706,
      "num_tokens": 16324025.0,
      "step": 6450
    },
    {
      "entropy": 6.422914981842041,
      "epoch": 0.7449509521061742,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000495512763588969,
      "loss": 6.3311,
      "mean_token_accuracy": 0.15080109387636184,
      "num_tokens": 16336152.0,
      "step": 6455
    },
    {
      "entropy": 6.359565687179566,
      "epoch": 0.7455279861511829,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004955045599144269,
      "loss": 6.3426,
      "mean_token_accuracy": 0.15447487831115722,
      "num_tokens": 16349118.0,
      "step": 6460
    },
    {
      "entropy": 6.373418140411377,
      "epoch": 0.7461050201961916,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004954963488233235,
      "loss": 6.3055,
      "mean_token_accuracy": 0.15982365757226943,
      "num_tokens": 16362986.0,
      "step": 6465
    },
    {
      "entropy": 6.465278196334839,
      "epoch": 0.7466820542412003,
      "grad_norm": 1.21875,
      "learning_rate": 0.000495488130315935,
      "loss": 6.4488,
      "mean_token_accuracy": 0.14422052949666977,
      "num_tokens": 16376384.0,
      "step": 6470
    },
    {
      "entropy": 6.436511468887329,
      "epoch": 0.7472590882862089,
      "grad_norm": 1.375,
      "learning_rate": 0.0004954799043925377,
      "loss": 6.2663,
      "mean_token_accuracy": 0.15960704237222673,
      "num_tokens": 16388137.0,
      "step": 6475
    },
    {
      "entropy": 6.442381572723389,
      "epoch": 0.7478361223312175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004954716710534082,
      "loss": 6.3868,
      "mean_token_accuracy": 0.15023239105939865,
      "num_tokens": 16400037.0,
      "step": 6480
    },
    {
      "entropy": 6.4429028034210205,
      "epoch": 0.7484131563762262,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004954634302988237,
      "loss": 6.3091,
      "mean_token_accuracy": 0.15836434438824654,
      "num_tokens": 16412065.0,
      "step": 6485
    },
    {
      "entropy": 6.423034954071045,
      "epoch": 0.7489901904212348,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954551821290612,
      "loss": 6.3568,
      "mean_token_accuracy": 0.15571586638689042,
      "num_tokens": 16424345.0,
      "step": 6490
    },
    {
      "entropy": 6.433213615417481,
      "epoch": 0.7495672244662435,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004954469265443981,
      "loss": 6.4097,
      "mean_token_accuracy": 0.1566193573176861,
      "num_tokens": 16436423.0,
      "step": 6495
    },
    {
      "entropy": 6.387920618057251,
      "epoch": 0.7501442585112522,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954386635451123,
      "loss": 6.3036,
      "mean_token_accuracy": 0.15891094803810119,
      "num_tokens": 16449496.0,
      "step": 6500
    },
    {
      "entropy": 6.435271167755127,
      "epoch": 0.7507212925562609,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004954303931314814,
      "loss": 6.3454,
      "mean_token_accuracy": 0.15708634853363038,
      "num_tokens": 16462193.0,
      "step": 6505
    },
    {
      "entropy": 6.427425193786621,
      "epoch": 0.7512983266012695,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954221153037837,
      "loss": 6.3555,
      "mean_token_accuracy": 0.15513187199831008,
      "num_tokens": 16474470.0,
      "step": 6510
    },
    {
      "entropy": 6.497897005081176,
      "epoch": 0.7518753606462781,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004954138300622978,
      "loss": 6.3487,
      "mean_token_accuracy": 0.15803459137678147,
      "num_tokens": 16487231.0,
      "step": 6515
    },
    {
      "entropy": 6.431686544418335,
      "epoch": 0.7524523946912868,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000495405537407302,
      "loss": 6.4133,
      "mean_token_accuracy": 0.1513650543987751,
      "num_tokens": 16500822.0,
      "step": 6520
    },
    {
      "entropy": 6.419403791427612,
      "epoch": 0.7530294287362954,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004953972373390755,
      "loss": 6.3641,
      "mean_token_accuracy": 0.1541321247816086,
      "num_tokens": 16514529.0,
      "step": 6525
    },
    {
      "entropy": 6.402029132843017,
      "epoch": 0.753606462781304,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004953889298578975,
      "loss": 6.3215,
      "mean_token_accuracy": 0.151659794151783,
      "num_tokens": 16527281.0,
      "step": 6530
    },
    {
      "entropy": 6.446967220306396,
      "epoch": 0.7541834968263128,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953806149640473,
      "loss": 6.3285,
      "mean_token_accuracy": 0.15949189513921738,
      "num_tokens": 16539763.0,
      "step": 6535
    },
    {
      "entropy": 6.376011753082276,
      "epoch": 0.7547605308713214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953722926578045,
      "loss": 6.3066,
      "mean_token_accuracy": 0.16653935611248016,
      "num_tokens": 16552197.0,
      "step": 6540
    },
    {
      "entropy": 6.45217113494873,
      "epoch": 0.7553375649163301,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004953639629394492,
      "loss": 6.3595,
      "mean_token_accuracy": 0.15816257297992706,
      "num_tokens": 16563988.0,
      "step": 6545
    },
    {
      "entropy": 6.346515941619873,
      "epoch": 0.7559145989613387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953556258092613,
      "loss": 6.354,
      "mean_token_accuracy": 0.152614526450634,
      "num_tokens": 16576982.0,
      "step": 6550
    },
    {
      "entropy": 6.500877237319946,
      "epoch": 0.7564916330063474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004953472812675216,
      "loss": 6.3918,
      "mean_token_accuracy": 0.15306739658117294,
      "num_tokens": 16590268.0,
      "step": 6555
    },
    {
      "entropy": 6.4272643566131595,
      "epoch": 0.757068667051356,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004953389293145104,
      "loss": 6.2367,
      "mean_token_accuracy": 0.16499679684638976,
      "num_tokens": 16602599.0,
      "step": 6560
    },
    {
      "entropy": 6.439320850372314,
      "epoch": 0.7576457010963646,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004953305699505088,
      "loss": 6.336,
      "mean_token_accuracy": 0.15458009764552116,
      "num_tokens": 16614278.0,
      "step": 6565
    },
    {
      "entropy": 6.435625171661377,
      "epoch": 0.7582227351413734,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953222031757979,
      "loss": 6.3623,
      "mean_token_accuracy": 0.15230781883001326,
      "num_tokens": 16626810.0,
      "step": 6570
    },
    {
      "entropy": 6.333204841613769,
      "epoch": 0.758799769186382,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953138289906592,
      "loss": 6.2958,
      "mean_token_accuracy": 0.15872435569763182,
      "num_tokens": 16639134.0,
      "step": 6575
    },
    {
      "entropy": 6.464828538894653,
      "epoch": 0.7593768032313907,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004953054473953742,
      "loss": 6.3366,
      "mean_token_accuracy": 0.15388649702072144,
      "num_tokens": 16652724.0,
      "step": 6580
    },
    {
      "entropy": 6.3356160640716555,
      "epoch": 0.7599538372763993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004952970583902251,
      "loss": 6.3039,
      "mean_token_accuracy": 0.15653215497732162,
      "num_tokens": 16666045.0,
      "step": 6585
    },
    {
      "entropy": 6.404564905166626,
      "epoch": 0.760530871321408,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004952886619754937,
      "loss": 6.4156,
      "mean_token_accuracy": 0.14934709966182708,
      "num_tokens": 16678422.0,
      "step": 6590
    },
    {
      "entropy": 6.406267356872559,
      "epoch": 0.7611079053664166,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004952802581514625,
      "loss": 6.3248,
      "mean_token_accuracy": 0.15902172029018402,
      "num_tokens": 16690243.0,
      "step": 6595
    },
    {
      "entropy": 6.393883800506591,
      "epoch": 0.7616849394114252,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004952718469184144,
      "loss": 6.303,
      "mean_token_accuracy": 0.16410621106624604,
      "num_tokens": 16702418.0,
      "step": 6600
    },
    {
      "entropy": 6.468479824066162,
      "epoch": 0.762261973456434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004952634282766322,
      "loss": 6.3958,
      "mean_token_accuracy": 0.15294192880392074,
      "num_tokens": 16713974.0,
      "step": 6605
    },
    {
      "entropy": 6.432626152038575,
      "epoch": 0.7628390075014426,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004952550022263988,
      "loss": 6.2699,
      "mean_token_accuracy": 0.1547722727060318,
      "num_tokens": 16726518.0,
      "step": 6610
    },
    {
      "entropy": 6.352759170532226,
      "epoch": 0.7634160415464513,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004952465687679979,
      "loss": 6.3028,
      "mean_token_accuracy": 0.1620583415031433,
      "num_tokens": 16738701.0,
      "step": 6615
    },
    {
      "entropy": 6.434079885482788,
      "epoch": 0.7639930755914599,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000495238127901713,
      "loss": 6.3445,
      "mean_token_accuracy": 0.14946923404932022,
      "num_tokens": 16751251.0,
      "step": 6620
    },
    {
      "entropy": 6.453442096710205,
      "epoch": 0.7645701096364685,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004952296796278282,
      "loss": 6.3561,
      "mean_token_accuracy": 0.15859053581953048,
      "num_tokens": 16764233.0,
      "step": 6625
    },
    {
      "entropy": 6.302823925018311,
      "epoch": 0.7651471436814772,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004952212239466274,
      "loss": 6.3387,
      "mean_token_accuracy": 0.16784807592630385,
      "num_tokens": 16778408.0,
      "step": 6630
    },
    {
      "entropy": 6.466341161727906,
      "epoch": 0.7657241777264858,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004952127608583953,
      "loss": 6.3576,
      "mean_token_accuracy": 0.1531306877732277,
      "num_tokens": 16789863.0,
      "step": 6635
    },
    {
      "entropy": 6.465719747543335,
      "epoch": 0.7663012117714946,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004952042903634162,
      "loss": 6.2966,
      "mean_token_accuracy": 0.16282862573862075,
      "num_tokens": 16804448.0,
      "step": 6640
    },
    {
      "entropy": 6.363318300247192,
      "epoch": 0.7668782458165032,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004951958124619752,
      "loss": 6.3713,
      "mean_token_accuracy": 0.15907371491193772,
      "num_tokens": 16817922.0,
      "step": 6645
    },
    {
      "entropy": 6.424112319946289,
      "epoch": 0.7674552798615119,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004951873271543573,
      "loss": 6.3752,
      "mean_token_accuracy": 0.16005789488554,
      "num_tokens": 16830322.0,
      "step": 6650
    },
    {
      "entropy": 6.532310724258423,
      "epoch": 0.7680323139065205,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004951788344408483,
      "loss": 6.3789,
      "mean_token_accuracy": 0.1579961434006691,
      "num_tokens": 16843489.0,
      "step": 6655
    },
    {
      "entropy": 6.482181167602539,
      "epoch": 0.7686093479515291,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004951703343217335,
      "loss": 6.3652,
      "mean_token_accuracy": 0.1490263119339943,
      "num_tokens": 16855256.0,
      "step": 6660
    },
    {
      "entropy": 6.410966539382935,
      "epoch": 0.7691863819965378,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004951618267972987,
      "loss": 6.3532,
      "mean_token_accuracy": 0.15816780030727387,
      "num_tokens": 16869927.0,
      "step": 6665
    },
    {
      "entropy": 6.371284675598145,
      "epoch": 0.7697634160415464,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004951533118678304,
      "loss": 6.2393,
      "mean_token_accuracy": 0.16636443585157396,
      "num_tokens": 16882095.0,
      "step": 6670
    },
    {
      "entropy": 6.5100563049316404,
      "epoch": 0.7703404500865552,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004951447895336147,
      "loss": 6.4317,
      "mean_token_accuracy": 0.15491466969251633,
      "num_tokens": 16893338.0,
      "step": 6675
    },
    {
      "entropy": 6.3982367515563965,
      "epoch": 0.7709174841315638,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004951362597949384,
      "loss": 6.3991,
      "mean_token_accuracy": 0.15819637775421141,
      "num_tokens": 16907138.0,
      "step": 6680
    },
    {
      "entropy": 6.452460479736328,
      "epoch": 0.7714945181765724,
      "grad_norm": 2.125,
      "learning_rate": 0.0004951277226520883,
      "loss": 6.3306,
      "mean_token_accuracy": 0.15423502773046494,
      "num_tokens": 16919756.0,
      "step": 6685
    },
    {
      "entropy": 6.4641800880432125,
      "epoch": 0.7720715522215811,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004951191781053517,
      "loss": 6.2621,
      "mean_token_accuracy": 0.16414030492305756,
      "num_tokens": 16932059.0,
      "step": 6690
    },
    {
      "entropy": 6.405013608932495,
      "epoch": 0.7726485862665897,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004951106261550157,
      "loss": 6.3545,
      "mean_token_accuracy": 0.15656148940324782,
      "num_tokens": 16944777.0,
      "step": 6695
    },
    {
      "entropy": 6.428842639923095,
      "epoch": 0.7732256203115984,
      "grad_norm": 1.9375,
      "learning_rate": 0.0004951020668013683,
      "loss": 6.3287,
      "mean_token_accuracy": 0.15988122075796127,
      "num_tokens": 16957111.0,
      "step": 6700
    },
    {
      "entropy": 6.433229970932007,
      "epoch": 0.773802654356607,
      "grad_norm": 2.0,
      "learning_rate": 0.0004950935000446972,
      "loss": 6.3975,
      "mean_token_accuracy": 0.14823095202445985,
      "num_tokens": 16968663.0,
      "step": 6705
    },
    {
      "entropy": 6.4144176006317135,
      "epoch": 0.7743796884016156,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004950849258852905,
      "loss": 6.3352,
      "mean_token_accuracy": 0.1517917647957802,
      "num_tokens": 16980668.0,
      "step": 6710
    },
    {
      "entropy": 6.46686315536499,
      "epoch": 0.7749567224466244,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004950763443234366,
      "loss": 6.362,
      "mean_token_accuracy": 0.16094649136066436,
      "num_tokens": 16993537.0,
      "step": 6715
    },
    {
      "entropy": 6.4711127281188965,
      "epoch": 0.775533756491633,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004950677553594243,
      "loss": 6.3256,
      "mean_token_accuracy": 0.15494734719395636,
      "num_tokens": 17006268.0,
      "step": 6720
    },
    {
      "entropy": 6.444059276580811,
      "epoch": 0.7761107905366417,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004950591589935422,
      "loss": 6.3484,
      "mean_token_accuracy": 0.15666031688451768,
      "num_tokens": 17018643.0,
      "step": 6725
    },
    {
      "entropy": 6.418145561218262,
      "epoch": 0.7766878245816503,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004950505552260798,
      "loss": 6.3034,
      "mean_token_accuracy": 0.16328714042901993,
      "num_tokens": 17030670.0,
      "step": 6730
    },
    {
      "entropy": 6.476660203933716,
      "epoch": 0.777264858626659,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004950419440573261,
      "loss": 6.3197,
      "mean_token_accuracy": 0.15912694185972215,
      "num_tokens": 17042617.0,
      "step": 6735
    },
    {
      "entropy": 6.406995248794556,
      "epoch": 0.7778418926716676,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000495033325487571,
      "loss": 6.3565,
      "mean_token_accuracy": 0.15629192739725112,
      "num_tokens": 17055217.0,
      "step": 6740
    },
    {
      "entropy": 6.397409391403198,
      "epoch": 0.7784189267166762,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004950246995171042,
      "loss": 6.2153,
      "mean_token_accuracy": 0.1760856807231903,
      "num_tokens": 17068614.0,
      "step": 6745
    },
    {
      "entropy": 6.403274583816528,
      "epoch": 0.778995960761685,
      "grad_norm": 1.5625,
      "learning_rate": 0.000495016066146216,
      "loss": 6.2584,
      "mean_token_accuracy": 0.1673291563987732,
      "num_tokens": 17079892.0,
      "step": 6750
    },
    {
      "entropy": 6.386370849609375,
      "epoch": 0.7795729948066936,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004950074253751968,
      "loss": 6.351,
      "mean_token_accuracy": 0.15970607101917267,
      "num_tokens": 17092113.0,
      "step": 6755
    },
    {
      "entropy": 6.390835618972778,
      "epoch": 0.7801500288517023,
      "grad_norm": 1.296875,
      "learning_rate": 0.000494998777204337,
      "loss": 6.3212,
      "mean_token_accuracy": 0.1628754422068596,
      "num_tokens": 17103387.0,
      "step": 6760
    },
    {
      "entropy": 6.480059051513672,
      "epoch": 0.7807270628967109,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004949901216339276,
      "loss": 6.3738,
      "mean_token_accuracy": 0.15862331688404083,
      "num_tokens": 17116008.0,
      "step": 6765
    },
    {
      "entropy": 6.376413249969483,
      "epoch": 0.7813040969417195,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004949814586642598,
      "loss": 6.2674,
      "mean_token_accuracy": 0.16962072998285294,
      "num_tokens": 17129224.0,
      "step": 6770
    },
    {
      "entropy": 6.316375637054444,
      "epoch": 0.7818811309867282,
      "grad_norm": 1.453125,
      "learning_rate": 0.000494972788295625,
      "loss": 6.233,
      "mean_token_accuracy": 0.15894538313150405,
      "num_tokens": 17141357.0,
      "step": 6775
    },
    {
      "entropy": 6.3543017387390135,
      "epoch": 0.7824581650317368,
      "grad_norm": 1.25,
      "learning_rate": 0.0004949641105283144,
      "loss": 6.2748,
      "mean_token_accuracy": 0.15785481184720992,
      "num_tokens": 17153672.0,
      "step": 6780
    },
    {
      "entropy": 6.419472122192383,
      "epoch": 0.7830351990767456,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004949554253626205,
      "loss": 6.3656,
      "mean_token_accuracy": 0.1508081555366516,
      "num_tokens": 17168429.0,
      "step": 6785
    },
    {
      "entropy": 6.4885454177856445,
      "epoch": 0.7836122331217542,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004949467327988351,
      "loss": 6.3155,
      "mean_token_accuracy": 0.15907162874937059,
      "num_tokens": 17181000.0,
      "step": 6790
    },
    {
      "entropy": 6.396276998519897,
      "epoch": 0.7841892671667628,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949380328372505,
      "loss": 6.3156,
      "mean_token_accuracy": 0.15487978011369705,
      "num_tokens": 17193984.0,
      "step": 6795
    },
    {
      "entropy": 6.482124853134155,
      "epoch": 0.7847663012117715,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004949293254781595,
      "loss": 6.327,
      "mean_token_accuracy": 0.15890616029500962,
      "num_tokens": 17207289.0,
      "step": 6800
    },
    {
      "entropy": 6.413140916824341,
      "epoch": 0.7853433352567801,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004949206107218547,
      "loss": 6.3719,
      "mean_token_accuracy": 0.15868546366691588,
      "num_tokens": 17220785.0,
      "step": 6805
    },
    {
      "entropy": 6.459430408477783,
      "epoch": 0.7859203693017888,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004949118885686296,
      "loss": 6.3403,
      "mean_token_accuracy": 0.1554628610610962,
      "num_tokens": 17233475.0,
      "step": 6810
    },
    {
      "entropy": 6.440314626693725,
      "epoch": 0.7864974033467974,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004949031590187774,
      "loss": 6.331,
      "mean_token_accuracy": 0.1579444646835327,
      "num_tokens": 17246652.0,
      "step": 6815
    },
    {
      "entropy": 6.4201226234436035,
      "epoch": 0.7870744373918062,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004948944220725915,
      "loss": 6.3246,
      "mean_token_accuracy": 0.158696049451828,
      "num_tokens": 17259966.0,
      "step": 6820
    },
    {
      "entropy": 6.3556853294372555,
      "epoch": 0.7876514714368148,
      "grad_norm": 1.328125,
      "learning_rate": 0.000494885677730366,
      "loss": 6.2917,
      "mean_token_accuracy": 0.16671997755765916,
      "num_tokens": 17272556.0,
      "step": 6825
    },
    {
      "entropy": 6.431224250793457,
      "epoch": 0.7882285054818234,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000494876925992395,
      "loss": 6.2959,
      "mean_token_accuracy": 0.15880678445100785,
      "num_tokens": 17285960.0,
      "step": 6830
    },
    {
      "entropy": 6.45187816619873,
      "epoch": 0.7888055395268321,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004948681668589728,
      "loss": 6.4111,
      "mean_token_accuracy": 0.14944198802113534,
      "num_tokens": 17299743.0,
      "step": 6835
    },
    {
      "entropy": 6.4086791515350345,
      "epoch": 0.7893825735718407,
      "grad_norm": 1.234375,
      "learning_rate": 0.000494859400330394,
      "loss": 6.2756,
      "mean_token_accuracy": 0.15799426585435866,
      "num_tokens": 17312317.0,
      "step": 6840
    },
    {
      "entropy": 6.4271081447601315,
      "epoch": 0.7899596076168494,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004948506264069535,
      "loss": 6.3429,
      "mean_token_accuracy": 0.15904499739408492,
      "num_tokens": 17326204.0,
      "step": 6845
    },
    {
      "entropy": 6.375916767120361,
      "epoch": 0.790536641661858,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004948418450889464,
      "loss": 6.2957,
      "mean_token_accuracy": 0.1635165348649025,
      "num_tokens": 17338362.0,
      "step": 6850
    },
    {
      "entropy": 6.390348434448242,
      "epoch": 0.7911136757068667,
      "grad_norm": 1.203125,
      "learning_rate": 0.000494833056376668,
      "loss": 6.3198,
      "mean_token_accuracy": 0.16242424845695497,
      "num_tokens": 17350713.0,
      "step": 6855
    },
    {
      "entropy": 6.476944398880005,
      "epoch": 0.7916907097518754,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004948242602704139,
      "loss": 6.3484,
      "mean_token_accuracy": 0.14923545345664024,
      "num_tokens": 17364055.0,
      "step": 6860
    },
    {
      "entropy": 6.442521095275879,
      "epoch": 0.792267743796884,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004948154567704801,
      "loss": 6.4049,
      "mean_token_accuracy": 0.161513988673687,
      "num_tokens": 17377940.0,
      "step": 6865
    },
    {
      "entropy": 6.425199937820435,
      "epoch": 0.7928447778418927,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004948066458771625,
      "loss": 6.3176,
      "mean_token_accuracy": 0.15831714868545532,
      "num_tokens": 17390348.0,
      "step": 6870
    },
    {
      "entropy": 6.433390998840332,
      "epoch": 0.7934218118869013,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004947978275907577,
      "loss": 6.319,
      "mean_token_accuracy": 0.15604161992669105,
      "num_tokens": 17403406.0,
      "step": 6875
    },
    {
      "entropy": 6.44692759513855,
      "epoch": 0.79399884593191,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000494789001911562,
      "loss": 6.3229,
      "mean_token_accuracy": 0.15025489330291747,
      "num_tokens": 17415306.0,
      "step": 6880
    },
    {
      "entropy": 6.500966596603393,
      "epoch": 0.7945758799769186,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004947801688398725,
      "loss": 6.3779,
      "mean_token_accuracy": 0.15381309092044831,
      "num_tokens": 17427901.0,
      "step": 6885
    },
    {
      "entropy": 6.417627954483033,
      "epoch": 0.7951529140219273,
      "grad_norm": 2.0,
      "learning_rate": 0.0004947713283759862,
      "loss": 6.3172,
      "mean_token_accuracy": 0.15818158090114592,
      "num_tokens": 17439812.0,
      "step": 6890
    },
    {
      "entropy": 6.404314374923706,
      "epoch": 0.795729948066936,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004947624805202003,
      "loss": 6.432,
      "mean_token_accuracy": 0.15916587486863137,
      "num_tokens": 17453317.0,
      "step": 6895
    },
    {
      "entropy": 6.379020404815674,
      "epoch": 0.7963069821119446,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004947536252728126,
      "loss": 6.3185,
      "mean_token_accuracy": 0.15128010660409927,
      "num_tokens": 17465563.0,
      "step": 6900
    },
    {
      "entropy": 6.4297808647155765,
      "epoch": 0.7968840161569533,
      "grad_norm": 1.625,
      "learning_rate": 0.0004947447626341209,
      "loss": 6.366,
      "mean_token_accuracy": 0.15720113217830659,
      "num_tokens": 17477672.0,
      "step": 6905
    },
    {
      "entropy": 6.426624536514282,
      "epoch": 0.7974610502019619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004947358926044232,
      "loss": 6.3237,
      "mean_token_accuracy": 0.15934155583381654,
      "num_tokens": 17489061.0,
      "step": 6910
    },
    {
      "entropy": 6.392905282974243,
      "epoch": 0.7980380842469705,
      "grad_norm": 1.625,
      "learning_rate": 0.0004947270151840179,
      "loss": 6.371,
      "mean_token_accuracy": 0.15421160608530043,
      "num_tokens": 17501734.0,
      "step": 6915
    },
    {
      "entropy": 6.442747926712036,
      "epoch": 0.7986151182919792,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004947181303732038,
      "loss": 6.2726,
      "mean_token_accuracy": 0.15750008672475815,
      "num_tokens": 17514387.0,
      "step": 6920
    },
    {
      "entropy": 6.403343772888183,
      "epoch": 0.7991921523369879,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004947092381722793,
      "loss": 6.2788,
      "mean_token_accuracy": 0.1647616818547249,
      "num_tokens": 17526483.0,
      "step": 6925
    },
    {
      "entropy": 6.44573745727539,
      "epoch": 0.7997691863819966,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004947003385815438,
      "loss": 6.307,
      "mean_token_accuracy": 0.16297112554311752,
      "num_tokens": 17539809.0,
      "step": 6930
    },
    {
      "entropy": 6.416499471664428,
      "epoch": 0.8003462204270052,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004946914316012964,
      "loss": 6.3167,
      "mean_token_accuracy": 0.15640088468790053,
      "num_tokens": 17552973.0,
      "step": 6935
    },
    {
      "entropy": 6.429599618911743,
      "epoch": 0.8009232544720138,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000494682517231837,
      "loss": 6.3671,
      "mean_token_accuracy": 0.16080797612667083,
      "num_tokens": 17567803.0,
      "step": 6940
    },
    {
      "entropy": 6.481288766860962,
      "epoch": 0.8015002885170225,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004946735954734652,
      "loss": 6.3257,
      "mean_token_accuracy": 0.156865856051445,
      "num_tokens": 17579718.0,
      "step": 6945
    },
    {
      "entropy": 6.400873327255249,
      "epoch": 0.8020773225620311,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004946646663264811,
      "loss": 6.3161,
      "mean_token_accuracy": 0.15859004408121108,
      "num_tokens": 17592384.0,
      "step": 6950
    },
    {
      "entropy": 6.439334154129028,
      "epoch": 0.8026543566070398,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004946557297911852,
      "loss": 6.225,
      "mean_token_accuracy": 0.16009974032640456,
      "num_tokens": 17604292.0,
      "step": 6955
    },
    {
      "entropy": 6.35125412940979,
      "epoch": 0.8032313906520485,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004946467858678777,
      "loss": 6.2405,
      "mean_token_accuracy": 0.16352516561746597,
      "num_tokens": 17616638.0,
      "step": 6960
    },
    {
      "entropy": 6.439220523834228,
      "epoch": 0.8038084246970572,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004946378345568597,
      "loss": 6.3572,
      "mean_token_accuracy": 0.14690545722842216,
      "num_tokens": 17630672.0,
      "step": 6965
    },
    {
      "entropy": 6.457352066040039,
      "epoch": 0.8043854587420658,
      "grad_norm": 3.109375,
      "learning_rate": 0.0004946288758584324,
      "loss": 6.3248,
      "mean_token_accuracy": 0.15353938937187195,
      "num_tokens": 17644215.0,
      "step": 6970
    },
    {
      "entropy": 6.430181646347046,
      "epoch": 0.8049624927870744,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004946199097728968,
      "loss": 6.4047,
      "mean_token_accuracy": 0.15424128770828247,
      "num_tokens": 17655982.0,
      "step": 6975
    },
    {
      "entropy": 6.427625846862793,
      "epoch": 0.8055395268320831,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004946109363005548,
      "loss": 6.3272,
      "mean_token_accuracy": 0.15317181050777434,
      "num_tokens": 17668059.0,
      "step": 6980
    },
    {
      "entropy": 6.449534893035889,
      "epoch": 0.8061165608770917,
      "grad_norm": 2.078125,
      "learning_rate": 0.000494601955441708,
      "loss": 6.3512,
      "mean_token_accuracy": 0.1613871321082115,
      "num_tokens": 17680924.0,
      "step": 6985
    },
    {
      "entropy": 6.417742967605591,
      "epoch": 0.8066935949221004,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004945929671966585,
      "loss": 6.3544,
      "mean_token_accuracy": 0.15113742798566818,
      "num_tokens": 17693381.0,
      "step": 6990
    },
    {
      "entropy": 6.323920249938965,
      "epoch": 0.8072706289671091,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004945839715657085,
      "loss": 6.2573,
      "mean_token_accuracy": 0.1607775092124939,
      "num_tokens": 17705900.0,
      "step": 6995
    },
    {
      "entropy": 6.417225408554077,
      "epoch": 0.8078476630121177,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004945749685491607,
      "loss": 6.2688,
      "mean_token_accuracy": 0.16660031527280808,
      "num_tokens": 17719139.0,
      "step": 7000
    },
    {
      "entropy": 6.399656534194946,
      "epoch": 0.8084246970571264,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004945659581473181,
      "loss": 6.2703,
      "mean_token_accuracy": 0.1602562978863716,
      "num_tokens": 17734312.0,
      "step": 7005
    },
    {
      "entropy": 6.474909067153931,
      "epoch": 0.809001731102135,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004945569403604833,
      "loss": 6.3247,
      "mean_token_accuracy": 0.1548770174384117,
      "num_tokens": 17746137.0,
      "step": 7010
    },
    {
      "entropy": 6.39998688697815,
      "epoch": 0.8095787651471437,
      "grad_norm": 1.75,
      "learning_rate": 0.0004945479151889601,
      "loss": 6.1753,
      "mean_token_accuracy": 0.1639669805765152,
      "num_tokens": 17758879.0,
      "step": 7015
    },
    {
      "entropy": 6.332425308227539,
      "epoch": 0.8101557991921523,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004945388826330517,
      "loss": 6.3147,
      "mean_token_accuracy": 0.16598912477493286,
      "num_tokens": 17770820.0,
      "step": 7020
    },
    {
      "entropy": 6.469837856292725,
      "epoch": 0.8107328332371609,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004945298426930621,
      "loss": 6.3626,
      "mean_token_accuracy": 0.15875112563371657,
      "num_tokens": 17783495.0,
      "step": 7025
    },
    {
      "entropy": 6.473739576339722,
      "epoch": 0.8113098672821697,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004945207953692952,
      "loss": 6.3452,
      "mean_token_accuracy": 0.15288871824741362,
      "num_tokens": 17795780.0,
      "step": 7030
    },
    {
      "entropy": 6.462626647949219,
      "epoch": 0.8118869013271783,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004945117406620556,
      "loss": 6.3471,
      "mean_token_accuracy": 0.15740556567907332,
      "num_tokens": 17809684.0,
      "step": 7035
    },
    {
      "entropy": 6.404884290695191,
      "epoch": 0.812463935372187,
      "grad_norm": 1.5,
      "learning_rate": 0.0004945026785716474,
      "loss": 6.3465,
      "mean_token_accuracy": 0.15713463425636293,
      "num_tokens": 17822719.0,
      "step": 7040
    },
    {
      "entropy": 6.485169458389282,
      "epoch": 0.8130409694171956,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004944936090983757,
      "loss": 6.3756,
      "mean_token_accuracy": 0.15452174842357635,
      "num_tokens": 17834883.0,
      "step": 7045
    },
    {
      "entropy": 6.423402452468872,
      "epoch": 0.8136180034622043,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004944845322425455,
      "loss": 6.3131,
      "mean_token_accuracy": 0.16450032442808152,
      "num_tokens": 17847009.0,
      "step": 7050
    },
    {
      "entropy": 6.4183858871459964,
      "epoch": 0.8141950375072129,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004944754480044621,
      "loss": 6.3147,
      "mean_token_accuracy": 0.15603942126035691,
      "num_tokens": 17859106.0,
      "step": 7055
    },
    {
      "entropy": 6.444196557998657,
      "epoch": 0.8147720715522215,
      "grad_norm": 2.0,
      "learning_rate": 0.0004944663563844311,
      "loss": 6.3201,
      "mean_token_accuracy": 0.15441031754016876,
      "num_tokens": 17871418.0,
      "step": 7060
    },
    {
      "entropy": 6.40857720375061,
      "epoch": 0.8153491055972303,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004944572573827581,
      "loss": 6.2908,
      "mean_token_accuracy": 0.1583213210105896,
      "num_tokens": 17884219.0,
      "step": 7065
    },
    {
      "entropy": 6.3658651351928714,
      "epoch": 0.8159261396422389,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004944481509997494,
      "loss": 6.3028,
      "mean_token_accuracy": 0.15765465348958968,
      "num_tokens": 17896206.0,
      "step": 7070
    },
    {
      "entropy": 6.43848180770874,
      "epoch": 0.8165031736872476,
      "grad_norm": 1.765625,
      "learning_rate": 0.000494439037235711,
      "loss": 6.3986,
      "mean_token_accuracy": 0.15387010723352432,
      "num_tokens": 17909488.0,
      "step": 7075
    },
    {
      "entropy": 6.43320894241333,
      "epoch": 0.8170802077322562,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004944299160909495,
      "loss": 6.346,
      "mean_token_accuracy": 0.15420840233564376,
      "num_tokens": 17922851.0,
      "step": 7080
    },
    {
      "entropy": 6.46990966796875,
      "epoch": 0.8176572417772648,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000494420787565772,
      "loss": 6.2682,
      "mean_token_accuracy": 0.16781508177518845,
      "num_tokens": 17934586.0,
      "step": 7085
    },
    {
      "entropy": 6.410624599456787,
      "epoch": 0.8182342758222735,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004944116516604852,
      "loss": 6.3653,
      "mean_token_accuracy": 0.15871597677469254,
      "num_tokens": 17947637.0,
      "step": 7090
    },
    {
      "entropy": 6.419895935058594,
      "epoch": 0.8188113098672821,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004944025083753965,
      "loss": 6.3346,
      "mean_token_accuracy": 0.15860212296247483,
      "num_tokens": 17959601.0,
      "step": 7095
    },
    {
      "entropy": 6.42613205909729,
      "epoch": 0.8193883439122909,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004943933577108133,
      "loss": 6.3047,
      "mean_token_accuracy": 0.1616590827703476,
      "num_tokens": 17972568.0,
      "step": 7100
    },
    {
      "entropy": 6.461557245254516,
      "epoch": 0.8199653779572995,
      "grad_norm": 1.25,
      "learning_rate": 0.0004943841996670436,
      "loss": 6.3088,
      "mean_token_accuracy": 0.1592638984322548,
      "num_tokens": 17984665.0,
      "step": 7105
    },
    {
      "entropy": 6.465728902816773,
      "epoch": 0.8205424120023082,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004943750342443953,
      "loss": 6.3551,
      "mean_token_accuracy": 0.1517217919230461,
      "num_tokens": 17997140.0,
      "step": 7110
    },
    {
      "entropy": 6.459466886520386,
      "epoch": 0.8211194460473168,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004943658614431767,
      "loss": 6.3335,
      "mean_token_accuracy": 0.15835410803556443,
      "num_tokens": 18009361.0,
      "step": 7115
    },
    {
      "entropy": 6.377575302124024,
      "epoch": 0.8216964800923254,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004943566812636963,
      "loss": 6.2799,
      "mean_token_accuracy": 0.15527833849191666,
      "num_tokens": 18021233.0,
      "step": 7120
    },
    {
      "entropy": 6.3838001728057865,
      "epoch": 0.8222735141373341,
      "grad_norm": 1.953125,
      "learning_rate": 0.000494347493706263,
      "loss": 6.1757,
      "mean_token_accuracy": 0.1724259674549103,
      "num_tokens": 18034086.0,
      "step": 7125
    },
    {
      "entropy": 6.393672943115234,
      "epoch": 0.8228505481823427,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004943382987711856,
      "loss": 6.2819,
      "mean_token_accuracy": 0.16438249498605728,
      "num_tokens": 18046604.0,
      "step": 7130
    },
    {
      "entropy": 6.414757633209229,
      "epoch": 0.8234275822273515,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004943290964587734,
      "loss": 6.3044,
      "mean_token_accuracy": 0.16008765250444412,
      "num_tokens": 18059185.0,
      "step": 7135
    },
    {
      "entropy": 6.397540616989136,
      "epoch": 0.8240046162723601,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004943198867693361,
      "loss": 6.2537,
      "mean_token_accuracy": 0.16008234918117523,
      "num_tokens": 18071707.0,
      "step": 7140
    },
    {
      "entropy": 6.433577537536621,
      "epoch": 0.8245816503173687,
      "grad_norm": 4.8125,
      "learning_rate": 0.0004943106697031833,
      "loss": 6.4092,
      "mean_token_accuracy": 0.15259905457496642,
      "num_tokens": 18084657.0,
      "step": 7145
    },
    {
      "entropy": 6.472084045410156,
      "epoch": 0.8251586843623774,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004943014452606251,
      "loss": 6.3736,
      "mean_token_accuracy": 0.1545234352350235,
      "num_tokens": 18097084.0,
      "step": 7150
    },
    {
      "entropy": 6.420034742355346,
      "epoch": 0.825735718407386,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004942922134419717,
      "loss": 6.283,
      "mean_token_accuracy": 0.15995100438594817,
      "num_tokens": 18109306.0,
      "step": 7155
    },
    {
      "entropy": 6.360563182830811,
      "epoch": 0.8263127524523947,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004942829742475336,
      "loss": 6.1958,
      "mean_token_accuracy": 0.15680315494537353,
      "num_tokens": 18120300.0,
      "step": 7160
    },
    {
      "entropy": 6.409770345687866,
      "epoch": 0.8268897864974033,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004942737276776217,
      "loss": 6.3318,
      "mean_token_accuracy": 0.1568691462278366,
      "num_tokens": 18132251.0,
      "step": 7165
    },
    {
      "entropy": 6.3930370807647705,
      "epoch": 0.827466820542412,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004942644737325468,
      "loss": 6.2378,
      "mean_token_accuracy": 0.16396623849868774,
      "num_tokens": 18146198.0,
      "step": 7170
    },
    {
      "entropy": 6.412364673614502,
      "epoch": 0.8280438545874207,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004942552124126202,
      "loss": 6.3532,
      "mean_token_accuracy": 0.15747751146554947,
      "num_tokens": 18159493.0,
      "step": 7175
    },
    {
      "entropy": 6.372334575653076,
      "epoch": 0.8286208886324293,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004942459437181535,
      "loss": 6.2454,
      "mean_token_accuracy": 0.16117880791425704,
      "num_tokens": 18172120.0,
      "step": 7180
    },
    {
      "entropy": 6.417788648605347,
      "epoch": 0.829197922677438,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004942366676494584,
      "loss": 6.2758,
      "mean_token_accuracy": 0.15891496539115907,
      "num_tokens": 18184704.0,
      "step": 7185
    },
    {
      "entropy": 6.5040661811828615,
      "epoch": 0.8297749567224466,
      "grad_norm": 1.375,
      "learning_rate": 0.0004942273842068469,
      "loss": 6.378,
      "mean_token_accuracy": 0.1573457822203636,
      "num_tokens": 18196770.0,
      "step": 7190
    },
    {
      "entropy": 6.42578158378601,
      "epoch": 0.8303519907674553,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004942180933906311,
      "loss": 6.3353,
      "mean_token_accuracy": 0.1560923770070076,
      "num_tokens": 18209987.0,
      "step": 7195
    },
    {
      "entropy": 6.427188396453857,
      "epoch": 0.8309290248124639,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004942087952011237,
      "loss": 6.3619,
      "mean_token_accuracy": 0.15685783326625824,
      "num_tokens": 18224864.0,
      "step": 7200
    },
    {
      "entropy": 6.477103185653687,
      "epoch": 0.8315060588574726,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004941994896386374,
      "loss": 6.3183,
      "mean_token_accuracy": 0.1615700140595436,
      "num_tokens": 18237270.0,
      "step": 7205
    },
    {
      "entropy": 6.358955335617066,
      "epoch": 0.8320830929024813,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004941901767034851,
      "loss": 6.3402,
      "mean_token_accuracy": 0.15311284065246583,
      "num_tokens": 18251227.0,
      "step": 7210
    },
    {
      "entropy": 6.375926876068116,
      "epoch": 0.8326601269474899,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004941808563959802,
      "loss": 6.2519,
      "mean_token_accuracy": 0.1678831934928894,
      "num_tokens": 18264762.0,
      "step": 7215
    },
    {
      "entropy": 6.457195186614991,
      "epoch": 0.8332371609924986,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004941715287164359,
      "loss": 6.2962,
      "mean_token_accuracy": 0.15596022307872773,
      "num_tokens": 18276836.0,
      "step": 7220
    },
    {
      "entropy": 6.384139442443848,
      "epoch": 0.8338141950375072,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004941621936651661,
      "loss": 6.3289,
      "mean_token_accuracy": 0.16242541372776031,
      "num_tokens": 18290362.0,
      "step": 7225
    },
    {
      "entropy": 6.4580583572387695,
      "epoch": 0.8343912290825158,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004941528512424849,
      "loss": 6.3579,
      "mean_token_accuracy": 0.1578863427042961,
      "num_tokens": 18303924.0,
      "step": 7230
    },
    {
      "entropy": 6.370277833938599,
      "epoch": 0.8349682631275245,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004941435014487064,
      "loss": 6.2306,
      "mean_token_accuracy": 0.16379570066928864,
      "num_tokens": 18316490.0,
      "step": 7235
    },
    {
      "entropy": 6.417813491821289,
      "epoch": 0.8355452971725332,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000494134144284145,
      "loss": 6.2853,
      "mean_token_accuracy": 0.15361817181110382,
      "num_tokens": 18328819.0,
      "step": 7240
    },
    {
      "entropy": 6.363092947006225,
      "epoch": 0.8361223312175419,
      "grad_norm": 1.375,
      "learning_rate": 0.0004941247797491156,
      "loss": 6.2611,
      "mean_token_accuracy": 0.16033381819725037,
      "num_tokens": 18341872.0,
      "step": 7245
    },
    {
      "entropy": 6.436460399627686,
      "epoch": 0.8366993652625505,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004941154078439329,
      "loss": 6.2839,
      "mean_token_accuracy": 0.15880754142999648,
      "num_tokens": 18354644.0,
      "step": 7250
    },
    {
      "entropy": 6.445900964736938,
      "epoch": 0.8372763993075591,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004941060285689126,
      "loss": 6.3164,
      "mean_token_accuracy": 0.1531373918056488,
      "num_tokens": 18367581.0,
      "step": 7255
    },
    {
      "entropy": 6.410998773574829,
      "epoch": 0.8378534333525678,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004940966419243695,
      "loss": 6.2879,
      "mean_token_accuracy": 0.16572379171848298,
      "num_tokens": 18381123.0,
      "step": 7260
    },
    {
      "entropy": 6.344087362289429,
      "epoch": 0.8384304673975764,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049408724791062,
      "loss": 6.2591,
      "mean_token_accuracy": 0.1606621190905571,
      "num_tokens": 18395711.0,
      "step": 7265
    },
    {
      "entropy": 6.464009857177734,
      "epoch": 0.8390075014425851,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004940778465279797,
      "loss": 6.3763,
      "mean_token_accuracy": 0.14919317811727523,
      "num_tokens": 18409434.0,
      "step": 7270
    },
    {
      "entropy": 6.4164632797241214,
      "epoch": 0.8395845354875938,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004940684377767647,
      "loss": 6.3223,
      "mean_token_accuracy": 0.15381991267204284,
      "num_tokens": 18423138.0,
      "step": 7275
    },
    {
      "entropy": 6.401902055740356,
      "epoch": 0.8401615695326025,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004940590216572916,
      "loss": 6.3188,
      "mean_token_accuracy": 0.15444015413522721,
      "num_tokens": 18435655.0,
      "step": 7280
    },
    {
      "entropy": 6.337902498245239,
      "epoch": 0.8407386035776111,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004940495981698772,
      "loss": 6.197,
      "mean_token_accuracy": 0.16857851296663284,
      "num_tokens": 18448344.0,
      "step": 7285
    },
    {
      "entropy": 6.393205165863037,
      "epoch": 0.8413156376226197,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004940401673148384,
      "loss": 6.2057,
      "mean_token_accuracy": 0.16652216762304306,
      "num_tokens": 18460409.0,
      "step": 7290
    },
    {
      "entropy": 6.415207719802856,
      "epoch": 0.8418926716676284,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004940307290924923,
      "loss": 6.3247,
      "mean_token_accuracy": 0.15722924172878266,
      "num_tokens": 18473650.0,
      "step": 7295
    },
    {
      "entropy": 6.35351300239563,
      "epoch": 0.842469705712637,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004940212835031565,
      "loss": 6.2472,
      "mean_token_accuracy": 0.16438708156347276,
      "num_tokens": 18487108.0,
      "step": 7300
    },
    {
      "entropy": 6.432998180389404,
      "epoch": 0.8430467397576457,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004940118305471486,
      "loss": 6.2857,
      "mean_token_accuracy": 0.15940231084823608,
      "num_tokens": 18500377.0,
      "step": 7305
    },
    {
      "entropy": 6.433759593963623,
      "epoch": 0.8436237738026544,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004940023702247866,
      "loss": 6.3466,
      "mean_token_accuracy": 0.1481998160481453,
      "num_tokens": 18513257.0,
      "step": 7310
    },
    {
      "entropy": 6.464117813110351,
      "epoch": 0.844200807847663,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004939929025363886,
      "loss": 6.3405,
      "mean_token_accuracy": 0.15689841657876968,
      "num_tokens": 18525438.0,
      "step": 7315
    },
    {
      "entropy": 6.4316305160522464,
      "epoch": 0.8447778418926717,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004939834274822731,
      "loss": 6.3755,
      "mean_token_accuracy": 0.16183712780475618,
      "num_tokens": 18537937.0,
      "step": 7320
    },
    {
      "entropy": 6.469800138473511,
      "epoch": 0.8453548759376803,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004939739450627588,
      "loss": 6.3821,
      "mean_token_accuracy": 0.15115718245506288,
      "num_tokens": 18551629.0,
      "step": 7325
    },
    {
      "entropy": 6.4402059555053714,
      "epoch": 0.845931909982689,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004939644552781647,
      "loss": 6.3297,
      "mean_token_accuracy": 0.15563774853944778,
      "num_tokens": 18564549.0,
      "step": 7330
    },
    {
      "entropy": 6.371536636352539,
      "epoch": 0.8465089440276976,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004939549581288099,
      "loss": 6.3173,
      "mean_token_accuracy": 0.16682110875844955,
      "num_tokens": 18576536.0,
      "step": 7335
    },
    {
      "entropy": 6.324753952026367,
      "epoch": 0.8470859780727062,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004939454536150138,
      "loss": 6.2766,
      "mean_token_accuracy": 0.16521506160497665,
      "num_tokens": 18588655.0,
      "step": 7340
    },
    {
      "entropy": 6.4212664604187015,
      "epoch": 0.847663012117715,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000493935941737096,
      "loss": 6.251,
      "mean_token_accuracy": 0.16551081836223602,
      "num_tokens": 18602152.0,
      "step": 7345
    },
    {
      "entropy": 6.347581577301026,
      "epoch": 0.8482400461627236,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004939264224953768,
      "loss": 6.2585,
      "mean_token_accuracy": 0.1667644664645195,
      "num_tokens": 18615391.0,
      "step": 7350
    },
    {
      "entropy": 6.302462100982666,
      "epoch": 0.8488170802077323,
      "grad_norm": 1.53125,
      "learning_rate": 0.000493916895890176,
      "loss": 6.3328,
      "mean_token_accuracy": 0.15009007528424262,
      "num_tokens": 18629987.0,
      "step": 7355
    },
    {
      "entropy": 6.5011743068695065,
      "epoch": 0.8493941142527409,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000493907361921814,
      "loss": 6.38,
      "mean_token_accuracy": 0.15558104068040848,
      "num_tokens": 18643064.0,
      "step": 7360
    },
    {
      "entropy": 6.517504692077637,
      "epoch": 0.8499711482977496,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004938978205906118,
      "loss": 6.3399,
      "mean_token_accuracy": 0.15750785619020463,
      "num_tokens": 18654950.0,
      "step": 7365
    },
    {
      "entropy": 6.375610494613648,
      "epoch": 0.8505481823427582,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004938882718968901,
      "loss": 6.2809,
      "mean_token_accuracy": 0.15807467252016066,
      "num_tokens": 18666318.0,
      "step": 7370
    },
    {
      "entropy": 6.356922674179077,
      "epoch": 0.8511252163877668,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004938787158409702,
      "loss": 6.242,
      "mean_token_accuracy": 0.16371146887540816,
      "num_tokens": 18677357.0,
      "step": 7375
    },
    {
      "entropy": 6.3947381496429445,
      "epoch": 0.8517022504327756,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004938691524231733,
      "loss": 6.2723,
      "mean_token_accuracy": 0.16164294630289078,
      "num_tokens": 18689563.0,
      "step": 7380
    },
    {
      "entropy": 6.312204122543335,
      "epoch": 0.8522792844777842,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004938595816438212,
      "loss": 6.2794,
      "mean_token_accuracy": 0.1561274692416191,
      "num_tokens": 18701442.0,
      "step": 7385
    },
    {
      "entropy": 6.441098594665528,
      "epoch": 0.8528563185227929,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004938500035032358,
      "loss": 6.335,
      "mean_token_accuracy": 0.15520687401294708,
      "num_tokens": 18714942.0,
      "step": 7390
    },
    {
      "entropy": 6.293469429016113,
      "epoch": 0.8534333525678015,
      "grad_norm": 3.25,
      "learning_rate": 0.0004938404180017392,
      "loss": 6.3419,
      "mean_token_accuracy": 0.16100662648677827,
      "num_tokens": 18727752.0,
      "step": 7395
    },
    {
      "entropy": 6.428264188766479,
      "epoch": 0.8540103866128101,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004938308251396539,
      "loss": 6.3428,
      "mean_token_accuracy": 0.1571537360548973,
      "num_tokens": 18741719.0,
      "step": 7400
    },
    {
      "entropy": 6.4747106552124025,
      "epoch": 0.8545874206578188,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004938212249173024,
      "loss": 6.2866,
      "mean_token_accuracy": 0.14972534030675888,
      "num_tokens": 18754472.0,
      "step": 7405
    },
    {
      "entropy": 6.4440333366394045,
      "epoch": 0.8551644547028274,
      "grad_norm": 1.875,
      "learning_rate": 0.0004938116173350078,
      "loss": 6.3288,
      "mean_token_accuracy": 0.15726516842842103,
      "num_tokens": 18766691.0,
      "step": 7410
    },
    {
      "entropy": 6.412831592559814,
      "epoch": 0.8557414887478362,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004938020023930932,
      "loss": 6.2983,
      "mean_token_accuracy": 0.15810909271240234,
      "num_tokens": 18780083.0,
      "step": 7415
    },
    {
      "entropy": 6.457475090026856,
      "epoch": 0.8563185227928448,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004937923800918817,
      "loss": 6.359,
      "mean_token_accuracy": 0.15542669594287872,
      "num_tokens": 18792559.0,
      "step": 7420
    },
    {
      "entropy": 6.3452540874481205,
      "epoch": 0.8568955568378535,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004937827504316974,
      "loss": 6.2176,
      "mean_token_accuracy": 0.16970468163490296,
      "num_tokens": 18805089.0,
      "step": 7425
    },
    {
      "entropy": 6.240821170806885,
      "epoch": 0.8574725908828621,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004937731134128639,
      "loss": 6.2161,
      "mean_token_accuracy": 0.15851088762283325,
      "num_tokens": 18818476.0,
      "step": 7430
    },
    {
      "entropy": 6.44542875289917,
      "epoch": 0.8580496249278707,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004937634690357054,
      "loss": 6.4014,
      "mean_token_accuracy": 0.15640219449996948,
      "num_tokens": 18830785.0,
      "step": 7435
    },
    {
      "entropy": 6.495653343200684,
      "epoch": 0.8586266589728794,
      "grad_norm": 1.375,
      "learning_rate": 0.0004937538173005462,
      "loss": 6.3712,
      "mean_token_accuracy": 0.1560351386666298,
      "num_tokens": 18844029.0,
      "step": 7440
    },
    {
      "entropy": 6.422978258132934,
      "epoch": 0.859203693017888,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004937441582077111,
      "loss": 6.3346,
      "mean_token_accuracy": 0.15770572125911714,
      "num_tokens": 18857468.0,
      "step": 7445
    },
    {
      "entropy": 6.425333786010742,
      "epoch": 0.8597807270628968,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004937344917575249,
      "loss": 6.2786,
      "mean_token_accuracy": 0.15311031341552733,
      "num_tokens": 18870487.0,
      "step": 7450
    },
    {
      "entropy": 6.4155045509338375,
      "epoch": 0.8603577611079054,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004937248179503127,
      "loss": 6.3045,
      "mean_token_accuracy": 0.1591097131371498,
      "num_tokens": 18883273.0,
      "step": 7455
    },
    {
      "entropy": 6.372232294082641,
      "epoch": 0.860934795152914,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004937151367863998,
      "loss": 6.2792,
      "mean_token_accuracy": 0.1714622050523758,
      "num_tokens": 18895863.0,
      "step": 7460
    },
    {
      "entropy": 6.392306852340698,
      "epoch": 0.8615118291979227,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000493705448266112,
      "loss": 6.3424,
      "mean_token_accuracy": 0.1597261071205139,
      "num_tokens": 18909406.0,
      "step": 7465
    },
    {
      "entropy": 6.38096022605896,
      "epoch": 0.8620888632429313,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004936957523897752,
      "loss": 6.3049,
      "mean_token_accuracy": 0.16298051476478576,
      "num_tokens": 18922794.0,
      "step": 7470
    },
    {
      "entropy": 6.395741033554077,
      "epoch": 0.86266589728794,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004936860491577153,
      "loss": 6.2997,
      "mean_token_accuracy": 0.1635509356856346,
      "num_tokens": 18936511.0,
      "step": 7475
    },
    {
      "entropy": 6.436820840835571,
      "epoch": 0.8632429313329486,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004936763385702588,
      "loss": 6.3391,
      "mean_token_accuracy": 0.15443028509616852,
      "num_tokens": 18948164.0,
      "step": 7480
    },
    {
      "entropy": 6.445798921585083,
      "epoch": 0.8638199653779572,
      "grad_norm": 2.765625,
      "learning_rate": 0.0004936666206277322,
      "loss": 6.3573,
      "mean_token_accuracy": 0.1542191758751869,
      "num_tokens": 18960608.0,
      "step": 7485
    },
    {
      "entropy": 6.390499258041382,
      "epoch": 0.864396999422966,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004936568953304624,
      "loss": 6.2226,
      "mean_token_accuracy": 0.17056952118873597,
      "num_tokens": 18972266.0,
      "step": 7490
    },
    {
      "entropy": 6.33178129196167,
      "epoch": 0.8649740334679746,
      "grad_norm": 2.953125,
      "learning_rate": 0.0004936471626787766,
      "loss": 6.2737,
      "mean_token_accuracy": 0.15900991410017012,
      "num_tokens": 18984518.0,
      "step": 7495
    },
    {
      "entropy": 6.370002889633179,
      "epoch": 0.8655510675129833,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004936374226730022,
      "loss": 6.2777,
      "mean_token_accuracy": 0.1658060982823372,
      "num_tokens": 18998611.0,
      "step": 7500
    },
    {
      "entropy": 6.413336277008057,
      "epoch": 0.8661281015579919,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004936276753134666,
      "loss": 6.2949,
      "mean_token_accuracy": 0.15621051490306853,
      "num_tokens": 19012117.0,
      "step": 7505
    },
    {
      "entropy": 6.277068948745727,
      "epoch": 0.8667051356030006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004936179206004976,
      "loss": 6.1665,
      "mean_token_accuracy": 0.16537234038114548,
      "num_tokens": 19024075.0,
      "step": 7510
    },
    {
      "entropy": 6.416962480545044,
      "epoch": 0.8672821696480092,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004936081585344236,
      "loss": 6.3606,
      "mean_token_accuracy": 0.15432593673467637,
      "num_tokens": 19037350.0,
      "step": 7515
    },
    {
      "entropy": 6.509782838821411,
      "epoch": 0.8678592036930178,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004935983891155727,
      "loss": 6.3553,
      "mean_token_accuracy": 0.15507148206233978,
      "num_tokens": 19049438.0,
      "step": 7520
    },
    {
      "entropy": 6.379511880874634,
      "epoch": 0.8684362377380266,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004935886123442737,
      "loss": 6.3226,
      "mean_token_accuracy": 0.15508285611867906,
      "num_tokens": 19061689.0,
      "step": 7525
    },
    {
      "entropy": 6.466004753112793,
      "epoch": 0.8690132717830352,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004935788282208551,
      "loss": 6.3727,
      "mean_token_accuracy": 0.15051396191120148,
      "num_tokens": 19075606.0,
      "step": 7530
    },
    {
      "entropy": 6.3925048351287845,
      "epoch": 0.8695903058280439,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004935690367456464,
      "loss": 6.2911,
      "mean_token_accuracy": 0.15860063582658768,
      "num_tokens": 19088248.0,
      "step": 7535
    },
    {
      "entropy": 6.416834545135498,
      "epoch": 0.8701673398730525,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004935592379189766,
      "loss": 6.2637,
      "mean_token_accuracy": 0.1581605017185211,
      "num_tokens": 19102480.0,
      "step": 7540
    },
    {
      "entropy": 6.363423204421997,
      "epoch": 0.8707443739180611,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004935494317411754,
      "loss": 6.2578,
      "mean_token_accuracy": 0.16764952838420868,
      "num_tokens": 19114454.0,
      "step": 7545
    },
    {
      "entropy": 6.3521568298339846,
      "epoch": 0.8713214079630698,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004935396182125726,
      "loss": 6.2943,
      "mean_token_accuracy": 0.16149284839630126,
      "num_tokens": 19126709.0,
      "step": 7550
    },
    {
      "entropy": 6.3308103561401365,
      "epoch": 0.8718984420080784,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004935297973334983,
      "loss": 6.2911,
      "mean_token_accuracy": 0.1629858657717705,
      "num_tokens": 19138831.0,
      "step": 7555
    },
    {
      "entropy": 6.431607007980347,
      "epoch": 0.8724754760530872,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004935199691042828,
      "loss": 6.1944,
      "mean_token_accuracy": 0.1633341908454895,
      "num_tokens": 19151555.0,
      "step": 7560
    },
    {
      "entropy": 6.300298261642456,
      "epoch": 0.8730525100980958,
      "grad_norm": 1.5,
      "learning_rate": 0.0004935101335252568,
      "loss": 6.2575,
      "mean_token_accuracy": 0.16245049238204956,
      "num_tokens": 19164557.0,
      "step": 7565
    },
    {
      "entropy": 6.392845678329468,
      "epoch": 0.8736295441431045,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004935002905967509,
      "loss": 6.3803,
      "mean_token_accuracy": 0.15307363122701645,
      "num_tokens": 19177081.0,
      "step": 7570
    },
    {
      "entropy": 6.405813980102539,
      "epoch": 0.8742065781881131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004934904403190963,
      "loss": 6.223,
      "mean_token_accuracy": 0.15920666754245758,
      "num_tokens": 19190209.0,
      "step": 7575
    },
    {
      "entropy": 6.415853786468506,
      "epoch": 0.8747836122331217,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004934805826926242,
      "loss": 6.2746,
      "mean_token_accuracy": 0.1588641032576561,
      "num_tokens": 19202262.0,
      "step": 7580
    },
    {
      "entropy": 6.336044120788574,
      "epoch": 0.8753606462781304,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004934707177176663,
      "loss": 6.3027,
      "mean_token_accuracy": 0.15697792172431946,
      "num_tokens": 19214752.0,
      "step": 7585
    },
    {
      "entropy": 6.4314311981201175,
      "epoch": 0.875937680323139,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004934608453945543,
      "loss": 6.3528,
      "mean_token_accuracy": 0.1546017959713936,
      "num_tokens": 19227320.0,
      "step": 7590
    },
    {
      "entropy": 6.463335561752319,
      "epoch": 0.8765147143681478,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004934509657236203,
      "loss": 6.33,
      "mean_token_accuracy": 0.15523393899202348,
      "num_tokens": 19239571.0,
      "step": 7595
    },
    {
      "entropy": 6.394824266433716,
      "epoch": 0.8770917484131564,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004934410787051965,
      "loss": 6.2782,
      "mean_token_accuracy": 0.16422777026891708,
      "num_tokens": 19252242.0,
      "step": 7600
    },
    {
      "entropy": 6.398877048492432,
      "epoch": 0.877668782458165,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004934311843396157,
      "loss": 6.3362,
      "mean_token_accuracy": 0.16020542681217192,
      "num_tokens": 19265501.0,
      "step": 7605
    },
    {
      "entropy": 6.389577150344849,
      "epoch": 0.8782458165031737,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004934212826272104,
      "loss": 6.1916,
      "mean_token_accuracy": 0.1710354968905449,
      "num_tokens": 19278152.0,
      "step": 7610
    },
    {
      "entropy": 6.268189096450806,
      "epoch": 0.8788228505481823,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004934113735683137,
      "loss": 6.3076,
      "mean_token_accuracy": 0.15756093189120293,
      "num_tokens": 19290555.0,
      "step": 7615
    },
    {
      "entropy": 6.378724002838135,
      "epoch": 0.879399884593191,
      "grad_norm": 1.234375,
      "learning_rate": 0.000493401457163259,
      "loss": 6.252,
      "mean_token_accuracy": 0.16543366014957428,
      "num_tokens": 19302866.0,
      "step": 7620
    },
    {
      "entropy": 6.493717765808105,
      "epoch": 0.8799769186381996,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004933915334123798,
      "loss": 6.349,
      "mean_token_accuracy": 0.15103889033198356,
      "num_tokens": 19315577.0,
      "step": 7625
    },
    {
      "entropy": 6.472537183761597,
      "epoch": 0.8805539526832084,
      "grad_norm": 2.28125,
      "learning_rate": 0.0004933816023160099,
      "loss": 6.3288,
      "mean_token_accuracy": 0.15646580904722213,
      "num_tokens": 19328593.0,
      "step": 7630
    },
    {
      "entropy": 6.432592248916626,
      "epoch": 0.881130986728217,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004933716638744832,
      "loss": 6.2837,
      "mean_token_accuracy": 0.1658412992954254,
      "num_tokens": 19341669.0,
      "step": 7635
    },
    {
      "entropy": 6.324086666107178,
      "epoch": 0.8817080207732256,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000493361718088134,
      "loss": 6.2859,
      "mean_token_accuracy": 0.15649446547031404,
      "num_tokens": 19354268.0,
      "step": 7640
    },
    {
      "entropy": 6.411100578308106,
      "epoch": 0.8822850548182343,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000493351764957297,
      "loss": 6.3361,
      "mean_token_accuracy": 0.15419866889715195,
      "num_tokens": 19367676.0,
      "step": 7645
    },
    {
      "entropy": 6.486693906784057,
      "epoch": 0.8828620888632429,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004933418044823068,
      "loss": 6.3086,
      "mean_token_accuracy": 0.16112906634807586,
      "num_tokens": 19380726.0,
      "step": 7650
    },
    {
      "entropy": 6.410530614852905,
      "epoch": 0.8834391229082516,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004933318366634984,
      "loss": 6.3102,
      "mean_token_accuracy": 0.15761423408985137,
      "num_tokens": 19392917.0,
      "step": 7655
    },
    {
      "entropy": 6.391311740875244,
      "epoch": 0.8840161569532602,
      "grad_norm": 2.25,
      "learning_rate": 0.0004933218615012072,
      "loss": 6.3448,
      "mean_token_accuracy": 0.1594384342432022,
      "num_tokens": 19406026.0,
      "step": 7660
    },
    {
      "entropy": 6.393275213241577,
      "epoch": 0.8845931909982689,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004933118789957687,
      "loss": 6.2815,
      "mean_token_accuracy": 0.1612059071660042,
      "num_tokens": 19418612.0,
      "step": 7665
    },
    {
      "entropy": 6.414052724838257,
      "epoch": 0.8851702250432776,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004933018891475186,
      "loss": 6.3248,
      "mean_token_accuracy": 0.1564637005329132,
      "num_tokens": 19430746.0,
      "step": 7670
    },
    {
      "entropy": 6.396469593048096,
      "epoch": 0.8857472590882862,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004932918919567927,
      "loss": 6.3689,
      "mean_token_accuracy": 0.1536492109298706,
      "num_tokens": 19443043.0,
      "step": 7675
    },
    {
      "entropy": 6.478463459014892,
      "epoch": 0.8863242931332949,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004932818874239275,
      "loss": 6.3276,
      "mean_token_accuracy": 0.15937264189124106,
      "num_tokens": 19456463.0,
      "step": 7680
    },
    {
      "entropy": 6.444648313522339,
      "epoch": 0.8869013271783035,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004932718755492595,
      "loss": 6.305,
      "mean_token_accuracy": 0.15255717635154725,
      "num_tokens": 19468108.0,
      "step": 7685
    },
    {
      "entropy": 6.367991638183594,
      "epoch": 0.8874783612233121,
      "grad_norm": 1.75,
      "learning_rate": 0.0004932618563331254,
      "loss": 6.3265,
      "mean_token_accuracy": 0.15850533843040465,
      "num_tokens": 19480741.0,
      "step": 7690
    },
    {
      "entropy": 6.35459451675415,
      "epoch": 0.8880553952683208,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004932518297758622,
      "loss": 6.2542,
      "mean_token_accuracy": 0.1637923687696457,
      "num_tokens": 19492996.0,
      "step": 7695
    },
    {
      "entropy": 6.439085531234741,
      "epoch": 0.8886324293133295,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004932417958778071,
      "loss": 6.2594,
      "mean_token_accuracy": 0.16098986119031905,
      "num_tokens": 19505513.0,
      "step": 7700
    },
    {
      "entropy": 6.417087125778198,
      "epoch": 0.8892094633583382,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004932317546392976,
      "loss": 6.3979,
      "mean_token_accuracy": 0.15393300205469132,
      "num_tokens": 19518244.0,
      "step": 7705
    },
    {
      "entropy": 6.320071363449097,
      "epoch": 0.8897864974033468,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004932217060606714,
      "loss": 6.2906,
      "mean_token_accuracy": 0.1595987156033516,
      "num_tokens": 19530987.0,
      "step": 7710
    },
    {
      "entropy": 6.3398231029510494,
      "epoch": 0.8903635314483554,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004932116501422665,
      "loss": 6.2154,
      "mean_token_accuracy": 0.1674814835190773,
      "num_tokens": 19543546.0,
      "step": 7715
    },
    {
      "entropy": 6.446526670455933,
      "epoch": 0.8909405654933641,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004932015868844211,
      "loss": 6.2946,
      "mean_token_accuracy": 0.1571994751691818,
      "num_tokens": 19556867.0,
      "step": 7720
    },
    {
      "entropy": 6.37026309967041,
      "epoch": 0.8915175995383727,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004931915162874738,
      "loss": 6.2828,
      "mean_token_accuracy": 0.15851514786481857,
      "num_tokens": 19569738.0,
      "step": 7725
    },
    {
      "entropy": 6.363567304611206,
      "epoch": 0.8920946335833814,
      "grad_norm": 1.75,
      "learning_rate": 0.0004931814383517632,
      "loss": 6.3218,
      "mean_token_accuracy": 0.15662410855293274,
      "num_tokens": 19581701.0,
      "step": 7730
    },
    {
      "entropy": 6.377171468734741,
      "epoch": 0.8926716676283901,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004931713530776284,
      "loss": 6.247,
      "mean_token_accuracy": 0.16711300909519194,
      "num_tokens": 19593698.0,
      "step": 7735
    },
    {
      "entropy": 6.349302959442139,
      "epoch": 0.8932487016733988,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004931612604654083,
      "loss": 6.303,
      "mean_token_accuracy": 0.15497682094573975,
      "num_tokens": 19605323.0,
      "step": 7740
    },
    {
      "entropy": 6.404062414169312,
      "epoch": 0.8938257357184074,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004931511605154428,
      "loss": 6.3447,
      "mean_token_accuracy": 0.1544952780008316,
      "num_tokens": 19616577.0,
      "step": 7745
    },
    {
      "entropy": 6.473753356933594,
      "epoch": 0.894402769763416,
      "grad_norm": 3.046875,
      "learning_rate": 0.0004931410532280711,
      "loss": 6.2382,
      "mean_token_accuracy": 0.15787958949804307,
      "num_tokens": 19629177.0,
      "step": 7750
    },
    {
      "entropy": 6.344061231613159,
      "epoch": 0.8949798038084247,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004931309386036337,
      "loss": 6.2839,
      "mean_token_accuracy": 0.15972928553819657,
      "num_tokens": 19642170.0,
      "step": 7755
    },
    {
      "entropy": 6.308340644836425,
      "epoch": 0.8955568378534333,
      "grad_norm": 4.59375,
      "learning_rate": 0.0004931208166424704,
      "loss": 6.2324,
      "mean_token_accuracy": 0.16119810193777084,
      "num_tokens": 19655648.0,
      "step": 7760
    },
    {
      "entropy": 6.403913593292236,
      "epoch": 0.896133871898442,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004931106873449219,
      "loss": 6.2052,
      "mean_token_accuracy": 0.16124669313430787,
      "num_tokens": 19667941.0,
      "step": 7765
    },
    {
      "entropy": 6.437334489822388,
      "epoch": 0.8967109059434507,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004931005507113285,
      "loss": 6.3193,
      "mean_token_accuracy": 0.1630260318517685,
      "num_tokens": 19680523.0,
      "step": 7770
    },
    {
      "entropy": 6.261597490310669,
      "epoch": 0.8972879399884593,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004930904067420317,
      "loss": 6.276,
      "mean_token_accuracy": 0.16263071596622466,
      "num_tokens": 19692984.0,
      "step": 7775
    },
    {
      "entropy": 6.432254409790039,
      "epoch": 0.897864974033468,
      "grad_norm": 4.71875,
      "learning_rate": 0.0004930802554373723,
      "loss": 6.2354,
      "mean_token_accuracy": 0.16485002189874648,
      "num_tokens": 19705926.0,
      "step": 7780
    },
    {
      "entropy": 6.449591398239136,
      "epoch": 0.8984420080784766,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004930700967976918,
      "loss": 6.2797,
      "mean_token_accuracy": 0.16124334186315536,
      "num_tokens": 19717322.0,
      "step": 7785
    },
    {
      "entropy": 6.39738130569458,
      "epoch": 0.8990190421234853,
      "grad_norm": 1.578125,
      "learning_rate": 0.000493059930823332,
      "loss": 6.2975,
      "mean_token_accuracy": 0.14967879951000213,
      "num_tokens": 19730577.0,
      "step": 7790
    },
    {
      "entropy": 6.44396858215332,
      "epoch": 0.8995960761684939,
      "grad_norm": 2.125,
      "learning_rate": 0.0004930497575146346,
      "loss": 6.3116,
      "mean_token_accuracy": 0.15643653869628907,
      "num_tokens": 19744054.0,
      "step": 7795
    },
    {
      "entropy": 6.342684650421143,
      "epoch": 0.9001731102135025,
      "grad_norm": 1.25,
      "learning_rate": 0.0004930395768719421,
      "loss": 6.2894,
      "mean_token_accuracy": 0.1550717294216156,
      "num_tokens": 19755898.0,
      "step": 7800
    },
    {
      "entropy": 6.396816730499268,
      "epoch": 0.9007501442585113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004930293888955966,
      "loss": 6.3557,
      "mean_token_accuracy": 0.15300133675336838,
      "num_tokens": 19769599.0,
      "step": 7805
    },
    {
      "entropy": 6.453741598129272,
      "epoch": 0.9013271783035199,
      "grad_norm": 1.34375,
      "learning_rate": 0.000493019193585941,
      "loss": 6.3723,
      "mean_token_accuracy": 0.14918754994869232,
      "num_tokens": 19783442.0,
      "step": 7810
    },
    {
      "entropy": 6.4418963432312015,
      "epoch": 0.9019042123485286,
      "grad_norm": 2.046875,
      "learning_rate": 0.000493008990943318,
      "loss": 6.3659,
      "mean_token_accuracy": 0.15801341980695724,
      "num_tokens": 19796039.0,
      "step": 7815
    },
    {
      "entropy": 6.478892612457275,
      "epoch": 0.9024812463935372,
      "grad_norm": 2.375,
      "learning_rate": 0.0004929987809680709,
      "loss": 6.3283,
      "mean_token_accuracy": 0.16183366179466246,
      "num_tokens": 19808695.0,
      "step": 7820
    },
    {
      "entropy": 6.390950679779053,
      "epoch": 0.9030582804385459,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004929885636605432,
      "loss": 6.2872,
      "mean_token_accuracy": 0.15758599042892457,
      "num_tokens": 19821173.0,
      "step": 7825
    },
    {
      "entropy": 6.370595455169678,
      "epoch": 0.9036353144835545,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004929783390210784,
      "loss": 6.3441,
      "mean_token_accuracy": 0.15149060785770416,
      "num_tokens": 19833804.0,
      "step": 7830
    },
    {
      "entropy": 6.405590677261353,
      "epoch": 0.9042123485285631,
      "grad_norm": 2.3125,
      "learning_rate": 0.0004929681070500204,
      "loss": 6.2947,
      "mean_token_accuracy": 0.1585235998034477,
      "num_tokens": 19845690.0,
      "step": 7835
    },
    {
      "entropy": 6.453483057022095,
      "epoch": 0.9047893825735719,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004929578677477135,
      "loss": 6.3082,
      "mean_token_accuracy": 0.1503335326910019,
      "num_tokens": 19859463.0,
      "step": 7840
    },
    {
      "entropy": 6.3136523246765135,
      "epoch": 0.9053664166185805,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004929476211145019,
      "loss": 6.1993,
      "mean_token_accuracy": 0.16918923556804658,
      "num_tokens": 19872168.0,
      "step": 7845
    },
    {
      "entropy": 6.331404352188111,
      "epoch": 0.9059434506635892,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004929373671507303,
      "loss": 6.2647,
      "mean_token_accuracy": 0.16442092657089233,
      "num_tokens": 19884927.0,
      "step": 7850
    },
    {
      "entropy": 6.3881467342376705,
      "epoch": 0.9065204847085978,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004929271058567436,
      "loss": 6.2781,
      "mean_token_accuracy": 0.16083337366580963,
      "num_tokens": 19897105.0,
      "step": 7855
    },
    {
      "entropy": 6.376621198654175,
      "epoch": 0.9070975187536064,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000492916837232887,
      "loss": 6.2943,
      "mean_token_accuracy": 0.1583924874663353,
      "num_tokens": 19909973.0,
      "step": 7860
    },
    {
      "entropy": 6.371397638320923,
      "epoch": 0.9076745527986151,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004929065612795058,
      "loss": 6.2893,
      "mean_token_accuracy": 0.1654267877340317,
      "num_tokens": 19923674.0,
      "step": 7865
    },
    {
      "entropy": 6.401323223114014,
      "epoch": 0.9082515868436237,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004928962779969456,
      "loss": 6.3279,
      "mean_token_accuracy": 0.1563215509057045,
      "num_tokens": 19935742.0,
      "step": 7870
    },
    {
      "entropy": 6.431182050704956,
      "epoch": 0.9088286208886325,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004928859873855524,
      "loss": 6.2985,
      "mean_token_accuracy": 0.15487526804208757,
      "num_tokens": 19947945.0,
      "step": 7875
    },
    {
      "entropy": 6.463040542602539,
      "epoch": 0.9094056549336411,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004928756894456723,
      "loss": 6.2667,
      "mean_token_accuracy": 0.16083851903676988,
      "num_tokens": 19961275.0,
      "step": 7880
    },
    {
      "entropy": 6.431319952011108,
      "epoch": 0.9099826889786498,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004928653841776515,
      "loss": 6.3134,
      "mean_token_accuracy": 0.15611430555582045,
      "num_tokens": 19973634.0,
      "step": 7885
    },
    {
      "entropy": 6.450580596923828,
      "epoch": 0.9105597230236584,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004928550715818368,
      "loss": 6.3138,
      "mean_token_accuracy": 0.15701612830162048,
      "num_tokens": 19985920.0,
      "step": 7890
    },
    {
      "entropy": 6.391114807128906,
      "epoch": 0.911136757068667,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004928447516585749,
      "loss": 6.3351,
      "mean_token_accuracy": 0.15266132205724717,
      "num_tokens": 19999081.0,
      "step": 7895
    },
    {
      "entropy": 6.3048357486724855,
      "epoch": 0.9117137911136757,
      "grad_norm": 1.28125,
      "learning_rate": 0.000492834424408213,
      "loss": 6.2833,
      "mean_token_accuracy": 0.16531692892313005,
      "num_tokens": 20012801.0,
      "step": 7900
    },
    {
      "entropy": 6.479332065582275,
      "epoch": 0.9122908251586843,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004928240898310984,
      "loss": 6.3693,
      "mean_token_accuracy": 0.15050050467252732,
      "num_tokens": 20024618.0,
      "step": 7905
    },
    {
      "entropy": 6.47349362373352,
      "epoch": 0.9128678592036931,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004928137479275789,
      "loss": 6.3076,
      "mean_token_accuracy": 0.16095187067985534,
      "num_tokens": 20038587.0,
      "step": 7910
    },
    {
      "entropy": 6.414200925827027,
      "epoch": 0.9134448932487017,
      "grad_norm": 2.015625,
      "learning_rate": 0.000492803398698002,
      "loss": 6.3259,
      "mean_token_accuracy": 0.15401843264698983,
      "num_tokens": 20051811.0,
      "step": 7915
    },
    {
      "entropy": 6.4483012676239015,
      "epoch": 0.9140219272937103,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004927930421427161,
      "loss": 6.3042,
      "mean_token_accuracy": 0.15708602219820023,
      "num_tokens": 20064463.0,
      "step": 7920
    },
    {
      "entropy": 6.372395849227905,
      "epoch": 0.914598961338719,
      "grad_norm": 5.53125,
      "learning_rate": 0.0004927826782620694,
      "loss": 6.3181,
      "mean_token_accuracy": 0.1550089493393898,
      "num_tokens": 20076614.0,
      "step": 7925
    },
    {
      "entropy": 6.468643712997436,
      "epoch": 0.9151759953837276,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004927723070564104,
      "loss": 6.4253,
      "mean_token_accuracy": 0.15085637867450713,
      "num_tokens": 20090395.0,
      "step": 7930
    },
    {
      "entropy": 6.419011640548706,
      "epoch": 0.9157530294287363,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004927619285260881,
      "loss": 6.3461,
      "mean_token_accuracy": 0.15481395423412322,
      "num_tokens": 20102736.0,
      "step": 7935
    },
    {
      "entropy": 6.410750818252564,
      "epoch": 0.9163300634737449,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004927515426714515,
      "loss": 6.321,
      "mean_token_accuracy": 0.15636474043130874,
      "num_tokens": 20116516.0,
      "step": 7940
    },
    {
      "entropy": 6.385376882553101,
      "epoch": 0.9169070975187537,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00049274114949285,
      "loss": 6.291,
      "mean_token_accuracy": 0.16122057735919954,
      "num_tokens": 20128742.0,
      "step": 7945
    },
    {
      "entropy": 6.438413190841675,
      "epoch": 0.9174841315637623,
      "grad_norm": 1.53125,
      "learning_rate": 0.000492730748990633,
      "loss": 6.2602,
      "mean_token_accuracy": 0.1578374445438385,
      "num_tokens": 20140590.0,
      "step": 7950
    },
    {
      "entropy": 6.3685081005096436,
      "epoch": 0.9180611656087709,
      "grad_norm": 1.5,
      "learning_rate": 0.0004927203411651504,
      "loss": 6.2941,
      "mean_token_accuracy": 0.15343682318925858,
      "num_tokens": 20153748.0,
      "step": 7955
    },
    {
      "entropy": 6.46742205619812,
      "epoch": 0.9186381996537796,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004927099260167523,
      "loss": 6.2976,
      "mean_token_accuracy": 0.1620119497179985,
      "num_tokens": 20167094.0,
      "step": 7960
    },
    {
      "entropy": 6.400900650024414,
      "epoch": 0.9192152336987882,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004926995035457889,
      "loss": 6.2755,
      "mean_token_accuracy": 0.15437212586402893,
      "num_tokens": 20180716.0,
      "step": 7965
    },
    {
      "entropy": 6.422587203979492,
      "epoch": 0.9197922677437969,
      "grad_norm": 1.921875,
      "learning_rate": 0.000492689073752611,
      "loss": 6.3306,
      "mean_token_accuracy": 0.158352267742157,
      "num_tokens": 20192214.0,
      "step": 7970
    },
    {
      "entropy": 6.322636461257934,
      "epoch": 0.9203693017888055,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004926786366375691,
      "loss": 6.186,
      "mean_token_accuracy": 0.159789976477623,
      "num_tokens": 20204789.0,
      "step": 7975
    },
    {
      "entropy": 6.395492315292358,
      "epoch": 0.9209463358338142,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004926681922010145,
      "loss": 6.2937,
      "mean_token_accuracy": 0.15341563820838927,
      "num_tokens": 20217318.0,
      "step": 7980
    },
    {
      "entropy": 6.503517580032349,
      "epoch": 0.9215233698788229,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004926577404432982,
      "loss": 6.2776,
      "mean_token_accuracy": 0.15594429075717925,
      "num_tokens": 20229757.0,
      "step": 7985
    },
    {
      "entropy": 6.401009654998779,
      "epoch": 0.9221004039238315,
      "grad_norm": 1.5,
      "learning_rate": 0.0004926472813647721,
      "loss": 6.3206,
      "mean_token_accuracy": 0.1593426927924156,
      "num_tokens": 20242042.0,
      "step": 7990
    },
    {
      "entropy": 6.317966318130493,
      "epoch": 0.9226774379688402,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004926368149657876,
      "loss": 6.2308,
      "mean_token_accuracy": 0.16747417151927949,
      "num_tokens": 20255188.0,
      "step": 7995
    },
    {
      "entropy": 6.4740152835845945,
      "epoch": 0.9232544720138488,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004926263412466972,
      "loss": 6.2961,
      "mean_token_accuracy": 0.16103923618793486,
      "num_tokens": 20267359.0,
      "step": 8000
    },
    {
      "entropy": 6.388199758529663,
      "epoch": 0.9238315060588574,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004926158602078527,
      "loss": 6.3689,
      "mean_token_accuracy": 0.15565205961465836,
      "num_tokens": 20280023.0,
      "step": 8005
    },
    {
      "entropy": 6.334804201126099,
      "epoch": 0.9244085401038661,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004926053718496069,
      "loss": 6.2985,
      "mean_token_accuracy": 0.15743525475263595,
      "num_tokens": 20293043.0,
      "step": 8010
    },
    {
      "entropy": 6.403419828414917,
      "epoch": 0.9249855741488748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004925948761723126,
      "loss": 6.247,
      "mean_token_accuracy": 0.15326788574457167,
      "num_tokens": 20304929.0,
      "step": 8015
    },
    {
      "entropy": 6.409779071807861,
      "epoch": 0.9255626081938835,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004925843731763226,
      "loss": 6.242,
      "mean_token_accuracy": 0.15896770507097244,
      "num_tokens": 20317024.0,
      "step": 8020
    },
    {
      "entropy": 6.434144783020019,
      "epoch": 0.9261396422388921,
      "grad_norm": 2.046875,
      "learning_rate": 0.0004925738628619904,
      "loss": 6.3182,
      "mean_token_accuracy": 0.1545105457305908,
      "num_tokens": 20329909.0,
      "step": 8025
    },
    {
      "entropy": 6.324703788757324,
      "epoch": 0.9267166762839008,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004925633452296693,
      "loss": 6.28,
      "mean_token_accuracy": 0.15805017352104186,
      "num_tokens": 20344224.0,
      "step": 8030
    },
    {
      "entropy": 6.506214904785156,
      "epoch": 0.9272937103289094,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004925528202797131,
      "loss": 6.3046,
      "mean_token_accuracy": 0.15081962794065476,
      "num_tokens": 20357393.0,
      "step": 8035
    },
    {
      "entropy": 6.424812507629395,
      "epoch": 0.927870744373918,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004925422880124761,
      "loss": 6.3061,
      "mean_token_accuracy": 0.16021231263875962,
      "num_tokens": 20369551.0,
      "step": 8040
    },
    {
      "entropy": 6.379309225082397,
      "epoch": 0.9284477784189267,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004925317484283121,
      "loss": 6.4167,
      "mean_token_accuracy": 0.14904050678014755,
      "num_tokens": 20382639.0,
      "step": 8045
    },
    {
      "entropy": 6.429602766036988,
      "epoch": 0.9290248124639354,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004925212015275758,
      "loss": 6.2348,
      "mean_token_accuracy": 0.16726955473423005,
      "num_tokens": 20394751.0,
      "step": 8050
    },
    {
      "entropy": 6.39623794555664,
      "epoch": 0.9296018465089441,
      "grad_norm": 1.46875,
      "learning_rate": 0.000492510647310622,
      "loss": 6.2848,
      "mean_token_accuracy": 0.1595268651843071,
      "num_tokens": 20408235.0,
      "step": 8055
    },
    {
      "entropy": 6.398084878921509,
      "epoch": 0.9301788805539527,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004925000857778055,
      "loss": 6.3392,
      "mean_token_accuracy": 0.15427375137805938,
      "num_tokens": 20423063.0,
      "step": 8060
    },
    {
      "entropy": 6.427191162109375,
      "epoch": 0.9307559145989613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004924895169294818,
      "loss": 6.2369,
      "mean_token_accuracy": 0.1577201783657074,
      "num_tokens": 20436225.0,
      "step": 8065
    },
    {
      "entropy": 6.4034092903137205,
      "epoch": 0.93133294864397,
      "grad_norm": 2.0,
      "learning_rate": 0.0004924789407660062,
      "loss": 6.3375,
      "mean_token_accuracy": 0.1600277066230774,
      "num_tokens": 20447957.0,
      "step": 8070
    },
    {
      "entropy": 6.381423425674439,
      "epoch": 0.9319099826889786,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004924683572877345,
      "loss": 6.2226,
      "mean_token_accuracy": 0.16029435247182847,
      "num_tokens": 20459836.0,
      "step": 8075
    },
    {
      "entropy": 6.533320760726928,
      "epoch": 0.9324870167339873,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004924577664950225,
      "loss": 6.3824,
      "mean_token_accuracy": 0.15089137703180314,
      "num_tokens": 20471907.0,
      "step": 8080
    },
    {
      "entropy": 6.440421152114868,
      "epoch": 0.933064050778996,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004924471683882266,
      "loss": 6.3825,
      "mean_token_accuracy": 0.15363826006650924,
      "num_tokens": 20485166.0,
      "step": 8085
    },
    {
      "entropy": 6.420388221740723,
      "epoch": 0.9336410848240047,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004924365629677031,
      "loss": 6.1845,
      "mean_token_accuracy": 0.16712835282087327,
      "num_tokens": 20498973.0,
      "step": 8090
    },
    {
      "entropy": 6.355935859680176,
      "epoch": 0.9342181188690133,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000492425950233809,
      "loss": 6.2747,
      "mean_token_accuracy": 0.16278841942548752,
      "num_tokens": 20511615.0,
      "step": 8095
    },
    {
      "entropy": 6.3108885288238525,
      "epoch": 0.9347951529140219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000492415330186901,
      "loss": 6.2678,
      "mean_token_accuracy": 0.15716547816991805,
      "num_tokens": 20523615.0,
      "step": 8100
    },
    {
      "entropy": 6.421422386169434,
      "epoch": 0.9353721869590306,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004924047028273364,
      "loss": 6.2418,
      "mean_token_accuracy": 0.15872604250907899,
      "num_tokens": 20537048.0,
      "step": 8105
    },
    {
      "entropy": 6.393527889251709,
      "epoch": 0.9359492210040392,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004923940681554725,
      "loss": 6.301,
      "mean_token_accuracy": 0.1614370346069336,
      "num_tokens": 20551041.0,
      "step": 8110
    },
    {
      "entropy": 6.415983819961548,
      "epoch": 0.9365262550490479,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004923834261716672,
      "loss": 6.3455,
      "mean_token_accuracy": 0.15478847175836563,
      "num_tokens": 20563123.0,
      "step": 8115
    },
    {
      "entropy": 6.448668527603149,
      "epoch": 0.9371032890940566,
      "grad_norm": 2.75,
      "learning_rate": 0.0004923727768762782,
      "loss": 6.2795,
      "mean_token_accuracy": 0.16725585907697677,
      "num_tokens": 20577487.0,
      "step": 8120
    },
    {
      "entropy": 6.402075386047363,
      "epoch": 0.9376803231390652,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000492362120269664,
      "loss": 6.2682,
      "mean_token_accuracy": 0.1627908691763878,
      "num_tokens": 20590125.0,
      "step": 8125
    },
    {
      "entropy": 6.418191576004029,
      "epoch": 0.9382573571840739,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004923514563521827,
      "loss": 6.3141,
      "mean_token_accuracy": 0.15729555934667588,
      "num_tokens": 20603035.0,
      "step": 8130
    },
    {
      "entropy": 6.403626775741577,
      "epoch": 0.9388343912290825,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004923407851241933,
      "loss": 6.2841,
      "mean_token_accuracy": 0.15334731638431548,
      "num_tokens": 20615280.0,
      "step": 8135
    },
    {
      "entropy": 6.4232524871826175,
      "epoch": 0.9394114252740912,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004923301065860545,
      "loss": 6.2766,
      "mean_token_accuracy": 0.15411915630102158,
      "num_tokens": 20627800.0,
      "step": 8140
    },
    {
      "entropy": 6.412774085998535,
      "epoch": 0.9399884593190998,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004923194207381254,
      "loss": 6.2985,
      "mean_token_accuracy": 0.15818233788013458,
      "num_tokens": 20639264.0,
      "step": 8145
    },
    {
      "entropy": 6.378032875061035,
      "epoch": 0.9405654933641084,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004923087275807656,
      "loss": 6.2204,
      "mean_token_accuracy": 0.1706040620803833,
      "num_tokens": 20652155.0,
      "step": 8150
    },
    {
      "entropy": 6.391241407394409,
      "epoch": 0.9411425274091172,
      "grad_norm": 3.296875,
      "learning_rate": 0.0004922980271143347,
      "loss": 6.3063,
      "mean_token_accuracy": 0.15853582173585892,
      "num_tokens": 20665642.0,
      "step": 8155
    },
    {
      "entropy": 6.422382640838623,
      "epoch": 0.9417195614541258,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004922873193391927,
      "loss": 6.2504,
      "mean_token_accuracy": 0.16132109612226486,
      "num_tokens": 20678685.0,
      "step": 8160
    },
    {
      "entropy": 6.333035278320312,
      "epoch": 0.9422965954991345,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004922766042556994,
      "loss": 6.21,
      "mean_token_accuracy": 0.15850515216588973,
      "num_tokens": 20690714.0,
      "step": 8165
    },
    {
      "entropy": 6.418743991851807,
      "epoch": 0.9428736295441431,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004922658818642156,
      "loss": 6.3091,
      "mean_token_accuracy": 0.1515625908970833,
      "num_tokens": 20703316.0,
      "step": 8170
    },
    {
      "entropy": 6.327837753295898,
      "epoch": 0.9434506635891518,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004922551521651018,
      "loss": 6.204,
      "mean_token_accuracy": 0.1694895327091217,
      "num_tokens": 20716144.0,
      "step": 8175
    },
    {
      "entropy": 6.418542194366455,
      "epoch": 0.9440276976341604,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004922444151587189,
      "loss": 6.2522,
      "mean_token_accuracy": 0.15802036374807357,
      "num_tokens": 20728089.0,
      "step": 8180
    },
    {
      "entropy": 6.298161745071411,
      "epoch": 0.944604731679169,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004922336708454279,
      "loss": 6.2609,
      "mean_token_accuracy": 0.16404679417610168,
      "num_tokens": 20741684.0,
      "step": 8185
    },
    {
      "entropy": 6.420363283157348,
      "epoch": 0.9451817657241778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004922229192255903,
      "loss": 6.3386,
      "mean_token_accuracy": 0.15321814119815827,
      "num_tokens": 20754840.0,
      "step": 8190
    },
    {
      "entropy": 6.379177618026733,
      "epoch": 0.9457587997691864,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004922121602995678,
      "loss": 6.2035,
      "mean_token_accuracy": 0.16860224455595016,
      "num_tokens": 20767528.0,
      "step": 8195
    },
    {
      "entropy": 6.300572204589844,
      "epoch": 0.9463358338141951,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004922013940677221,
      "loss": 6.2703,
      "mean_token_accuracy": 0.16143476068973542,
      "num_tokens": 20778886.0,
      "step": 8200
    },
    {
      "entropy": 6.394140338897705,
      "epoch": 0.9469128678592037,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004921906205304155,
      "loss": 6.2467,
      "mean_token_accuracy": 0.1543475106358528,
      "num_tokens": 20792928.0,
      "step": 8205
    },
    {
      "entropy": 6.387519454956054,
      "epoch": 0.9474899019042123,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004921798396880101,
      "loss": 6.2372,
      "mean_token_accuracy": 0.1540800616145134,
      "num_tokens": 20805766.0,
      "step": 8210
    },
    {
      "entropy": 6.3942780017852785,
      "epoch": 0.948066935949221,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004921690515408688,
      "loss": 6.273,
      "mean_token_accuracy": 0.1582356408238411,
      "num_tokens": 20819008.0,
      "step": 8215
    },
    {
      "entropy": 6.369224262237549,
      "epoch": 0.9486439699942296,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004921582560893543,
      "loss": 6.3415,
      "mean_token_accuracy": 0.1553818851709366,
      "num_tokens": 20832706.0,
      "step": 8220
    },
    {
      "entropy": 6.4350669384002686,
      "epoch": 0.9492210040392384,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004921474533338297,
      "loss": 6.2547,
      "mean_token_accuracy": 0.16016023606061935,
      "num_tokens": 20844568.0,
      "step": 8225
    },
    {
      "entropy": 6.399309062957764,
      "epoch": 0.949798038084247,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004921366432746585,
      "loss": 6.2834,
      "mean_token_accuracy": 0.1573599874973297,
      "num_tokens": 20856157.0,
      "step": 8230
    },
    {
      "entropy": 6.362299489974975,
      "epoch": 0.9503750721292556,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000492125825912204,
      "loss": 6.2712,
      "mean_token_accuracy": 0.16097963899374007,
      "num_tokens": 20868039.0,
      "step": 8235
    },
    {
      "entropy": 6.463052082061767,
      "epoch": 0.9509521061742643,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004921150012468302,
      "loss": 6.2857,
      "mean_token_accuracy": 0.15774607434868812,
      "num_tokens": 20880190.0,
      "step": 8240
    },
    {
      "entropy": 6.41200098991394,
      "epoch": 0.9515291402192729,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004921041692789013,
      "loss": 6.2571,
      "mean_token_accuracy": 0.16535381972789764,
      "num_tokens": 20892996.0,
      "step": 8245
    },
    {
      "entropy": 6.455173110961914,
      "epoch": 0.9521061742642816,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004920933300087813,
      "loss": 6.3289,
      "mean_token_accuracy": 0.15742640495300292,
      "num_tokens": 20906080.0,
      "step": 8250
    },
    {
      "entropy": 6.430334520339966,
      "epoch": 0.9526832083092902,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004920824834368353,
      "loss": 6.3609,
      "mean_token_accuracy": 0.15268936306238173,
      "num_tokens": 20917944.0,
      "step": 8255
    },
    {
      "entropy": 6.340414762496948,
      "epoch": 0.953260242354299,
      "grad_norm": 2.46875,
      "learning_rate": 0.0004920716295634275,
      "loss": 6.2397,
      "mean_token_accuracy": 0.17364375591278075,
      "num_tokens": 20929046.0,
      "step": 8260
    },
    {
      "entropy": 6.455630874633789,
      "epoch": 0.9538372763993076,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004920607683889235,
      "loss": 6.3942,
      "mean_token_accuracy": 0.1561593271791935,
      "num_tokens": 20941643.0,
      "step": 8265
    },
    {
      "entropy": 6.538067293167114,
      "epoch": 0.9544143104443162,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004920498999136882,
      "loss": 6.3306,
      "mean_token_accuracy": 0.1545989155769348,
      "num_tokens": 20954620.0,
      "step": 8270
    },
    {
      "entropy": 6.4215727806091305,
      "epoch": 0.9549913444893249,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004920390241380874,
      "loss": 6.3429,
      "mean_token_accuracy": 0.15070486068725586,
      "num_tokens": 20968033.0,
      "step": 8275
    },
    {
      "entropy": 6.422377824783325,
      "epoch": 0.9555683785343335,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004920281410624868,
      "loss": 6.2661,
      "mean_token_accuracy": 0.16499000191688537,
      "num_tokens": 20980586.0,
      "step": 8280
    },
    {
      "entropy": 6.349878549575806,
      "epoch": 0.9561454125793422,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004920172506872525,
      "loss": 6.2084,
      "mean_token_accuracy": 0.16240258812904357,
      "num_tokens": 20992346.0,
      "step": 8285
    },
    {
      "entropy": 6.268904733657837,
      "epoch": 0.9567224466243508,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004920063530127508,
      "loss": 6.1527,
      "mean_token_accuracy": 0.17184782773256302,
      "num_tokens": 21004399.0,
      "step": 8290
    },
    {
      "entropy": 6.457413578033448,
      "epoch": 0.9572994806693594,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004919954480393482,
      "loss": 6.3675,
      "mean_token_accuracy": 0.1487828180193901,
      "num_tokens": 21017066.0,
      "step": 8295
    },
    {
      "entropy": 6.539893627166748,
      "epoch": 0.9578765147143682,
      "grad_norm": 2.53125,
      "learning_rate": 0.0004919845357674114,
      "loss": 6.4393,
      "mean_token_accuracy": 0.14757169634103776,
      "num_tokens": 21031664.0,
      "step": 8300
    },
    {
      "entropy": 6.500674772262573,
      "epoch": 0.9584535487593768,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004919736161973076,
      "loss": 6.272,
      "mean_token_accuracy": 0.1634308248758316,
      "num_tokens": 21043980.0,
      "step": 8305
    },
    {
      "entropy": 6.353540134429932,
      "epoch": 0.9590305828043855,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000491962689329404,
      "loss": 6.3015,
      "mean_token_accuracy": 0.15528810173273086,
      "num_tokens": 21056916.0,
      "step": 8310
    },
    {
      "entropy": 6.502210283279419,
      "epoch": 0.9596076168493941,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004919517551640681,
      "loss": 6.3659,
      "mean_token_accuracy": 0.1515677571296692,
      "num_tokens": 21070027.0,
      "step": 8315
    },
    {
      "entropy": 6.526283597946167,
      "epoch": 0.9601846508944027,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004919408137016677,
      "loss": 6.3125,
      "mean_token_accuracy": 0.15769415199756623,
      "num_tokens": 21082596.0,
      "step": 8320
    },
    {
      "entropy": 6.284257698059082,
      "epoch": 0.9607616849394114,
      "grad_norm": 3.359375,
      "learning_rate": 0.0004919298649425708,
      "loss": 6.1683,
      "mean_token_accuracy": 0.16701944917440414,
      "num_tokens": 21098160.0,
      "step": 8325
    },
    {
      "entropy": 6.364248371124267,
      "epoch": 0.96133871898442,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004919189088871456,
      "loss": 6.251,
      "mean_token_accuracy": 0.15590550601482392,
      "num_tokens": 21111562.0,
      "step": 8330
    },
    {
      "entropy": 6.41587643623352,
      "epoch": 0.9619157530294288,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004919079455357608,
      "loss": 6.2395,
      "mean_token_accuracy": 0.1601430892944336,
      "num_tokens": 21123177.0,
      "step": 8335
    },
    {
      "entropy": 6.406910991668701,
      "epoch": 0.9624927870744374,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004918969748887847,
      "loss": 6.2987,
      "mean_token_accuracy": 0.15528492927551268,
      "num_tokens": 21135390.0,
      "step": 8340
    },
    {
      "entropy": 6.391875457763672,
      "epoch": 0.9630698211194461,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004918859969465868,
      "loss": 6.2587,
      "mean_token_accuracy": 0.16090984791517257,
      "num_tokens": 21148718.0,
      "step": 8345
    },
    {
      "entropy": 6.409095191955567,
      "epoch": 0.9636468551644547,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004918750117095361,
      "loss": 6.2722,
      "mean_token_accuracy": 0.15558131486177446,
      "num_tokens": 21161057.0,
      "step": 8350
    },
    {
      "entropy": 6.4546857357025145,
      "epoch": 0.9642238892094633,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004918640191780021,
      "loss": 6.2665,
      "mean_token_accuracy": 0.15689287930727006,
      "num_tokens": 21175048.0,
      "step": 8355
    },
    {
      "entropy": 6.404264736175537,
      "epoch": 0.964800923254472,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004918530193523546,
      "loss": 6.2031,
      "mean_token_accuracy": 0.16269493997097015,
      "num_tokens": 21187995.0,
      "step": 8360
    },
    {
      "entropy": 6.339152526855469,
      "epoch": 0.9653779572994806,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004918420122329634,
      "loss": 6.208,
      "mean_token_accuracy": 0.16081032305955886,
      "num_tokens": 21200685.0,
      "step": 8365
    },
    {
      "entropy": 6.348997116088867,
      "epoch": 0.9659549913444894,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004918309978201989,
      "loss": 6.1992,
      "mean_token_accuracy": 0.1665423795580864,
      "num_tokens": 21212709.0,
      "step": 8370
    },
    {
      "entropy": 6.451622152328492,
      "epoch": 0.966532025389498,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004918199761144315,
      "loss": 6.3389,
      "mean_token_accuracy": 0.15161336064338685,
      "num_tokens": 21224846.0,
      "step": 8375
    },
    {
      "entropy": 6.403325319290161,
      "epoch": 0.9671090594345066,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004918089471160318,
      "loss": 6.3308,
      "mean_token_accuracy": 0.16391047835350037,
      "num_tokens": 21238490.0,
      "step": 8380
    },
    {
      "entropy": 6.386717939376831,
      "epoch": 0.9676860934795153,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004917979108253709,
      "loss": 6.2033,
      "mean_token_accuracy": 0.16050710380077363,
      "num_tokens": 21250064.0,
      "step": 8385
    },
    {
      "entropy": 6.421000051498413,
      "epoch": 0.9682631275245239,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00049178686724282,
      "loss": 6.2117,
      "mean_token_accuracy": 0.15899699330329894,
      "num_tokens": 21262777.0,
      "step": 8390
    },
    {
      "entropy": 6.395197439193725,
      "epoch": 0.9688401615695326,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004917758163687506,
      "loss": 6.2687,
      "mean_token_accuracy": 0.16623124480247498,
      "num_tokens": 21274837.0,
      "step": 8395
    },
    {
      "entropy": 6.374613809585571,
      "epoch": 0.9694171956145412,
      "grad_norm": 1.625,
      "learning_rate": 0.0004917647582035341,
      "loss": 6.3607,
      "mean_token_accuracy": 0.15184668079018593,
      "num_tokens": 21287683.0,
      "step": 8400
    },
    {
      "entropy": 6.376201486587524,
      "epoch": 0.96999422965955,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004917536927475428,
      "loss": 6.1818,
      "mean_token_accuracy": 0.16200344264507294,
      "num_tokens": 21300053.0,
      "step": 8405
    },
    {
      "entropy": 6.440302991867066,
      "epoch": 0.9705712637045586,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004917426200011486,
      "loss": 6.2611,
      "mean_token_accuracy": 0.1626146748661995,
      "num_tokens": 21312143.0,
      "step": 8410
    },
    {
      "entropy": 6.382162284851074,
      "epoch": 0.9711482977495672,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000491731539964724,
      "loss": 6.264,
      "mean_token_accuracy": 0.1642318233847618,
      "num_tokens": 21324530.0,
      "step": 8415
    },
    {
      "entropy": 6.407947683334351,
      "epoch": 0.9717253317945759,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004917204526386418,
      "loss": 6.3134,
      "mean_token_accuracy": 0.16119913905858993,
      "num_tokens": 21337766.0,
      "step": 8420
    },
    {
      "entropy": 6.442886304855347,
      "epoch": 0.9723023658395845,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004917093580232748,
      "loss": 6.2729,
      "mean_token_accuracy": 0.16371907740831376,
      "num_tokens": 21350304.0,
      "step": 8425
    },
    {
      "entropy": 6.390522384643555,
      "epoch": 0.9728793998845932,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004916982561189962,
      "loss": 6.2756,
      "mean_token_accuracy": 0.15955062806606293,
      "num_tokens": 21362030.0,
      "step": 8430
    },
    {
      "entropy": 6.373282623291016,
      "epoch": 0.9734564339296018,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004916871469261794,
      "loss": 6.2956,
      "mean_token_accuracy": 0.1568213276565075,
      "num_tokens": 21374279.0,
      "step": 8435
    },
    {
      "entropy": 6.364431238174438,
      "epoch": 0.9740334679746105,
      "grad_norm": 2.609375,
      "learning_rate": 0.000491676030445198,
      "loss": 6.3004,
      "mean_token_accuracy": 0.16820743083953857,
      "num_tokens": 21388220.0,
      "step": 8440
    },
    {
      "entropy": 6.4249077320098875,
      "epoch": 0.9746105020196192,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004916649066764259,
      "loss": 6.249,
      "mean_token_accuracy": 0.16421249657869338,
      "num_tokens": 21400770.0,
      "step": 8445
    },
    {
      "entropy": 6.374565935134887,
      "epoch": 0.9751875360646278,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004916537756202375,
      "loss": 6.2807,
      "mean_token_accuracy": 0.16066452264785766,
      "num_tokens": 21413078.0,
      "step": 8450
    },
    {
      "entropy": 6.362772464752197,
      "epoch": 0.9757645701096365,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004916426372770069,
      "loss": 6.3373,
      "mean_token_accuracy": 0.15732649713754654,
      "num_tokens": 21427285.0,
      "step": 8455
    },
    {
      "entropy": 6.42603907585144,
      "epoch": 0.9763416041546451,
      "grad_norm": 1.625,
      "learning_rate": 0.0004916314916471087,
      "loss": 6.3029,
      "mean_token_accuracy": 0.15744611397385597,
      "num_tokens": 21440659.0,
      "step": 8460
    },
    {
      "entropy": 6.410357141494751,
      "epoch": 0.9769186381996537,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004916203387309179,
      "loss": 6.246,
      "mean_token_accuracy": 0.152059967815876,
      "num_tokens": 21452301.0,
      "step": 8465
    },
    {
      "entropy": 6.278444290161133,
      "epoch": 0.9774956722446624,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004916091785288096,
      "loss": 6.1978,
      "mean_token_accuracy": 0.16501330584287643,
      "num_tokens": 21466716.0,
      "step": 8470
    },
    {
      "entropy": 6.429250001907349,
      "epoch": 0.9780727062896711,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004915980110411593,
      "loss": 6.29,
      "mean_token_accuracy": 0.15631196796894073,
      "num_tokens": 21479314.0,
      "step": 8475
    },
    {
      "entropy": 6.427707147598267,
      "epoch": 0.9786497403346798,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004915868362683425,
      "loss": 6.2077,
      "mean_token_accuracy": 0.16927620470523835,
      "num_tokens": 21491968.0,
      "step": 8480
    },
    {
      "entropy": 6.281297016143799,
      "epoch": 0.9792267743796884,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004915756542107349,
      "loss": 6.298,
      "mean_token_accuracy": 0.15860208868980408,
      "num_tokens": 21505744.0,
      "step": 8485
    },
    {
      "entropy": 6.404678392410278,
      "epoch": 0.979803808424697,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004915644648687127,
      "loss": 6.3305,
      "mean_token_accuracy": 0.15427371114492416,
      "num_tokens": 21520981.0,
      "step": 8490
    },
    {
      "entropy": 6.421433591842652,
      "epoch": 0.9803808424697057,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004915532682426524,
      "loss": 6.2965,
      "mean_token_accuracy": 0.15904518067836762,
      "num_tokens": 21533712.0,
      "step": 8495
    },
    {
      "entropy": 6.336588764190674,
      "epoch": 0.9809578765147143,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004915420643329306,
      "loss": 6.1778,
      "mean_token_accuracy": 0.16007357686758042,
      "num_tokens": 21546875.0,
      "step": 8500
    },
    {
      "entropy": 6.358702325820923,
      "epoch": 0.981534910559723,
      "grad_norm": 1.265625,
      "learning_rate": 0.000491530853139924,
      "loss": 6.2287,
      "mean_token_accuracy": 0.16115213334560394,
      "num_tokens": 21559154.0,
      "step": 8505
    },
    {
      "entropy": 6.469691753387451,
      "epoch": 0.9821119446047317,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004915196346640095,
      "loss": 6.3238,
      "mean_token_accuracy": 0.15400136709213258,
      "num_tokens": 21573164.0,
      "step": 8510
    },
    {
      "entropy": 6.439447975158691,
      "epoch": 0.9826889786497404,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004915084089055648,
      "loss": 6.3306,
      "mean_token_accuracy": 0.14860097840428352,
      "num_tokens": 21585449.0,
      "step": 8515
    },
    {
      "entropy": 6.328643274307251,
      "epoch": 0.983266012694749,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004914971758649673,
      "loss": 6.27,
      "mean_token_accuracy": 0.16568298786878585,
      "num_tokens": 21598120.0,
      "step": 8520
    },
    {
      "entropy": 6.421794176101685,
      "epoch": 0.9838430467397576,
      "grad_norm": 3.203125,
      "learning_rate": 0.0004914859355425948,
      "loss": 6.3032,
      "mean_token_accuracy": 0.15581102222204207,
      "num_tokens": 21609912.0,
      "step": 8525
    },
    {
      "entropy": 6.466002559661865,
      "epoch": 0.9844200807847663,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004914746879388255,
      "loss": 6.2581,
      "mean_token_accuracy": 0.16066803485155107,
      "num_tokens": 21623242.0,
      "step": 8530
    },
    {
      "entropy": 6.35989294052124,
      "epoch": 0.9849971148297749,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004914634330540373,
      "loss": 6.2925,
      "mean_token_accuracy": 0.16082993894815445,
      "num_tokens": 21635527.0,
      "step": 8535
    },
    {
      "entropy": 6.380307149887085,
      "epoch": 0.9855741488747836,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004914521708886093,
      "loss": 6.2392,
      "mean_token_accuracy": 0.16111573278903962,
      "num_tokens": 21648994.0,
      "step": 8540
    },
    {
      "entropy": 6.379744577407837,
      "epoch": 0.9861511829197923,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004914409014429201,
      "loss": 6.2758,
      "mean_token_accuracy": 0.15881554186344146,
      "num_tokens": 21661512.0,
      "step": 8545
    },
    {
      "entropy": 6.381563758850097,
      "epoch": 0.986728216964801,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004914296247173486,
      "loss": 6.3171,
      "mean_token_accuracy": 0.15678785592317582,
      "num_tokens": 21674900.0,
      "step": 8550
    },
    {
      "entropy": 6.508550548553467,
      "epoch": 0.9873052510098096,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004914183407122741,
      "loss": 6.5073,
      "mean_token_accuracy": 0.1529560297727585,
      "num_tokens": 21688451.0,
      "step": 8555
    },
    {
      "entropy": 6.460909605026245,
      "epoch": 0.9878822850548182,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004914070494280763,
      "loss": 6.2937,
      "mean_token_accuracy": 0.15463425517082213,
      "num_tokens": 21700350.0,
      "step": 8560
    },
    {
      "entropy": 6.382816743850708,
      "epoch": 0.9884593190998269,
      "grad_norm": 2.421875,
      "learning_rate": 0.0004913957508651349,
      "loss": 6.2653,
      "mean_token_accuracy": 0.16228749752044677,
      "num_tokens": 21713211.0,
      "step": 8565
    },
    {
      "entropy": 6.36720757484436,
      "epoch": 0.9890363531448355,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004913844450238299,
      "loss": 6.2327,
      "mean_token_accuracy": 0.16315484046936035,
      "num_tokens": 21726105.0,
      "step": 8570
    },
    {
      "entropy": 6.412167978286743,
      "epoch": 0.9896133871898442,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004913731319045416,
      "loss": 6.2817,
      "mean_token_accuracy": 0.16310435831546782,
      "num_tokens": 21738075.0,
      "step": 8575
    },
    {
      "entropy": 6.411454057693481,
      "epoch": 0.9901904212348529,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004913618115076505,
      "loss": 6.2542,
      "mean_token_accuracy": 0.16109486371278764,
      "num_tokens": 21750614.0,
      "step": 8580
    },
    {
      "entropy": 6.256771802902222,
      "epoch": 0.9907674552798615,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004913504838335372,
      "loss": 6.2713,
      "mean_token_accuracy": 0.1610737442970276,
      "num_tokens": 21763670.0,
      "step": 8585
    },
    {
      "entropy": 6.344682216644287,
      "epoch": 0.9913444893248702,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004913391488825829,
      "loss": 6.238,
      "mean_token_accuracy": 0.16902547478675842,
      "num_tokens": 21776623.0,
      "step": 8590
    },
    {
      "entropy": 6.429253101348877,
      "epoch": 0.9919215233698788,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004913278066551689,
      "loss": 6.2113,
      "mean_token_accuracy": 0.1598364844918251,
      "num_tokens": 21787903.0,
      "step": 8595
    },
    {
      "entropy": 6.308300876617432,
      "epoch": 0.9924985574148875,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004913164571516765,
      "loss": 6.1519,
      "mean_token_accuracy": 0.1660546824336052,
      "num_tokens": 21800409.0,
      "step": 8600
    },
    {
      "entropy": 6.34729495048523,
      "epoch": 0.9930755914598961,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004913051003724876,
      "loss": 6.3579,
      "mean_token_accuracy": 0.15751274973154067,
      "num_tokens": 21812741.0,
      "step": 8605
    },
    {
      "entropy": 6.407994890213013,
      "epoch": 0.9936526255049047,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004912937363179839,
      "loss": 6.2875,
      "mean_token_accuracy": 0.16249522119760512,
      "num_tokens": 21826180.0,
      "step": 8610
    },
    {
      "entropy": 6.504053926467895,
      "epoch": 0.9942296595499135,
      "grad_norm": 1.5,
      "learning_rate": 0.000491282364988548,
      "loss": 6.3495,
      "mean_token_accuracy": 0.15898843556642533,
      "num_tokens": 21838961.0,
      "step": 8615
    },
    {
      "entropy": 6.395296335220337,
      "epoch": 0.9948066935949221,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004912709863845621,
      "loss": 6.2812,
      "mean_token_accuracy": 0.16122212558984755,
      "num_tokens": 21850942.0,
      "step": 8620
    },
    {
      "entropy": 6.361412048339844,
      "epoch": 0.9953837276399308,
      "grad_norm": 1.328125,
      "learning_rate": 0.000491259600506409,
      "loss": 6.1996,
      "mean_token_accuracy": 0.16818103045225144,
      "num_tokens": 21862666.0,
      "step": 8625
    },
    {
      "entropy": 6.396421813964844,
      "epoch": 0.9959607616849394,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004912482073544716,
      "loss": 6.2305,
      "mean_token_accuracy": 0.16952535808086394,
      "num_tokens": 21874866.0,
      "step": 8630
    },
    {
      "entropy": 6.403103637695312,
      "epoch": 0.996537795729948,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004912368069291333,
      "loss": 6.3302,
      "mean_token_accuracy": 0.1602864071726799,
      "num_tokens": 21887886.0,
      "step": 8635
    },
    {
      "entropy": 6.4391790390014645,
      "epoch": 0.9971148297749567,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004912253992307773,
      "loss": 6.2807,
      "mean_token_accuracy": 0.15628421008586885,
      "num_tokens": 21901096.0,
      "step": 8640
    },
    {
      "entropy": 6.439533281326294,
      "epoch": 0.9976918638199653,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004912139842597875,
      "loss": 6.3617,
      "mean_token_accuracy": 0.16195413023233413,
      "num_tokens": 21914018.0,
      "step": 8645
    },
    {
      "entropy": 6.36651029586792,
      "epoch": 0.9982688978649741,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004912025620165477,
      "loss": 6.2606,
      "mean_token_accuracy": 0.16470131576061248,
      "num_tokens": 21925488.0,
      "step": 8650
    },
    {
      "entropy": 6.478236961364746,
      "epoch": 0.9988459319099827,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004911911325014421,
      "loss": 6.3725,
      "mean_token_accuracy": 0.14617064893245696,
      "num_tokens": 21938268.0,
      "step": 8655
    },
    {
      "entropy": 6.4387367248535154,
      "epoch": 0.9994229659549914,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004911796957148552,
      "loss": 6.3445,
      "mean_token_accuracy": 0.16030535250902175,
      "num_tokens": 21951219.0,
      "step": 8660
    },
    {
      "entropy": 6.430270862579346,
      "epoch": 1.0,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004911682516571715,
      "loss": 6.2959,
      "mean_token_accuracy": 0.15840745270252227,
      "num_tokens": 21963117.0,
      "step": 8665
    },
    {
      "entropy": 6.426940965652466,
      "epoch": 1.0005770340450086,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004911568003287761,
      "loss": 6.2376,
      "mean_token_accuracy": 0.1605121448636055,
      "num_tokens": 21974187.0,
      "step": 8670
    },
    {
      "entropy": 6.415137052536011,
      "epoch": 1.0011540680900173,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004911453417300541,
      "loss": 6.2486,
      "mean_token_accuracy": 0.16247110962867736,
      "num_tokens": 21987393.0,
      "step": 8675
    },
    {
      "entropy": 6.38947024345398,
      "epoch": 1.001731102135026,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004911338758613909,
      "loss": 6.1214,
      "mean_token_accuracy": 0.16808023154735566,
      "num_tokens": 21998638.0,
      "step": 8680
    },
    {
      "entropy": 6.340237426757812,
      "epoch": 1.0023081361800346,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004911224027231722,
      "loss": 6.1783,
      "mean_token_accuracy": 0.16497329473495484,
      "num_tokens": 22012231.0,
      "step": 8685
    },
    {
      "entropy": 6.314101123809815,
      "epoch": 1.0028851702250432,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004911109223157838,
      "loss": 6.0206,
      "mean_token_accuracy": 0.1724448785185814,
      "num_tokens": 22024929.0,
      "step": 8690
    },
    {
      "entropy": 6.372276830673218,
      "epoch": 1.0034622042700518,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004910994346396118,
      "loss": 6.1297,
      "mean_token_accuracy": 0.16675959825515746,
      "num_tokens": 22037873.0,
      "step": 8695
    },
    {
      "entropy": 6.307498407363892,
      "epoch": 1.0040392383150605,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004910879396950427,
      "loss": 6.1082,
      "mean_token_accuracy": 0.16449054181575776,
      "num_tokens": 22049572.0,
      "step": 8700
    },
    {
      "entropy": 6.38808240890503,
      "epoch": 1.0046162723600693,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000491076437482463,
      "loss": 6.2122,
      "mean_token_accuracy": 0.15654898583889007,
      "num_tokens": 22063290.0,
      "step": 8705
    },
    {
      "entropy": 6.271512460708618,
      "epoch": 1.005193306405078,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004910649280022599,
      "loss": 6.0864,
      "mean_token_accuracy": 0.1705736681818962,
      "num_tokens": 22076387.0,
      "step": 8710
    },
    {
      "entropy": 6.347558546066284,
      "epoch": 1.0057703404500866,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004910534112548201,
      "loss": 6.1711,
      "mean_token_accuracy": 0.16337237358093262,
      "num_tokens": 22088780.0,
      "step": 8715
    },
    {
      "entropy": 6.276888036727906,
      "epoch": 1.0063473744950953,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004910418872405312,
      "loss": 6.0633,
      "mean_token_accuracy": 0.1731952100992203,
      "num_tokens": 22102150.0,
      "step": 8720
    },
    {
      "entropy": 6.3603862762451175,
      "epoch": 1.006924408540104,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004910303559597806,
      "loss": 6.1491,
      "mean_token_accuracy": 0.16267468482255937,
      "num_tokens": 22115250.0,
      "step": 8725
    },
    {
      "entropy": 6.325836277008056,
      "epoch": 1.0075014425851125,
      "grad_norm": 2.125,
      "learning_rate": 0.0004910188174129564,
      "loss": 6.1948,
      "mean_token_accuracy": 0.1561719983816147,
      "num_tokens": 22129015.0,
      "step": 8730
    },
    {
      "entropy": 6.2975523471832275,
      "epoch": 1.0080784766301212,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004910072716004466,
      "loss": 6.1954,
      "mean_token_accuracy": 0.16403840184211732,
      "num_tokens": 22141716.0,
      "step": 8735
    },
    {
      "entropy": 6.385508728027344,
      "epoch": 1.0086555106751298,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004909957185226394,
      "loss": 6.2501,
      "mean_token_accuracy": 0.1517787382006645,
      "num_tokens": 22153568.0,
      "step": 8740
    },
    {
      "entropy": 6.447339200973511,
      "epoch": 1.0092325447201385,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004909841581799236,
      "loss": 6.2239,
      "mean_token_accuracy": 0.1596845343708992,
      "num_tokens": 22165443.0,
      "step": 8745
    },
    {
      "entropy": 6.389078998565674,
      "epoch": 1.009809578765147,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004909725905726879,
      "loss": 6.1835,
      "mean_token_accuracy": 0.1615777239203453,
      "num_tokens": 22179068.0,
      "step": 8750
    },
    {
      "entropy": 6.350357580184936,
      "epoch": 1.0103866128101557,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004909610157013214,
      "loss": 6.2675,
      "mean_token_accuracy": 0.15202371701598166,
      "num_tokens": 22191597.0,
      "step": 8755
    },
    {
      "entropy": 6.410714101791382,
      "epoch": 1.0109636468551644,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004909494335662134,
      "loss": 6.1896,
      "mean_token_accuracy": 0.16322282552719117,
      "num_tokens": 22203724.0,
      "step": 8760
    },
    {
      "entropy": 6.303729724884033,
      "epoch": 1.011540680900173,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004909378441677535,
      "loss": 6.1458,
      "mean_token_accuracy": 0.16659850329160691,
      "num_tokens": 22216100.0,
      "step": 8765
    },
    {
      "entropy": 6.359002780914307,
      "epoch": 1.0121177149451817,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004909262475063314,
      "loss": 6.1722,
      "mean_token_accuracy": 0.15876885056495665,
      "num_tokens": 22228162.0,
      "step": 8770
    },
    {
      "entropy": 6.332960414886474,
      "epoch": 1.0126947489901905,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004909146435823373,
      "loss": 6.157,
      "mean_token_accuracy": 0.16825098544359207,
      "num_tokens": 22239756.0,
      "step": 8775
    },
    {
      "entropy": 6.333130693435669,
      "epoch": 1.0132717830351992,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004909030323961613,
      "loss": 6.1633,
      "mean_token_accuracy": 0.15986109226942063,
      "num_tokens": 22252781.0,
      "step": 8780
    },
    {
      "entropy": 6.3398871421813965,
      "epoch": 1.0138488170802078,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004908914139481942,
      "loss": 6.2258,
      "mean_token_accuracy": 0.15804215222597123,
      "num_tokens": 22265171.0,
      "step": 8785
    },
    {
      "entropy": 6.337258434295654,
      "epoch": 1.0144258511252164,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004908797882388265,
      "loss": 6.1805,
      "mean_token_accuracy": 0.16364909410476686,
      "num_tokens": 22278121.0,
      "step": 8790
    },
    {
      "entropy": 6.275689315795899,
      "epoch": 1.015002885170225,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004908681552684495,
      "loss": 6.0378,
      "mean_token_accuracy": 0.166185662150383,
      "num_tokens": 22290074.0,
      "step": 8795
    },
    {
      "entropy": 6.295356369018554,
      "epoch": 1.0155799192152337,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004908565150374542,
      "loss": 6.1924,
      "mean_token_accuracy": 0.16460571438074112,
      "num_tokens": 22303088.0,
      "step": 8800
    },
    {
      "entropy": 6.413890361785889,
      "epoch": 1.0161569532602424,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004908448675462323,
      "loss": 6.2355,
      "mean_token_accuracy": 0.1506106361746788,
      "num_tokens": 22317693.0,
      "step": 8805
    },
    {
      "entropy": 6.359114170074463,
      "epoch": 1.016733987305251,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004908332127951756,
      "loss": 6.1808,
      "mean_token_accuracy": 0.1655200481414795,
      "num_tokens": 22330876.0,
      "step": 8810
    },
    {
      "entropy": 6.337749481201172,
      "epoch": 1.0173110213502596,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004908215507846757,
      "loss": 6.1308,
      "mean_token_accuracy": 0.1639953449368477,
      "num_tokens": 22342620.0,
      "step": 8815
    },
    {
      "entropy": 6.328289365768432,
      "epoch": 1.0178880553952683,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004908098815151254,
      "loss": 6.1631,
      "mean_token_accuracy": 0.1632054179906845,
      "num_tokens": 22353733.0,
      "step": 8820
    },
    {
      "entropy": 6.461244201660156,
      "epoch": 1.018465089440277,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004907982049869168,
      "loss": 6.2178,
      "mean_token_accuracy": 0.15410459488630296,
      "num_tokens": 22366857.0,
      "step": 8825
    },
    {
      "entropy": 6.299141931533813,
      "epoch": 1.0190421234852856,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004907865212004428,
      "loss": 6.0471,
      "mean_token_accuracy": 0.16742806285619735,
      "num_tokens": 22377658.0,
      "step": 8830
    },
    {
      "entropy": 6.279093837738037,
      "epoch": 1.0196191575302942,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004907748301560963,
      "loss": 6.206,
      "mean_token_accuracy": 0.15780699700117112,
      "num_tokens": 22390060.0,
      "step": 8835
    },
    {
      "entropy": 6.34080319404602,
      "epoch": 1.0201961915753028,
      "grad_norm": 2.0,
      "learning_rate": 0.0004907631318542707,
      "loss": 6.1819,
      "mean_token_accuracy": 0.1587137073278427,
      "num_tokens": 22402903.0,
      "step": 8840
    },
    {
      "entropy": 6.349684619903565,
      "epoch": 1.0207732256203117,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004907514262953594,
      "loss": 6.1471,
      "mean_token_accuracy": 0.16755905002355576,
      "num_tokens": 22416036.0,
      "step": 8845
    },
    {
      "entropy": 6.399058294296265,
      "epoch": 1.0213502596653203,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004907397134797559,
      "loss": 6.16,
      "mean_token_accuracy": 0.16025111079216003,
      "num_tokens": 22428000.0,
      "step": 8850
    },
    {
      "entropy": 6.285398197174072,
      "epoch": 1.021927293710329,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004907279934078543,
      "loss": 6.1808,
      "mean_token_accuracy": 0.16548313647508622,
      "num_tokens": 22440029.0,
      "step": 8855
    },
    {
      "entropy": 6.307600975036621,
      "epoch": 1.0225043277553376,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004907162660800488,
      "loss": 6.2091,
      "mean_token_accuracy": 0.1566235363483429,
      "num_tokens": 22452906.0,
      "step": 8860
    },
    {
      "entropy": 6.368782329559326,
      "epoch": 1.0230813618003463,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004907045314967338,
      "loss": 6.0737,
      "mean_token_accuracy": 0.17112531810998916,
      "num_tokens": 22465853.0,
      "step": 8865
    },
    {
      "entropy": 6.359477853775024,
      "epoch": 1.023658395845355,
      "grad_norm": 4.71875,
      "learning_rate": 0.0004906927896583041,
      "loss": 6.1674,
      "mean_token_accuracy": 0.1638613685965538,
      "num_tokens": 22478583.0,
      "step": 8870
    },
    {
      "entropy": 6.2635619163513185,
      "epoch": 1.0242354298903635,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004906810405651546,
      "loss": 6.185,
      "mean_token_accuracy": 0.16508891731500625,
      "num_tokens": 22491665.0,
      "step": 8875
    },
    {
      "entropy": 6.332400894165039,
      "epoch": 1.0248124639353722,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004906692842176803,
      "loss": 6.1711,
      "mean_token_accuracy": 0.16148089468479157,
      "num_tokens": 22504015.0,
      "step": 8880
    },
    {
      "entropy": 6.441442584991455,
      "epoch": 1.0253894979803808,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004906575206162769,
      "loss": 6.2513,
      "mean_token_accuracy": 0.15498681217432023,
      "num_tokens": 22516385.0,
      "step": 8885
    },
    {
      "entropy": 6.370885753631592,
      "epoch": 1.0259665320253895,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004906457497613399,
      "loss": 6.1514,
      "mean_token_accuracy": 0.16406670212745667,
      "num_tokens": 22527988.0,
      "step": 8890
    },
    {
      "entropy": 6.2743278503417965,
      "epoch": 1.026543566070398,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004906339716532651,
      "loss": 6.1883,
      "mean_token_accuracy": 0.16317007690668106,
      "num_tokens": 22540613.0,
      "step": 8895
    },
    {
      "entropy": 6.298226737976075,
      "epoch": 1.0271206001154067,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004906221862924488,
      "loss": 6.1085,
      "mean_token_accuracy": 0.16695300936698915,
      "num_tokens": 22552137.0,
      "step": 8900
    },
    {
      "entropy": 6.369357299804688,
      "epoch": 1.0276976341604154,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004906103936792875,
      "loss": 6.2491,
      "mean_token_accuracy": 0.15076114386320114,
      "num_tokens": 22564677.0,
      "step": 8905
    },
    {
      "entropy": 6.419727420806884,
      "epoch": 1.028274668205424,
      "grad_norm": 2.515625,
      "learning_rate": 0.0004905985938141777,
      "loss": 6.1798,
      "mean_token_accuracy": 0.1652047887444496,
      "num_tokens": 22578236.0,
      "step": 8910
    },
    {
      "entropy": 6.350459432601928,
      "epoch": 1.0288517022504329,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004905867866975163,
      "loss": 6.1683,
      "mean_token_accuracy": 0.16516325920820235,
      "num_tokens": 22592533.0,
      "step": 8915
    },
    {
      "entropy": 6.358260679244995,
      "epoch": 1.0294287362954415,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004905749723297003,
      "loss": 6.2329,
      "mean_token_accuracy": 0.15685783475637435,
      "num_tokens": 22606062.0,
      "step": 8920
    },
    {
      "entropy": 6.399117088317871,
      "epoch": 1.0300057703404502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004905631507111271,
      "loss": 6.2389,
      "mean_token_accuracy": 0.15823203176259995,
      "num_tokens": 22619342.0,
      "step": 8925
    },
    {
      "entropy": 6.342332744598389,
      "epoch": 1.0305828043854588,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004905513218421946,
      "loss": 6.1936,
      "mean_token_accuracy": 0.16186433136463166,
      "num_tokens": 22632003.0,
      "step": 8930
    },
    {
      "entropy": 6.3654186725616455,
      "epoch": 1.0311598384304674,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004905394857233004,
      "loss": 6.1736,
      "mean_token_accuracy": 0.16698863506317138,
      "num_tokens": 22644179.0,
      "step": 8935
    },
    {
      "entropy": 6.423052978515625,
      "epoch": 1.031736872475476,
      "grad_norm": 1.875,
      "learning_rate": 0.0004905276423548426,
      "loss": 6.2233,
      "mean_token_accuracy": 0.16120436564087867,
      "num_tokens": 22657544.0,
      "step": 8940
    },
    {
      "entropy": 6.401259517669677,
      "epoch": 1.0323139065204847,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004905157917372197,
      "loss": 6.2665,
      "mean_token_accuracy": 0.15423907488584518,
      "num_tokens": 22670789.0,
      "step": 8945
    },
    {
      "entropy": 6.326565361022949,
      "epoch": 1.0328909405654934,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004905039338708302,
      "loss": 6.1676,
      "mean_token_accuracy": 0.15901732444763184,
      "num_tokens": 22682491.0,
      "step": 8950
    },
    {
      "entropy": 6.43274712562561,
      "epoch": 1.033467974610502,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004904920687560728,
      "loss": 6.2071,
      "mean_token_accuracy": 0.16146688759326935,
      "num_tokens": 22695235.0,
      "step": 8955
    },
    {
      "entropy": 6.322785186767578,
      "epoch": 1.0340450086555106,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004904801963933469,
      "loss": 6.2042,
      "mean_token_accuracy": 0.16150673478841782,
      "num_tokens": 22707222.0,
      "step": 8960
    },
    {
      "entropy": 6.356104993820191,
      "epoch": 1.0346220427005193,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004904683167830515,
      "loss": 6.2446,
      "mean_token_accuracy": 0.1534208372235298,
      "num_tokens": 22718820.0,
      "step": 8965
    },
    {
      "entropy": 6.398207187652588,
      "epoch": 1.035199076745528,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004904564299255862,
      "loss": 6.1547,
      "mean_token_accuracy": 0.16409973949193954,
      "num_tokens": 22731265.0,
      "step": 8970
    },
    {
      "entropy": 6.334744119644165,
      "epoch": 1.0357761107905366,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004904445358213511,
      "loss": 6.139,
      "mean_token_accuracy": 0.1598776698112488,
      "num_tokens": 22744222.0,
      "step": 8975
    },
    {
      "entropy": 6.229028511047363,
      "epoch": 1.0363531448355452,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004904326344707461,
      "loss": 6.1449,
      "mean_token_accuracy": 0.1658041372895241,
      "num_tokens": 22756190.0,
      "step": 8980
    },
    {
      "entropy": 6.297280120849609,
      "epoch": 1.036930178880554,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004904207258741712,
      "loss": 6.084,
      "mean_token_accuracy": 0.1679148942232132,
      "num_tokens": 22767943.0,
      "step": 8985
    },
    {
      "entropy": 6.369158220291138,
      "epoch": 1.0375072129255627,
      "grad_norm": 1.5,
      "learning_rate": 0.0004904088100320274,
      "loss": 6.1554,
      "mean_token_accuracy": 0.1659896582365036,
      "num_tokens": 22780972.0,
      "step": 8990
    },
    {
      "entropy": 6.400546598434448,
      "epoch": 1.0380842469705713,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004903968869447151,
      "loss": 6.2692,
      "mean_token_accuracy": 0.1603499099612236,
      "num_tokens": 22792584.0,
      "step": 8995
    },
    {
      "entropy": 6.354660606384277,
      "epoch": 1.03866128101558,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004903849566126356,
      "loss": 6.1999,
      "mean_token_accuracy": 0.16254472136497497,
      "num_tokens": 22807385.0,
      "step": 9000
    },
    {
      "epoch": 1.03866128101558,
      "eval_entropy": 6.251975536007484,
      "eval_loss": 6.3294453620910645,
      "eval_mean_token_accuracy": 0.16032058560999918,
      "eval_num_tokens": 22807385.0,
      "eval_runtime": 22.7515,
      "eval_samples_per_second": 1236.663,
      "eval_steps_per_second": 154.583,
      "step": 9000
    },
    {
      "entropy": 6.375589609146118,
      "epoch": 1.0392383150605886,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903730190361902,
      "loss": 6.2186,
      "mean_token_accuracy": 0.1562452420592308,
      "num_tokens": 22820655.0,
      "step": 9005
    },
    {
      "entropy": 6.387510585784912,
      "epoch": 1.0398153491055973,
      "grad_norm": 1.609375,
      "learning_rate": 0.00049036107421578,
      "loss": 6.1143,
      "mean_token_accuracy": 0.16923735439777374,
      "num_tokens": 22833043.0,
      "step": 9010
    },
    {
      "entropy": 6.322872257232666,
      "epoch": 1.040392383150606,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004903491221518073,
      "loss": 6.2247,
      "mean_token_accuracy": 0.16158089339733123,
      "num_tokens": 22844414.0,
      "step": 9015
    },
    {
      "entropy": 6.334288787841797,
      "epoch": 1.0409694171956145,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004903371628446737,
      "loss": 6.121,
      "mean_token_accuracy": 0.16435754895210267,
      "num_tokens": 22855966.0,
      "step": 9020
    },
    {
      "entropy": 6.290288877487183,
      "epoch": 1.0415464512406232,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903251962947817,
      "loss": 6.1463,
      "mean_token_accuracy": 0.16436477303504943,
      "num_tokens": 22869300.0,
      "step": 9025
    },
    {
      "entropy": 6.392134428024292,
      "epoch": 1.0421234852856318,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004903132225025335,
      "loss": 6.1891,
      "mean_token_accuracy": 0.1661277323961258,
      "num_tokens": 22881040.0,
      "step": 9030
    },
    {
      "entropy": 6.345999336242675,
      "epoch": 1.0427005193306405,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004903012414683322,
      "loss": 6.1113,
      "mean_token_accuracy": 0.1675960272550583,
      "num_tokens": 22892367.0,
      "step": 9035
    },
    {
      "entropy": 6.356339693069458,
      "epoch": 1.043277553375649,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004902892531925805,
      "loss": 6.2527,
      "mean_token_accuracy": 0.1567448541522026,
      "num_tokens": 22904737.0,
      "step": 9040
    },
    {
      "entropy": 6.377230978012085,
      "epoch": 1.0438545874206577,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004902772576756818,
      "loss": 6.2375,
      "mean_token_accuracy": 0.15742040425539017,
      "num_tokens": 22917775.0,
      "step": 9045
    },
    {
      "entropy": 6.351797676086425,
      "epoch": 1.0444316214656664,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004902652549180394,
      "loss": 6.2575,
      "mean_token_accuracy": 0.15253747701644899,
      "num_tokens": 22930496.0,
      "step": 9050
    },
    {
      "entropy": 6.380230855941773,
      "epoch": 1.0450086555106752,
      "grad_norm": 1.75,
      "learning_rate": 0.0004902532449200571,
      "loss": 6.2036,
      "mean_token_accuracy": 0.15602196007966995,
      "num_tokens": 22943787.0,
      "step": 9055
    },
    {
      "entropy": 6.360061550140381,
      "epoch": 1.0455856895556839,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004902412276821386,
      "loss": 6.1759,
      "mean_token_accuracy": 0.16011980623006822,
      "num_tokens": 22956399.0,
      "step": 9060
    },
    {
      "entropy": 6.331561517715454,
      "epoch": 1.0461627236006925,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004902292032046887,
      "loss": 6.2718,
      "mean_token_accuracy": 0.15832901746034622,
      "num_tokens": 22969993.0,
      "step": 9065
    },
    {
      "entropy": 6.431485795974732,
      "epoch": 1.0467397576457012,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004902171714881112,
      "loss": 6.2884,
      "mean_token_accuracy": 0.15681524723768234,
      "num_tokens": 22983807.0,
      "step": 9070
    },
    {
      "entropy": 6.39180817604065,
      "epoch": 1.0473167916907098,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004902051325328112,
      "loss": 6.235,
      "mean_token_accuracy": 0.1634105086326599,
      "num_tokens": 22996238.0,
      "step": 9075
    },
    {
      "entropy": 6.372631359100342,
      "epoch": 1.0478938257357184,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004901930863391934,
      "loss": 6.2263,
      "mean_token_accuracy": 0.1627142459154129,
      "num_tokens": 23009084.0,
      "step": 9080
    },
    {
      "entropy": 6.471501731872559,
      "epoch": 1.048470859780727,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000490181032907663,
      "loss": 6.289,
      "mean_token_accuracy": 0.15503096878528594,
      "num_tokens": 23022799.0,
      "step": 9085
    },
    {
      "entropy": 6.345464706420898,
      "epoch": 1.0490478938257357,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004901689722386255,
      "loss": 6.224,
      "mean_token_accuracy": 0.1631155714392662,
      "num_tokens": 23035566.0,
      "step": 9090
    },
    {
      "entropy": 6.2709979057312015,
      "epoch": 1.0496249278707444,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004901569043324864,
      "loss": 6.1007,
      "mean_token_accuracy": 0.17416824996471406,
      "num_tokens": 23050156.0,
      "step": 9095
    },
    {
      "entropy": 6.332347631454468,
      "epoch": 1.050201961915753,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004901448291896518,
      "loss": 6.2069,
      "mean_token_accuracy": 0.16119893938302993,
      "num_tokens": 23063121.0,
      "step": 9100
    },
    {
      "entropy": 6.363145351409912,
      "epoch": 1.0507789959607616,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004901327468105277,
      "loss": 6.0885,
      "mean_token_accuracy": 0.1751132309436798,
      "num_tokens": 23075236.0,
      "step": 9105
    },
    {
      "entropy": 6.300115585327148,
      "epoch": 1.0513560300057703,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004901206571955205,
      "loss": 6.2693,
      "mean_token_accuracy": 0.16072097718715667,
      "num_tokens": 23089091.0,
      "step": 9110
    },
    {
      "entropy": 6.385873746871948,
      "epoch": 1.051933064050779,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004901085603450369,
      "loss": 6.2176,
      "mean_token_accuracy": 0.15178074762225152,
      "num_tokens": 23101642.0,
      "step": 9115
    },
    {
      "entropy": 6.3707091331481935,
      "epoch": 1.0525100980957875,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004900964562594838,
      "loss": 6.1347,
      "mean_token_accuracy": 0.16434525996446608,
      "num_tokens": 23113427.0,
      "step": 9120
    },
    {
      "entropy": 6.324672508239746,
      "epoch": 1.0530871321407964,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000490084344939268,
      "loss": 6.1162,
      "mean_token_accuracy": 0.16822165101766587,
      "num_tokens": 23125857.0,
      "step": 9125
    },
    {
      "entropy": 6.295746231079102,
      "epoch": 1.053664166185805,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004900722263847973,
      "loss": 6.1563,
      "mean_token_accuracy": 0.16180892288684845,
      "num_tokens": 23138541.0,
      "step": 9130
    },
    {
      "entropy": 6.283871793746949,
      "epoch": 1.0542412002308137,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004900601005964791,
      "loss": 6.1741,
      "mean_token_accuracy": 0.1658548966050148,
      "num_tokens": 23151155.0,
      "step": 9135
    },
    {
      "entropy": 6.381706619262696,
      "epoch": 1.0548182342758223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004900479675747212,
      "loss": 6.179,
      "mean_token_accuracy": 0.15879111289978026,
      "num_tokens": 23165478.0,
      "step": 9140
    },
    {
      "entropy": 6.3661736965179445,
      "epoch": 1.055395268320831,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004900358273199316,
      "loss": 6.1881,
      "mean_token_accuracy": 0.15336017161607743,
      "num_tokens": 23178120.0,
      "step": 9145
    },
    {
      "entropy": 6.3876782894134525,
      "epoch": 1.0559723023658396,
      "grad_norm": 1.359375,
      "learning_rate": 0.000490023679832519,
      "loss": 6.2258,
      "mean_token_accuracy": 0.15789849311113358,
      "num_tokens": 23191275.0,
      "step": 9150
    },
    {
      "entropy": 6.295345401763916,
      "epoch": 1.0565493364108482,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004900115251128916,
      "loss": 6.192,
      "mean_token_accuracy": 0.15735087543725967,
      "num_tokens": 23204518.0,
      "step": 9155
    },
    {
      "entropy": 6.415962743759155,
      "epoch": 1.057126370455857,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004899993631614583,
      "loss": 6.2385,
      "mean_token_accuracy": 0.15383946001529694,
      "num_tokens": 23217488.0,
      "step": 9160
    },
    {
      "entropy": 6.418709421157837,
      "epoch": 1.0577034045008655,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004899871939786283,
      "loss": 6.2066,
      "mean_token_accuracy": 0.16239057183265687,
      "num_tokens": 23230101.0,
      "step": 9165
    },
    {
      "entropy": 6.295369005203247,
      "epoch": 1.0582804385458742,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004899750175648107,
      "loss": 6.34,
      "mean_token_accuracy": 0.15164581537246705,
      "num_tokens": 23243747.0,
      "step": 9170
    },
    {
      "entropy": 6.400159597396851,
      "epoch": 1.0588574725908828,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004899628339204154,
      "loss": 6.1433,
      "mean_token_accuracy": 0.1617654398083687,
      "num_tokens": 23256593.0,
      "step": 9175
    },
    {
      "entropy": 6.389183235168457,
      "epoch": 1.0594345066358914,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004899506430458518,
      "loss": 6.2726,
      "mean_token_accuracy": 0.15196809992194177,
      "num_tokens": 23269359.0,
      "step": 9180
    },
    {
      "entropy": 6.304968214035034,
      "epoch": 1.0600115406809,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004899384449415302,
      "loss": 6.1119,
      "mean_token_accuracy": 0.16750582456588745,
      "num_tokens": 23283462.0,
      "step": 9185
    },
    {
      "entropy": 6.301669359207153,
      "epoch": 1.0605885747259087,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004899262396078606,
      "loss": 6.2079,
      "mean_token_accuracy": 0.16235196888446807,
      "num_tokens": 23295920.0,
      "step": 9190
    },
    {
      "entropy": 6.443309926986695,
      "epoch": 1.0611656087709176,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004899140270452539,
      "loss": 6.1902,
      "mean_token_accuracy": 0.16026893705129625,
      "num_tokens": 23307433.0,
      "step": 9195
    },
    {
      "entropy": 6.360797882080078,
      "epoch": 1.0617426428159262,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004899018072541204,
      "loss": 6.1686,
      "mean_token_accuracy": 0.1604166179895401,
      "num_tokens": 23320265.0,
      "step": 9200
    },
    {
      "entropy": 6.2488456726074215,
      "epoch": 1.0623196768609349,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004898895802348715,
      "loss": 6.0442,
      "mean_token_accuracy": 0.16586936563253402,
      "num_tokens": 23331951.0,
      "step": 9205
    },
    {
      "entropy": 6.276108264923096,
      "epoch": 1.0628967109059435,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004898773459879183,
      "loss": 6.2023,
      "mean_token_accuracy": 0.15901616886258124,
      "num_tokens": 23344862.0,
      "step": 9210
    },
    {
      "entropy": 6.370738840103149,
      "epoch": 1.0634737449509521,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004898651045136724,
      "loss": 6.1776,
      "mean_token_accuracy": 0.15551743656396866,
      "num_tokens": 23357212.0,
      "step": 9215
    },
    {
      "entropy": 6.334261131286621,
      "epoch": 1.0640507789959608,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004898528558125453,
      "loss": 6.1547,
      "mean_token_accuracy": 0.16699836254119874,
      "num_tokens": 23370114.0,
      "step": 9220
    },
    {
      "entropy": 6.3864648818969725,
      "epoch": 1.0646278130409694,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004898405998849492,
      "loss": 6.1723,
      "mean_token_accuracy": 0.15924877747893335,
      "num_tokens": 23382295.0,
      "step": 9225
    },
    {
      "entropy": 6.2677223682403564,
      "epoch": 1.065204847085978,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004898283367312962,
      "loss": 6.1904,
      "mean_token_accuracy": 0.1685797780752182,
      "num_tokens": 23395477.0,
      "step": 9230
    },
    {
      "entropy": 6.282564926147461,
      "epoch": 1.0657818811309867,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004898160663519988,
      "loss": 6.1465,
      "mean_token_accuracy": 0.16062789410352707,
      "num_tokens": 23407868.0,
      "step": 9235
    },
    {
      "entropy": 6.367104148864746,
      "epoch": 1.0663589151759953,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004898037887474697,
      "loss": 6.2365,
      "mean_token_accuracy": 0.1604292392730713,
      "num_tokens": 23422156.0,
      "step": 9240
    },
    {
      "entropy": 6.323754930496216,
      "epoch": 1.066935949221004,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004897915039181219,
      "loss": 6.1416,
      "mean_token_accuracy": 0.1669705554842949,
      "num_tokens": 23433951.0,
      "step": 9245
    },
    {
      "entropy": 6.293605089187622,
      "epoch": 1.0675129832660126,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004897792118643685,
      "loss": 6.1905,
      "mean_token_accuracy": 0.16886758506298066,
      "num_tokens": 23447174.0,
      "step": 9250
    },
    {
      "entropy": 6.372821855545044,
      "epoch": 1.0680900173110213,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004897669125866231,
      "loss": 6.1598,
      "mean_token_accuracy": 0.16654045432806014,
      "num_tokens": 23458502.0,
      "step": 9255
    },
    {
      "entropy": 6.351552820205688,
      "epoch": 1.06866705135603,
      "grad_norm": 1.25,
      "learning_rate": 0.0004897546060852993,
      "loss": 6.2567,
      "mean_token_accuracy": 0.15953566431999205,
      "num_tokens": 23471572.0,
      "step": 9260
    },
    {
      "entropy": 6.332253694534302,
      "epoch": 1.0692440854010385,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004897422923608108,
      "loss": 6.1396,
      "mean_token_accuracy": 0.16287099570035934,
      "num_tokens": 23484454.0,
      "step": 9265
    },
    {
      "entropy": 6.344149446487426,
      "epoch": 1.0698211194460474,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004897299714135721,
      "loss": 6.2177,
      "mean_token_accuracy": 0.15650669634342193,
      "num_tokens": 23497022.0,
      "step": 9270
    },
    {
      "entropy": 6.39455246925354,
      "epoch": 1.070398153491056,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004897176432439974,
      "loss": 6.2155,
      "mean_token_accuracy": 0.16501809060573577,
      "num_tokens": 23510714.0,
      "step": 9275
    },
    {
      "entropy": 6.408887195587158,
      "epoch": 1.0709751875360647,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004897053078525016,
      "loss": 6.196,
      "mean_token_accuracy": 0.15856562554836273,
      "num_tokens": 23522927.0,
      "step": 9280
    },
    {
      "entropy": 6.287023735046387,
      "epoch": 1.0715522215810733,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004896929652394993,
      "loss": 6.0978,
      "mean_token_accuracy": 0.16473745703697204,
      "num_tokens": 23536042.0,
      "step": 9285
    },
    {
      "entropy": 6.364973068237305,
      "epoch": 1.072129255626082,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004896806154054057,
      "loss": 6.2441,
      "mean_token_accuracy": 0.15745574235916138,
      "num_tokens": 23549002.0,
      "step": 9290
    },
    {
      "entropy": 6.400933647155762,
      "epoch": 1.0727062896710906,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004896682583506363,
      "loss": 6.2686,
      "mean_token_accuracy": 0.15413917899131774,
      "num_tokens": 23563080.0,
      "step": 9295
    },
    {
      "entropy": 6.418299388885498,
      "epoch": 1.0732833237160992,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004896558940756067,
      "loss": 6.2007,
      "mean_token_accuracy": 0.15697152465581893,
      "num_tokens": 23577431.0,
      "step": 9300
    },
    {
      "entropy": 6.302762603759765,
      "epoch": 1.0738603577611079,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004896435225807327,
      "loss": 6.1245,
      "mean_token_accuracy": 0.16009364873170853,
      "num_tokens": 23591191.0,
      "step": 9305
    },
    {
      "entropy": 6.322522497177124,
      "epoch": 1.0744373918061165,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004896311438664304,
      "loss": 6.0839,
      "mean_token_accuracy": 0.17335233688354493,
      "num_tokens": 23603195.0,
      "step": 9310
    },
    {
      "entropy": 6.314301586151123,
      "epoch": 1.0750144258511252,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004896187579331163,
      "loss": 6.1446,
      "mean_token_accuracy": 0.16357824355363845,
      "num_tokens": 23615278.0,
      "step": 9315
    },
    {
      "entropy": 6.371065235137939,
      "epoch": 1.0755914598961338,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004896063647812068,
      "loss": 6.2056,
      "mean_token_accuracy": 0.16097504198551177,
      "num_tokens": 23627311.0,
      "step": 9320
    },
    {
      "entropy": 6.334024095535279,
      "epoch": 1.0761684939411424,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004895939644111189,
      "loss": 6.1931,
      "mean_token_accuracy": 0.1588624119758606,
      "num_tokens": 23640338.0,
      "step": 9325
    },
    {
      "entropy": 6.397532653808594,
      "epoch": 1.076745527986151,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004895815568232694,
      "loss": 6.2591,
      "mean_token_accuracy": 0.15749427080154418,
      "num_tokens": 23654541.0,
      "step": 9330
    },
    {
      "entropy": 6.384134292602539,
      "epoch": 1.07732256203116,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004895691420180759,
      "loss": 6.2222,
      "mean_token_accuracy": 0.16097208261489868,
      "num_tokens": 23666837.0,
      "step": 9335
    },
    {
      "entropy": 6.424416351318359,
      "epoch": 1.0778995960761686,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000489556719995956,
      "loss": 6.22,
      "mean_token_accuracy": 0.15995656698942184,
      "num_tokens": 23679270.0,
      "step": 9340
    },
    {
      "entropy": 6.345190954208374,
      "epoch": 1.0784766301211772,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004895442907573274,
      "loss": 6.2035,
      "mean_token_accuracy": 0.1626092180609703,
      "num_tokens": 23691640.0,
      "step": 9345
    },
    {
      "entropy": 6.308082056045532,
      "epoch": 1.0790536641661859,
      "grad_norm": 1.140625,
      "learning_rate": 0.000489531854302608,
      "loss": 6.2063,
      "mean_token_accuracy": 0.1572630599141121,
      "num_tokens": 23704545.0,
      "step": 9350
    },
    {
      "entropy": 6.277027750015259,
      "epoch": 1.0796306982111945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004895194106322164,
      "loss": 6.1235,
      "mean_token_accuracy": 0.17247570157051087,
      "num_tokens": 23716319.0,
      "step": 9355
    },
    {
      "entropy": 6.384001874923706,
      "epoch": 1.0802077322562031,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004895069597465709,
      "loss": 6.2092,
      "mean_token_accuracy": 0.16064286977052689,
      "num_tokens": 23728633.0,
      "step": 9360
    },
    {
      "entropy": 6.416384887695313,
      "epoch": 1.0807847663012118,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004894945016460904,
      "loss": 6.2165,
      "mean_token_accuracy": 0.1609300494194031,
      "num_tokens": 23741370.0,
      "step": 9365
    },
    {
      "entropy": 6.349208211898803,
      "epoch": 1.0813618003462204,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004894820363311938,
      "loss": 6.1808,
      "mean_token_accuracy": 0.16120522022247313,
      "num_tokens": 23754026.0,
      "step": 9370
    },
    {
      "entropy": 6.358585834503174,
      "epoch": 1.081938834391229,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004894695638023005,
      "loss": 6.2496,
      "mean_token_accuracy": 0.1550566226243973,
      "num_tokens": 23766950.0,
      "step": 9375
    },
    {
      "entropy": 6.364051389694214,
      "epoch": 1.0825158684362377,
      "grad_norm": 1.25,
      "learning_rate": 0.00048945708405983,
      "loss": 6.2438,
      "mean_token_accuracy": 0.16189382672309877,
      "num_tokens": 23780858.0,
      "step": 9380
    },
    {
      "entropy": 6.405634260177612,
      "epoch": 1.0830929024812463,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004894445971042019,
      "loss": 6.1763,
      "mean_token_accuracy": 0.1622998982667923,
      "num_tokens": 23793213.0,
      "step": 9385
    },
    {
      "entropy": 6.302175426483155,
      "epoch": 1.083669936526255,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004894321029358364,
      "loss": 6.1903,
      "mean_token_accuracy": 0.1612499751150608,
      "num_tokens": 23804754.0,
      "step": 9390
    },
    {
      "entropy": 6.287882852554321,
      "epoch": 1.0842469705712636,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004894196015551536,
      "loss": 6.2081,
      "mean_token_accuracy": 0.16173621714115144,
      "num_tokens": 23817503.0,
      "step": 9395
    },
    {
      "entropy": 6.3686400890350345,
      "epoch": 1.0848240046162723,
      "grad_norm": 1.40625,
      "learning_rate": 0.000489407092962574,
      "loss": 6.2052,
      "mean_token_accuracy": 0.1568010240793228,
      "num_tokens": 23830263.0,
      "step": 9400
    },
    {
      "entropy": 6.36795973777771,
      "epoch": 1.085401038661281,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004893945771585183,
      "loss": 6.1213,
      "mean_token_accuracy": 0.16961509138345718,
      "num_tokens": 23843734.0,
      "step": 9405
    },
    {
      "entropy": 6.360470724105835,
      "epoch": 1.0859780727062898,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004893820541434075,
      "loss": 6.2461,
      "mean_token_accuracy": 0.16234959214925765,
      "num_tokens": 23856857.0,
      "step": 9410
    },
    {
      "entropy": 6.344463777542114,
      "epoch": 1.0865551067512984,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004893695239176629,
      "loss": 6.1358,
      "mean_token_accuracy": 0.16343702822923661,
      "num_tokens": 23869961.0,
      "step": 9415
    },
    {
      "entropy": 6.357371282577515,
      "epoch": 1.087132140796307,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004893569864817057,
      "loss": 6.2664,
      "mean_token_accuracy": 0.15537794083356857,
      "num_tokens": 23884551.0,
      "step": 9420
    },
    {
      "entropy": 6.3662666320800785,
      "epoch": 1.0877091748413157,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004893444418359579,
      "loss": 6.2702,
      "mean_token_accuracy": 0.15453500151634217,
      "num_tokens": 23897240.0,
      "step": 9425
    },
    {
      "entropy": 6.384363794326783,
      "epoch": 1.0882862088863243,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000489331889980841,
      "loss": 6.1507,
      "mean_token_accuracy": 0.1617793247103691,
      "num_tokens": 23909677.0,
      "step": 9430
    },
    {
      "entropy": 6.378599834442139,
      "epoch": 1.088863242931333,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004893193309167778,
      "loss": 6.2053,
      "mean_token_accuracy": 0.15970574617385863,
      "num_tokens": 23921552.0,
      "step": 9435
    },
    {
      "entropy": 6.402880382537842,
      "epoch": 1.0894402769763416,
      "grad_norm": 4.65625,
      "learning_rate": 0.0004893067646441902,
      "loss": 6.2554,
      "mean_token_accuracy": 0.15138119906187059,
      "num_tokens": 23936315.0,
      "step": 9440
    },
    {
      "entropy": 6.350248575210571,
      "epoch": 1.0900173110213502,
      "grad_norm": 1.6875,
      "learning_rate": 0.000489294191163501,
      "loss": 6.229,
      "mean_token_accuracy": 0.15604216903448104,
      "num_tokens": 23948939.0,
      "step": 9445
    },
    {
      "entropy": 6.410356950759888,
      "epoch": 1.0905943450663589,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004892816104751331,
      "loss": 6.2336,
      "mean_token_accuracy": 0.15622987747192382,
      "num_tokens": 23961657.0,
      "step": 9450
    },
    {
      "entropy": 6.4098145961761475,
      "epoch": 1.0911713791113675,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004892690225795096,
      "loss": 6.3006,
      "mean_token_accuracy": 0.14848166555166245,
      "num_tokens": 23974085.0,
      "step": 9455
    },
    {
      "entropy": 6.303513717651367,
      "epoch": 1.0917484131563762,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004892564274770542,
      "loss": 6.1443,
      "mean_token_accuracy": 0.1626450762152672,
      "num_tokens": 23987038.0,
      "step": 9460
    },
    {
      "entropy": 6.442585802078247,
      "epoch": 1.0923254472013848,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004892438251681901,
      "loss": 6.2233,
      "mean_token_accuracy": 0.16099887639284133,
      "num_tokens": 24001051.0,
      "step": 9465
    },
    {
      "entropy": 6.355003547668457,
      "epoch": 1.0929024812463934,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004892312156533413,
      "loss": 6.1834,
      "mean_token_accuracy": 0.15782278925180435,
      "num_tokens": 24013817.0,
      "step": 9470
    },
    {
      "entropy": 6.28466534614563,
      "epoch": 1.0934795152914023,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004892185989329321,
      "loss": 6.1254,
      "mean_token_accuracy": 0.164686119556427,
      "num_tokens": 24025810.0,
      "step": 9475
    },
    {
      "entropy": 6.277740526199341,
      "epoch": 1.094056549336411,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004892059750073868,
      "loss": 6.1332,
      "mean_token_accuracy": 0.16929626762866973,
      "num_tokens": 24039796.0,
      "step": 9480
    },
    {
      "entropy": 6.366404342651367,
      "epoch": 1.0946335833814196,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004891933438771299,
      "loss": 6.1861,
      "mean_token_accuracy": 0.16250718981027604,
      "num_tokens": 24052721.0,
      "step": 9485
    },
    {
      "entropy": 6.3012937068939205,
      "epoch": 1.0952106174264282,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004891807055425862,
      "loss": 6.0907,
      "mean_token_accuracy": 0.17106976062059404,
      "num_tokens": 24064515.0,
      "step": 9490
    },
    {
      "entropy": 6.254371690750122,
      "epoch": 1.0957876514714369,
      "grad_norm": 1.375,
      "learning_rate": 0.0004891680600041809,
      "loss": 6.1401,
      "mean_token_accuracy": 0.15985100865364074,
      "num_tokens": 24078537.0,
      "step": 9495
    },
    {
      "entropy": 6.410112905502319,
      "epoch": 1.0963646855164455,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004891554072623392,
      "loss": 6.2528,
      "mean_token_accuracy": 0.15630560666322707,
      "num_tokens": 24092067.0,
      "step": 9500
    },
    {
      "entropy": 6.368152618408203,
      "epoch": 1.0969417195614541,
      "grad_norm": 2.21875,
      "learning_rate": 0.0004891427473174869,
      "loss": 6.2061,
      "mean_token_accuracy": 0.15990003943443298,
      "num_tokens": 24105625.0,
      "step": 9505
    },
    {
      "entropy": 6.396222496032715,
      "epoch": 1.0975187536064628,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891300801700496,
      "loss": 6.2431,
      "mean_token_accuracy": 0.15077004134654998,
      "num_tokens": 24119714.0,
      "step": 9510
    },
    {
      "entropy": 6.329865121841431,
      "epoch": 1.0980957876514714,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004891174058204534,
      "loss": 6.132,
      "mean_token_accuracy": 0.1595097690820694,
      "num_tokens": 24132873.0,
      "step": 9515
    },
    {
      "entropy": 6.31348352432251,
      "epoch": 1.09867282169648,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004891047242691246,
      "loss": 6.1466,
      "mean_token_accuracy": 0.15776203125715255,
      "num_tokens": 24145550.0,
      "step": 9520
    },
    {
      "entropy": 6.367596292495728,
      "epoch": 1.0992498557414887,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004890920355164897,
      "loss": 6.1561,
      "mean_token_accuracy": 0.15608249455690384,
      "num_tokens": 24158232.0,
      "step": 9525
    },
    {
      "entropy": 6.392232513427734,
      "epoch": 1.0998268897864973,
      "grad_norm": 5.21875,
      "learning_rate": 0.0004890793395629756,
      "loss": 6.2952,
      "mean_token_accuracy": 0.1520036354660988,
      "num_tokens": 24170780.0,
      "step": 9530
    },
    {
      "entropy": 6.346212100982666,
      "epoch": 1.100403923831506,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004890666364090093,
      "loss": 6.1696,
      "mean_token_accuracy": 0.16291800290346145,
      "num_tokens": 24182347.0,
      "step": 9535
    },
    {
      "entropy": 6.402856874465942,
      "epoch": 1.1009809578765146,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004890539260550181,
      "loss": 6.2753,
      "mean_token_accuracy": 0.15678591281175613,
      "num_tokens": 24195215.0,
      "step": 9540
    },
    {
      "entropy": 6.30384635925293,
      "epoch": 1.1015579919215233,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890412085014292,
      "loss": 6.1616,
      "mean_token_accuracy": 0.16906683444976806,
      "num_tokens": 24208214.0,
      "step": 9545
    },
    {
      "entropy": 6.382722663879394,
      "epoch": 1.1021350259665321,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004890284837486706,
      "loss": 6.187,
      "mean_token_accuracy": 0.1664133921265602,
      "num_tokens": 24220883.0,
      "step": 9550
    },
    {
      "entropy": 6.351633214950562,
      "epoch": 1.1027120600115408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004890157517971704,
      "loss": 6.1088,
      "mean_token_accuracy": 0.16918679922819138,
      "num_tokens": 24233918.0,
      "step": 9555
    },
    {
      "entropy": 6.356971645355225,
      "epoch": 1.1032890940565494,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004890030126473566,
      "loss": 6.247,
      "mean_token_accuracy": 0.15816302448511124,
      "num_tokens": 24247307.0,
      "step": 9560
    },
    {
      "entropy": 6.337526273727417,
      "epoch": 1.103866128101558,
      "grad_norm": 2.25,
      "learning_rate": 0.0004889902662996578,
      "loss": 6.1669,
      "mean_token_accuracy": 0.15724012702703477,
      "num_tokens": 24260149.0,
      "step": 9565
    },
    {
      "entropy": 6.375165748596191,
      "epoch": 1.1044431621465667,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004889775127545027,
      "loss": 6.238,
      "mean_token_accuracy": 0.15898802280426025,
      "num_tokens": 24273675.0,
      "step": 9570
    },
    {
      "entropy": 6.340470027923584,
      "epoch": 1.1050201961915753,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004889647520123202,
      "loss": 6.1316,
      "mean_token_accuracy": 0.1621945694088936,
      "num_tokens": 24285575.0,
      "step": 9575
    },
    {
      "entropy": 6.410197305679321,
      "epoch": 1.105597230236584,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004889519840735396,
      "loss": 6.187,
      "mean_token_accuracy": 0.16513626426458358,
      "num_tokens": 24297908.0,
      "step": 9580
    },
    {
      "entropy": 6.233777141571045,
      "epoch": 1.1061742642815926,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004889392089385903,
      "loss": 6.1212,
      "mean_token_accuracy": 0.16366908550262452,
      "num_tokens": 24310714.0,
      "step": 9585
    },
    {
      "entropy": 6.280074691772461,
      "epoch": 1.1067512983266012,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004889264266079019,
      "loss": 6.1752,
      "mean_token_accuracy": 0.16871743351221086,
      "num_tokens": 24322367.0,
      "step": 9590
    },
    {
      "entropy": 6.287104368209839,
      "epoch": 1.1073283323716099,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004889136370819045,
      "loss": 6.1854,
      "mean_token_accuracy": 0.15897441729903222,
      "num_tokens": 24334256.0,
      "step": 9595
    },
    {
      "entropy": 6.3352649211883545,
      "epoch": 1.1079053664166185,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004889008403610281,
      "loss": 6.1601,
      "mean_token_accuracy": 0.1653594493865967,
      "num_tokens": 24346968.0,
      "step": 9600
    },
    {
      "entropy": 6.436962127685547,
      "epoch": 1.1084824004616272,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004888880364457033,
      "loss": 6.2534,
      "mean_token_accuracy": 0.1597901850938797,
      "num_tokens": 24360468.0,
      "step": 9605
    },
    {
      "entropy": 6.369411945343018,
      "epoch": 1.1090594345066358,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004888752253363604,
      "loss": 6.1868,
      "mean_token_accuracy": 0.16491993069648742,
      "num_tokens": 24372330.0,
      "step": 9610
    },
    {
      "entropy": 6.2847589492797855,
      "epoch": 1.1096364685516447,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004888624070334308,
      "loss": 6.1342,
      "mean_token_accuracy": 0.16332882195711135,
      "num_tokens": 24384864.0,
      "step": 9615
    },
    {
      "entropy": 6.321532726287842,
      "epoch": 1.1102135025966533,
      "grad_norm": 1.25,
      "learning_rate": 0.0004888495815373452,
      "loss": 6.0681,
      "mean_token_accuracy": 0.16914285868406295,
      "num_tokens": 24397408.0,
      "step": 9620
    },
    {
      "entropy": 6.317392492294312,
      "epoch": 1.110790536641662,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004888367488485351,
      "loss": 6.1335,
      "mean_token_accuracy": 0.1612032026052475,
      "num_tokens": 24409113.0,
      "step": 9625
    },
    {
      "entropy": 6.314182138442993,
      "epoch": 1.1113675706866706,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004888239089674323,
      "loss": 6.1601,
      "mean_token_accuracy": 0.16142479479312896,
      "num_tokens": 24420818.0,
      "step": 9630
    },
    {
      "entropy": 6.383041763305664,
      "epoch": 1.1119446047316792,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004888110618944686,
      "loss": 6.2779,
      "mean_token_accuracy": 0.15037421733140946,
      "num_tokens": 24434452.0,
      "step": 9635
    },
    {
      "entropy": 6.427024078369141,
      "epoch": 1.1125216387766879,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004887982076300759,
      "loss": 6.2702,
      "mean_token_accuracy": 0.15480797290802,
      "num_tokens": 24448674.0,
      "step": 9640
    },
    {
      "entropy": 6.376167249679566,
      "epoch": 1.1130986728216965,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004887853461746867,
      "loss": 6.1727,
      "mean_token_accuracy": 0.15854671001434326,
      "num_tokens": 24461388.0,
      "step": 9645
    },
    {
      "entropy": 6.356609535217285,
      "epoch": 1.1136757068667051,
      "grad_norm": 1.375,
      "learning_rate": 0.0004887724775287336,
      "loss": 6.1436,
      "mean_token_accuracy": 0.16579209119081498,
      "num_tokens": 24474304.0,
      "step": 9650
    },
    {
      "entropy": 6.258968782424927,
      "epoch": 1.1142527409117138,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004887596016926494,
      "loss": 6.117,
      "mean_token_accuracy": 0.16495985984802247,
      "num_tokens": 24487175.0,
      "step": 9655
    },
    {
      "entropy": 6.345869302749634,
      "epoch": 1.1148297749567224,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004887467186668673,
      "loss": 6.1624,
      "mean_token_accuracy": 0.1574894145131111,
      "num_tokens": 24500160.0,
      "step": 9660
    },
    {
      "entropy": 6.4012657642364506,
      "epoch": 1.115406809001731,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004887338284518204,
      "loss": 6.2653,
      "mean_token_accuracy": 0.15609754621982574,
      "num_tokens": 24513358.0,
      "step": 9665
    },
    {
      "entropy": 6.367801094055176,
      "epoch": 1.1159838430467397,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004887209310479423,
      "loss": 6.1507,
      "mean_token_accuracy": 0.16932276785373687,
      "num_tokens": 24526254.0,
      "step": 9670
    },
    {
      "entropy": 6.327196407318115,
      "epoch": 1.1165608770917483,
      "grad_norm": 1.375,
      "learning_rate": 0.0004887080264556668,
      "loss": 6.082,
      "mean_token_accuracy": 0.16644769310951232,
      "num_tokens": 24540023.0,
      "step": 9675
    },
    {
      "entropy": 6.344431686401367,
      "epoch": 1.117137911136757,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004886951146754282,
      "loss": 6.1884,
      "mean_token_accuracy": 0.1614872083067894,
      "num_tokens": 24553270.0,
      "step": 9680
    },
    {
      "entropy": 6.389737462997436,
      "epoch": 1.1177149451817656,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004886821957076603,
      "loss": 6.1574,
      "mean_token_accuracy": 0.16487915813922882,
      "num_tokens": 24565576.0,
      "step": 9685
    },
    {
      "entropy": 6.313331651687622,
      "epoch": 1.1182919792267745,
      "grad_norm": 1.890625,
      "learning_rate": 0.000488669269552798,
      "loss": 6.2514,
      "mean_token_accuracy": 0.16085973232984543,
      "num_tokens": 24578825.0,
      "step": 9690
    },
    {
      "entropy": 6.3549232006073,
      "epoch": 1.1188690132717831,
      "grad_norm": 1.25,
      "learning_rate": 0.000488656336211276,
      "loss": 6.2449,
      "mean_token_accuracy": 0.15805209577083587,
      "num_tokens": 24592058.0,
      "step": 9695
    },
    {
      "entropy": 6.404150485992432,
      "epoch": 1.1194460473167918,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004886433956835292,
      "loss": 6.1607,
      "mean_token_accuracy": 0.16113292276859284,
      "num_tokens": 24605956.0,
      "step": 9700
    },
    {
      "entropy": 6.3350914478302,
      "epoch": 1.1200230813618004,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004886304479699929,
      "loss": 6.1736,
      "mean_token_accuracy": 0.16226267218589782,
      "num_tokens": 24618904.0,
      "step": 9705
    },
    {
      "entropy": 6.391560220718384,
      "epoch": 1.120600115406809,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004886174930711027,
      "loss": 6.3022,
      "mean_token_accuracy": 0.16147204041481017,
      "num_tokens": 24632720.0,
      "step": 9710
    },
    {
      "entropy": 6.404454040527344,
      "epoch": 1.1211771494518177,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004886045309872941,
      "loss": 6.2417,
      "mean_token_accuracy": 0.15741440653800964,
      "num_tokens": 24644863.0,
      "step": 9715
    },
    {
      "entropy": 6.412680292129517,
      "epoch": 1.1217541834968263,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004885915617190034,
      "loss": 6.2241,
      "mean_token_accuracy": 0.15678914189338683,
      "num_tokens": 24656884.0,
      "step": 9720
    },
    {
      "entropy": 6.399668502807617,
      "epoch": 1.122331217541835,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004885785852666664,
      "loss": 6.2334,
      "mean_token_accuracy": 0.15685741007328033,
      "num_tokens": 24671008.0,
      "step": 9725
    },
    {
      "entropy": 6.36036114692688,
      "epoch": 1.1229082515868436,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004885656016307199,
      "loss": 6.2253,
      "mean_token_accuracy": 0.15478505194187164,
      "num_tokens": 24685787.0,
      "step": 9730
    },
    {
      "entropy": 6.463143396377563,
      "epoch": 1.1234852856318522,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004885526108116004,
      "loss": 6.2564,
      "mean_token_accuracy": 0.16071103662252426,
      "num_tokens": 24697233.0,
      "step": 9735
    },
    {
      "entropy": 6.365033102035523,
      "epoch": 1.1240623196768609,
      "grad_norm": 1.5,
      "learning_rate": 0.000488539612809745,
      "loss": 6.1716,
      "mean_token_accuracy": 0.162790547311306,
      "num_tokens": 24709409.0,
      "step": 9740
    },
    {
      "entropy": 6.303580236434937,
      "epoch": 1.1246393537218695,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004885266076255907,
      "loss": 6.2484,
      "mean_token_accuracy": 0.15676273554563522,
      "num_tokens": 24722006.0,
      "step": 9745
    },
    {
      "entropy": 6.456765222549438,
      "epoch": 1.1252163877668782,
      "grad_norm": 1.578125,
      "learning_rate": 0.000488513595259575,
      "loss": 6.2482,
      "mean_token_accuracy": 0.15536370426416396,
      "num_tokens": 24736060.0,
      "step": 9750
    },
    {
      "entropy": 6.362082052230835,
      "epoch": 1.125793421811887,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004885005757121357,
      "loss": 6.222,
      "mean_token_accuracy": 0.1529646933078766,
      "num_tokens": 24748008.0,
      "step": 9755
    },
    {
      "entropy": 6.346718120574951,
      "epoch": 1.1263704558568954,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004884875489837105,
      "loss": 6.2068,
      "mean_token_accuracy": 0.15854579210281372,
      "num_tokens": 24760631.0,
      "step": 9760
    },
    {
      "entropy": 6.198451948165894,
      "epoch": 1.1269474899019043,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004884745150747378,
      "loss": 6.0406,
      "mean_token_accuracy": 0.17826538532972336,
      "num_tokens": 24772646.0,
      "step": 9765
    },
    {
      "entropy": 6.310153341293335,
      "epoch": 1.127524523946913,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004884614739856556,
      "loss": 6.2322,
      "mean_token_accuracy": 0.15513329356908798,
      "num_tokens": 24784571.0,
      "step": 9770
    },
    {
      "entropy": 6.441531229019165,
      "epoch": 1.1281015579919216,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004884484257169028,
      "loss": 6.218,
      "mean_token_accuracy": 0.16059867888689042,
      "num_tokens": 24797810.0,
      "step": 9775
    },
    {
      "entropy": 6.338571166992187,
      "epoch": 1.1286785920369302,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884353702689183,
      "loss": 6.2187,
      "mean_token_accuracy": 0.16038416177034379,
      "num_tokens": 24810677.0,
      "step": 9780
    },
    {
      "entropy": 6.347803831100464,
      "epoch": 1.1292556260819389,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004884223076421411,
      "loss": 6.1244,
      "mean_token_accuracy": 0.16551823019981385,
      "num_tokens": 24823138.0,
      "step": 9785
    },
    {
      "entropy": 6.200040245056153,
      "epoch": 1.1298326601269475,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004884092378370106,
      "loss": 6.1284,
      "mean_token_accuracy": 0.17274417281150817,
      "num_tokens": 24834548.0,
      "step": 9790
    },
    {
      "entropy": 6.300217199325561,
      "epoch": 1.1304096941719561,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004883961608539664,
      "loss": 6.2222,
      "mean_token_accuracy": 0.16022792905569078,
      "num_tokens": 24847326.0,
      "step": 9795
    },
    {
      "entropy": 6.326699066162109,
      "epoch": 1.1309867282169648,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004883830766934484,
      "loss": 6.07,
      "mean_token_accuracy": 0.16643578335642814,
      "num_tokens": 24860417.0,
      "step": 9800
    },
    {
      "entropy": 6.307373332977295,
      "epoch": 1.1315637622619734,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004883699853558965,
      "loss": 6.1698,
      "mean_token_accuracy": 0.15934834480285645,
      "num_tokens": 24872759.0,
      "step": 9805
    },
    {
      "entropy": 6.354069089889526,
      "epoch": 1.132140796306982,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004883568868417511,
      "loss": 6.1756,
      "mean_token_accuracy": 0.15413878858089447,
      "num_tokens": 24885278.0,
      "step": 9810
    },
    {
      "entropy": 6.389302158355713,
      "epoch": 1.1327178303519907,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004883437811514528,
      "loss": 6.253,
      "mean_token_accuracy": 0.15297829508781433,
      "num_tokens": 24898596.0,
      "step": 9815
    },
    {
      "entropy": 6.396500825881958,
      "epoch": 1.1332948643969993,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004883306682854424,
      "loss": 6.1644,
      "mean_token_accuracy": 0.163560388982296,
      "num_tokens": 24911755.0,
      "step": 9820
    },
    {
      "entropy": 6.341898441314697,
      "epoch": 1.133871898442008,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004883175482441611,
      "loss": 6.2487,
      "mean_token_accuracy": 0.1547734707593918,
      "num_tokens": 24925197.0,
      "step": 9825
    },
    {
      "entropy": 6.277676725387574,
      "epoch": 1.1344489324870168,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004883044210280499,
      "loss": 6.0629,
      "mean_token_accuracy": 0.17149295508861542,
      "num_tokens": 24938204.0,
      "step": 9830
    },
    {
      "entropy": 6.299826765060425,
      "epoch": 1.1350259665320255,
      "grad_norm": 2.703125,
      "learning_rate": 0.0004882912866375505,
      "loss": 6.2069,
      "mean_token_accuracy": 0.16771370023489,
      "num_tokens": 24949382.0,
      "step": 9835
    },
    {
      "entropy": 6.248973083496094,
      "epoch": 1.1356030005770341,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048827814507310455,
      "loss": 6.1424,
      "mean_token_accuracy": 0.15670545101165773,
      "num_tokens": 24961685.0,
      "step": 9840
    },
    {
      "entropy": 6.33607850074768,
      "epoch": 1.1361800346220428,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048826499633515416,
      "loss": 6.1412,
      "mean_token_accuracy": 0.1669670268893242,
      "num_tokens": 24973619.0,
      "step": 9845
    },
    {
      "entropy": 6.331471538543701,
      "epoch": 1.1367570686670514,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048825184042414156,
      "loss": 6.1318,
      "mean_token_accuracy": 0.1675176814198494,
      "num_tokens": 24985156.0,
      "step": 9850
    },
    {
      "entropy": 6.311733245849609,
      "epoch": 1.13733410271206,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004882386773405092,
      "loss": 6.1588,
      "mean_token_accuracy": 0.17118309289216996,
      "num_tokens": 24997891.0,
      "step": 9855
    },
    {
      "entropy": 6.35788254737854,
      "epoch": 1.1379111367570687,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004882255070846999,
      "loss": 6.1654,
      "mean_token_accuracy": 0.16182998716831207,
      "num_tokens": 25009824.0,
      "step": 9860
    },
    {
      "entropy": 6.362553501129151,
      "epoch": 1.1384881708020773,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048821232965715663,
      "loss": 6.2016,
      "mean_token_accuracy": 0.16237591803073884,
      "num_tokens": 25021574.0,
      "step": 9865
    },
    {
      "entropy": 6.32572078704834,
      "epoch": 1.139065204847086,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004881991450583225,
      "loss": 6.1044,
      "mean_token_accuracy": 0.1635513037443161,
      "num_tokens": 25033315.0,
      "step": 9870
    },
    {
      "entropy": 6.2788228511810305,
      "epoch": 1.1396422388920946,
      "grad_norm": 1.578125,
      "learning_rate": 0.000488185953288641,
      "loss": 6.1806,
      "mean_token_accuracy": 0.16586582809686662,
      "num_tokens": 25046284.0,
      "step": 9875
    },
    {
      "entropy": 6.362651920318603,
      "epoch": 1.1402192729371032,
      "grad_norm": 1.625,
      "learning_rate": 0.0004881727543485559,
      "loss": 6.175,
      "mean_token_accuracy": 0.15997907221317292,
      "num_tokens": 25058484.0,
      "step": 9880
    },
    {
      "entropy": 6.3163597106933596,
      "epoch": 1.1407963069821119,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048815954823851107,
      "loss": 6.0916,
      "mean_token_accuracy": 0.1663131073117256,
      "num_tokens": 25070555.0,
      "step": 9885
    },
    {
      "entropy": 6.389148378372193,
      "epoch": 1.1413733410271205,
      "grad_norm": 1.375,
      "learning_rate": 0.00048814633495895067,
      "loss": 6.2474,
      "mean_token_accuracy": 0.15366053134202956,
      "num_tokens": 25083492.0,
      "step": 9890
    },
    {
      "entropy": 6.3734739303588865,
      "epoch": 1.1419503750721294,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004881331145103192,
      "loss": 6.2516,
      "mean_token_accuracy": 0.15353629887104034,
      "num_tokens": 25096332.0,
      "step": 9895
    },
    {
      "entropy": 6.382818603515625,
      "epoch": 1.1425274091171378,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0004881198868930614,
      "loss": 6.1785,
      "mean_token_accuracy": 0.15331874191761016,
      "num_tokens": 25110447.0,
      "step": 9900
    },
    {
      "entropy": 6.479769325256347,
      "epoch": 1.1431044431621467,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048810665210762195,
      "loss": 6.2074,
      "mean_token_accuracy": 0.16485313773155214,
      "num_tokens": 25123852.0,
      "step": 9905
    },
    {
      "entropy": 6.382903051376343,
      "epoch": 1.1436814772071553,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048809341015444615,
      "loss": 6.2096,
      "mean_token_accuracy": 0.15939941108226777,
      "num_tokens": 25136151.0,
      "step": 9910
    },
    {
      "entropy": 6.3541422367095945,
      "epoch": 1.144258511252164,
      "grad_norm": 1.984375,
      "learning_rate": 0.00048808016103397934,
      "loss": 6.168,
      "mean_token_accuracy": 0.16297494173049926,
      "num_tokens": 25149642.0,
      "step": 9915
    },
    {
      "entropy": 6.3219459533691404,
      "epoch": 1.1448355452971726,
      "grad_norm": 2.671875,
      "learning_rate": 0.0004880669047466671,
      "loss": 6.1206,
      "mean_token_accuracy": 0.17158780246973038,
      "num_tokens": 25162022.0,
      "step": 9920
    },
    {
      "entropy": 6.376477193832398,
      "epoch": 1.1454125793421812,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048805364129295554,
      "loss": 6.2791,
      "mean_token_accuracy": 0.1544181153178215,
      "num_tokens": 25175404.0,
      "step": 9925
    },
    {
      "entropy": 6.315047931671143,
      "epoch": 1.1459896133871899,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048804037067329037,
      "loss": 6.0942,
      "mean_token_accuracy": 0.16884433776140212,
      "num_tokens": 25188148.0,
      "step": 9930
    },
    {
      "entropy": 6.355247545242309,
      "epoch": 1.1465666474321985,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004880270928881183,
      "loss": 6.1734,
      "mean_token_accuracy": 0.1672818660736084,
      "num_tokens": 25199960.0,
      "step": 9935
    },
    {
      "entropy": 6.2450096130371096,
      "epoch": 1.1471436814772071,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0004880138079378857,
      "loss": 6.1302,
      "mean_token_accuracy": 0.16437119990587234,
      "num_tokens": 25211054.0,
      "step": 9940
    },
    {
      "entropy": 6.339324235916138,
      "epoch": 1.1477207155222158,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004880005158230395,
      "loss": 6.0889,
      "mean_token_accuracy": 0.17020729631185533,
      "num_tokens": 25222553.0,
      "step": 9945
    },
    {
      "entropy": 6.344603872299194,
      "epoch": 1.1482977495672244,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004879872165440268,
      "loss": 6.1685,
      "mean_token_accuracy": 0.155762979388237,
      "num_tokens": 25235472.0,
      "step": 9950
    },
    {
      "entropy": 6.270013475418091,
      "epoch": 1.148874783612233,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004879739101012948,
      "loss": 6.2306,
      "mean_token_accuracy": 0.1624113589525223,
      "num_tokens": 25250209.0,
      "step": 9955
    },
    {
      "entropy": 6.363943099975586,
      "epoch": 1.1494518176572417,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004879605964952911,
      "loss": 6.2106,
      "mean_token_accuracy": 0.1601069986820221,
      "num_tokens": 25262064.0,
      "step": 9960
    },
    {
      "entropy": 6.420304775238037,
      "epoch": 1.1500288517022503,
      "grad_norm": 2.328125,
      "learning_rate": 0.00048794727572646366,
      "loss": 6.1949,
      "mean_token_accuracy": 0.1625082701444626,
      "num_tokens": 25274833.0,
      "step": 9965
    },
    {
      "entropy": 6.368671655654907,
      "epoch": 1.1506058857472592,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004879339477952603,
      "loss": 6.2022,
      "mean_token_accuracy": 0.16648445650935173,
      "num_tokens": 25288397.0,
      "step": 9970
    },
    {
      "entropy": 6.307248640060425,
      "epoch": 1.1511829197922678,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048792061270212935,
      "loss": 6.133,
      "mean_token_accuracy": 0.17006682157516478,
      "num_tokens": 25300887.0,
      "step": 9975
    },
    {
      "entropy": 6.401169538497925,
      "epoch": 1.1517599538372765,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004879072704475193,
      "loss": 6.2349,
      "mean_token_accuracy": 0.16358886212110518,
      "num_tokens": 25314686.0,
      "step": 9980
    },
    {
      "entropy": 6.440695905685425,
      "epoch": 1.1523369878822851,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048789392103187906,
      "loss": 6.2611,
      "mean_token_accuracy": 0.16165925413370133,
      "num_tokens": 25328253.0,
      "step": 9985
    },
    {
      "entropy": 6.404678010940552,
      "epoch": 1.1529140219272938,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004878805644556575,
      "loss": 6.1962,
      "mean_token_accuracy": 0.16324919015169143,
      "num_tokens": 25340583.0,
      "step": 9990
    },
    {
      "entropy": 6.297409296035767,
      "epoch": 1.1534910559723024,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000487867200719304,
      "loss": 6.1262,
      "mean_token_accuracy": 0.1642940953373909,
      "num_tokens": 25352513.0,
      "step": 9995
    },
    {
      "entropy": 6.331809711456299,
      "epoch": 1.154068090017311,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004878538298232678,
      "loss": 6.2146,
      "mean_token_accuracy": 0.15754812508821486,
      "num_tokens": 25365044.0,
      "step": 10000
    },
    {
      "entropy": 6.367120790481567,
      "epoch": 1.1546451240623197,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004878404517679988,
      "loss": 6.1761,
      "mean_token_accuracy": 0.15582527220249176,
      "num_tokens": 25377381.0,
      "step": 10005
    },
    {
      "entropy": 6.400685453414917,
      "epoch": 1.1552221581073283,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048782706655394695,
      "loss": 6.18,
      "mean_token_accuracy": 0.1643991768360138,
      "num_tokens": 25390723.0,
      "step": 10010
    },
    {
      "entropy": 6.296201992034912,
      "epoch": 1.155799192152337,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004878136741815624,
      "loss": 6.1212,
      "mean_token_accuracy": 0.16537229716777802,
      "num_tokens": 25403501.0,
      "step": 10015
    },
    {
      "entropy": 6.271032047271729,
      "epoch": 1.1563762261973456,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004878002746512956,
      "loss": 6.1031,
      "mean_token_accuracy": 0.17037995159626007,
      "num_tokens": 25415565.0,
      "step": 10020
    },
    {
      "entropy": 6.283905696868897,
      "epoch": 1.1569532602423542,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004877868679635974,
      "loss": 6.1687,
      "mean_token_accuracy": 0.16751373708248138,
      "num_tokens": 25428261.0,
      "step": 10025
    },
    {
      "entropy": 6.377343845367432,
      "epoch": 1.1575302942873629,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004877734541189185,
      "loss": 6.181,
      "mean_token_accuracy": 0.15702162981033324,
      "num_tokens": 25439758.0,
      "step": 10030
    },
    {
      "entropy": 6.360610103607177,
      "epoch": 1.1581073283323717,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048776003311771013,
      "loss": 6.2364,
      "mean_token_accuracy": 0.15337282419204712,
      "num_tokens": 25452162.0,
      "step": 10035
    },
    {
      "entropy": 6.4076014995574955,
      "epoch": 1.1586843623773802,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004877466049604238,
      "loss": 6.2545,
      "mean_token_accuracy": 0.15962868332862853,
      "num_tokens": 25463741.0,
      "step": 10040
    },
    {
      "entropy": 6.320121145248413,
      "epoch": 1.159261396422389,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048773316964751106,
      "loss": 6.1871,
      "mean_token_accuracy": 0.15946279913187028,
      "num_tokens": 25476466.0,
      "step": 10045
    },
    {
      "entropy": 6.422423934936523,
      "epoch": 1.1598384304673977,
      "grad_norm": 2.09375,
      "learning_rate": 0.0004877197271794239,
      "loss": 6.1692,
      "mean_token_accuracy": 0.1618300974369049,
      "num_tokens": 25488453.0,
      "step": 10050
    },
    {
      "entropy": 6.3750358581542965,
      "epoch": 1.1604154645124063,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004877062775566142,
      "loss": 6.1942,
      "mean_token_accuracy": 0.16088102161884307,
      "num_tokens": 25501803.0,
      "step": 10055
    },
    {
      "entropy": 6.349880695343018,
      "epoch": 1.160992498557415,
      "grad_norm": 2.0,
      "learning_rate": 0.00048769282077953464,
      "loss": 6.2892,
      "mean_token_accuracy": 0.15862567573785782,
      "num_tokens": 25515844.0,
      "step": 10060
    },
    {
      "entropy": 6.369120979309082,
      "epoch": 1.1615695326024236,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048767935684863775,
      "loss": 6.183,
      "mean_token_accuracy": 0.16268929988145828,
      "num_tokens": 25528529.0,
      "step": 10065
    },
    {
      "entropy": 6.397540330886841,
      "epoch": 1.1621465666474322,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004876658857643762,
      "loss": 6.2204,
      "mean_token_accuracy": 0.15767233669757844,
      "num_tokens": 25540110.0,
      "step": 10070
    },
    {
      "entropy": 6.364495658874512,
      "epoch": 1.1627236006924409,
      "grad_norm": 1.375,
      "learning_rate": 0.0004876524075272034,
      "loss": 6.1908,
      "mean_token_accuracy": 0.158960822224617,
      "num_tokens": 25553332.0,
      "step": 10075
    },
    {
      "entropy": 6.348095035552978,
      "epoch": 1.1633006347374495,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048763892213757234,
      "loss": 6.1686,
      "mean_token_accuracy": 0.16090085208415986,
      "num_tokens": 25565852.0,
      "step": 10080
    },
    {
      "entropy": 6.36542158126831,
      "epoch": 1.1638776687824581,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048762542959593683,
      "loss": 6.1672,
      "mean_token_accuracy": 0.1625974088907242,
      "num_tokens": 25579642.0,
      "step": 10085
    },
    {
      "entropy": 6.2937578678131105,
      "epoch": 1.1644547028274668,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004876119299027506,
      "loss": 6.1556,
      "mean_token_accuracy": 0.16866278797388076,
      "num_tokens": 25591452.0,
      "step": 10090
    },
    {
      "entropy": 6.290343618392944,
      "epoch": 1.1650317368724754,
      "grad_norm": 1.25,
      "learning_rate": 0.00048759842305846766,
      "loss": 6.1279,
      "mean_token_accuracy": 0.16313421428203584,
      "num_tokens": 25603426.0,
      "step": 10095
    },
    {
      "entropy": 6.355361557006836,
      "epoch": 1.165608770917484,
      "grad_norm": 1.5,
      "learning_rate": 0.0004875849090635425,
      "loss": 6.2365,
      "mean_token_accuracy": 0.15997890681028365,
      "num_tokens": 25617014.0,
      "step": 10100
    },
    {
      "entropy": 6.37194595336914,
      "epoch": 1.1661858049624927,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048757138791842943,
      "loss": 6.1972,
      "mean_token_accuracy": 0.16384346783161163,
      "num_tokens": 25629916.0,
      "step": 10105
    },
    {
      "entropy": 6.3468346118927,
      "epoch": 1.1667628390075016,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004875578596235832,
      "loss": 6.1866,
      "mean_token_accuracy": 0.16534270495176315,
      "num_tokens": 25642301.0,
      "step": 10110
    },
    {
      "entropy": 6.327104473114014,
      "epoch": 1.1673398730525102,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004875443241794591,
      "loss": 6.1448,
      "mean_token_accuracy": 0.16186633408069612,
      "num_tokens": 25653976.0,
      "step": 10115
    },
    {
      "entropy": 6.33780460357666,
      "epoch": 1.1679169070975188,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048753078158651227,
      "loss": 6.1864,
      "mean_token_accuracy": 0.1554411083459854,
      "num_tokens": 25666078.0,
      "step": 10120
    },
    {
      "entropy": 6.343927383422852,
      "epoch": 1.1684939411425275,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000487517231845198,
      "loss": 6.2353,
      "mean_token_accuracy": 0.15490031838417054,
      "num_tokens": 25679780.0,
      "step": 10125
    },
    {
      "entropy": 6.408223009109497,
      "epoch": 1.169070975187536,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004875036749559724,
      "loss": 6.176,
      "mean_token_accuracy": 0.1544538825750351,
      "num_tokens": 25694179.0,
      "step": 10130
    },
    {
      "entropy": 6.275277757644654,
      "epoch": 1.1696480092325447,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048749011091929115,
      "loss": 6.1543,
      "mean_token_accuracy": 0.17059712260961532,
      "num_tokens": 25707281.0,
      "step": 10135
    },
    {
      "entropy": 6.383667278289795,
      "epoch": 1.1702250432775534,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004874765397356105,
      "loss": 6.1877,
      "mean_token_accuracy": 0.16059649288654326,
      "num_tokens": 25719901.0,
      "step": 10140
    },
    {
      "entropy": 6.355267190933228,
      "epoch": 1.170802077322562,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004874629614053871,
      "loss": 6.1934,
      "mean_token_accuracy": 0.1627344861626625,
      "num_tokens": 25732063.0,
      "step": 10145
    },
    {
      "entropy": 6.267529439926148,
      "epoch": 1.1713791113675707,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004874493759290775,
      "loss": 6.1543,
      "mean_token_accuracy": 0.16306477189064025,
      "num_tokens": 25743966.0,
      "step": 10150
    },
    {
      "entropy": 6.147643136978149,
      "epoch": 1.1719561454125793,
      "grad_norm": 2.96875,
      "learning_rate": 0.00048743578330713854,
      "loss": 5.9489,
      "mean_token_accuracy": 0.1810133144259453,
      "num_tokens": 25757053.0,
      "step": 10155
    },
    {
      "entropy": 6.152535343170166,
      "epoch": 1.172533179457588,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004874221835400277,
      "loss": 6.0279,
      "mean_token_accuracy": 0.17614102661609649,
      "num_tokens": 25769430.0,
      "step": 10160
    },
    {
      "entropy": 6.3164385795593265,
      "epoch": 1.1731102135025966,
      "grad_norm": 1.25,
      "learning_rate": 0.0004874085766282022,
      "loss": 6.1467,
      "mean_token_accuracy": 0.1642584428191185,
      "num_tokens": 25781687.0,
      "step": 10165
    },
    {
      "entropy": 6.316783618927002,
      "epoch": 1.1736872475476052,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048739496257211966,
      "loss": 6.1701,
      "mean_token_accuracy": 0.1690356597304344,
      "num_tokens": 25793646.0,
      "step": 10170
    },
    {
      "entropy": 6.33498969078064,
      "epoch": 1.1742642815926139,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048738134137223815,
      "loss": 6.2383,
      "mean_token_accuracy": 0.1570924326777458,
      "num_tokens": 25807757.0,
      "step": 10175
    },
    {
      "entropy": 6.342399215698242,
      "epoch": 1.1748413156376225,
      "grad_norm": 2.546875,
      "learning_rate": 0.00048736771302901566,
      "loss": 6.1804,
      "mean_token_accuracy": 0.16174473017454147,
      "num_tokens": 25820063.0,
      "step": 10180
    },
    {
      "entropy": 6.319576597213745,
      "epoch": 1.1754183496826314,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048735407754291063,
      "loss": 6.123,
      "mean_token_accuracy": 0.16988920718431472,
      "num_tokens": 25832974.0,
      "step": 10185
    },
    {
      "entropy": 6.379149007797241,
      "epoch": 1.17599538372764,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048734043491438175,
      "loss": 6.2003,
      "mean_token_accuracy": 0.15487931072711944,
      "num_tokens": 25845939.0,
      "step": 10190
    },
    {
      "entropy": 6.3196595191955565,
      "epoch": 1.1765724177726486,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048732678514388773,
      "loss": 6.1455,
      "mean_token_accuracy": 0.16049207746982574,
      "num_tokens": 25857480.0,
      "step": 10195
    },
    {
      "entropy": 6.291083765029907,
      "epoch": 1.1771494518176573,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004873131282318878,
      "loss": 6.0832,
      "mean_token_accuracy": 0.1705268293619156,
      "num_tokens": 25869384.0,
      "step": 10200
    },
    {
      "entropy": 6.30240650177002,
      "epoch": 1.177726485862666,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048729946417884126,
      "loss": 6.1705,
      "mean_token_accuracy": 0.16226058602333068,
      "num_tokens": 25881705.0,
      "step": 10205
    },
    {
      "entropy": 6.36404013633728,
      "epoch": 1.1783035199076746,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004872857929852076,
      "loss": 6.1753,
      "mean_token_accuracy": 0.15568970143795013,
      "num_tokens": 25894072.0,
      "step": 10210
    },
    {
      "entropy": 6.363600587844848,
      "epoch": 1.1788805539526832,
      "grad_norm": 1.5,
      "learning_rate": 0.0004872721146514469,
      "loss": 6.0612,
      "mean_token_accuracy": 0.17028955072164537,
      "num_tokens": 25907358.0,
      "step": 10215
    },
    {
      "entropy": 6.243066787719727,
      "epoch": 1.1794575879976918,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000487258429178019,
      "loss": 6.1396,
      "mean_token_accuracy": 0.17397456020116805,
      "num_tokens": 25918746.0,
      "step": 10220
    },
    {
      "entropy": 6.3143209457397464,
      "epoch": 1.1800346220427005,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048724473656538427,
      "loss": 6.1761,
      "mean_token_accuracy": 0.1687172085046768,
      "num_tokens": 25930759.0,
      "step": 10225
    },
    {
      "entropy": 6.290168237686157,
      "epoch": 1.1806116560877091,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004872310368140032,
      "loss": 6.0937,
      "mean_token_accuracy": 0.16777861267328262,
      "num_tokens": 25942364.0,
      "step": 10230
    },
    {
      "entropy": 6.269335699081421,
      "epoch": 1.1811886901327178,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004872173299243368,
      "loss": 6.1418,
      "mean_token_accuracy": 0.17016907334327697,
      "num_tokens": 25954254.0,
      "step": 10235
    },
    {
      "entropy": 6.335594272613525,
      "epoch": 1.1817657241777264,
      "grad_norm": 2.15625,
      "learning_rate": 0.00048720361589684575,
      "loss": 6.2474,
      "mean_token_accuracy": 0.16507446467876435,
      "num_tokens": 25966708.0,
      "step": 10240
    },
    {
      "entropy": 6.332195568084717,
      "epoch": 1.182342758222735,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048718989473199147,
      "loss": 6.1933,
      "mean_token_accuracy": 0.1564444601535797,
      "num_tokens": 25979301.0,
      "step": 10245
    },
    {
      "entropy": 6.362717962265014,
      "epoch": 1.182919792267744,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004871761664302356,
      "loss": 6.1521,
      "mean_token_accuracy": 0.16474512219429016,
      "num_tokens": 25992195.0,
      "step": 10250
    },
    {
      "entropy": 6.284235668182373,
      "epoch": 1.1834968263127525,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004871624309920397,
      "loss": 6.041,
      "mean_token_accuracy": 0.1756657689809799,
      "num_tokens": 26005078.0,
      "step": 10255
    },
    {
      "entropy": 6.291142082214355,
      "epoch": 1.1840738603577612,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048714868841786586,
      "loss": 6.1701,
      "mean_token_accuracy": 0.16087482869625092,
      "num_tokens": 26017227.0,
      "step": 10260
    },
    {
      "entropy": 6.382811117172241,
      "epoch": 1.1846508944027698,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048713493870817626,
      "loss": 6.2511,
      "mean_token_accuracy": 0.16183987408876419,
      "num_tokens": 26030782.0,
      "step": 10265
    },
    {
      "entropy": 6.279626083374024,
      "epoch": 1.1852279284477785,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048712118186343336,
      "loss": 6.0983,
      "mean_token_accuracy": 0.1711464300751686,
      "num_tokens": 26042327.0,
      "step": 10270
    },
    {
      "entropy": 6.340450572967529,
      "epoch": 1.185804962492787,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048710741788409985,
      "loss": 6.1993,
      "mean_token_accuracy": 0.15717075914144515,
      "num_tokens": 26055298.0,
      "step": 10275
    },
    {
      "entropy": 6.387430858612061,
      "epoch": 1.1863819965377957,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004870936467706387,
      "loss": 6.2193,
      "mean_token_accuracy": 0.15629870891571046,
      "num_tokens": 26067389.0,
      "step": 10280
    },
    {
      "entropy": 6.354550886154175,
      "epoch": 1.1869590305828044,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004870798685235131,
      "loss": 6.1647,
      "mean_token_accuracy": 0.1612406626343727,
      "num_tokens": 26080527.0,
      "step": 10285
    },
    {
      "entropy": 6.320342922210694,
      "epoch": 1.187536064627813,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048706608314318654,
      "loss": 6.2134,
      "mean_token_accuracy": 0.15583257228136063,
      "num_tokens": 26093684.0,
      "step": 10290
    },
    {
      "entropy": 6.312257146835327,
      "epoch": 1.1881130986728217,
      "grad_norm": 1.125,
      "learning_rate": 0.0004870522906301225,
      "loss": 6.1227,
      "mean_token_accuracy": 0.16510868221521377,
      "num_tokens": 26105235.0,
      "step": 10295
    },
    {
      "entropy": 6.306887769699097,
      "epoch": 1.1886901327178303,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000487038490984785,
      "loss": 6.1452,
      "mean_token_accuracy": 0.1623164415359497,
      "num_tokens": 26117542.0,
      "step": 10300
    },
    {
      "entropy": 6.345963191986084,
      "epoch": 1.189267166762839,
      "grad_norm": 1.8125,
      "learning_rate": 0.00048702468420763826,
      "loss": 6.2605,
      "mean_token_accuracy": 0.15688117891550063,
      "num_tokens": 26130722.0,
      "step": 10305
    },
    {
      "entropy": 6.332888174057007,
      "epoch": 1.1898442008078476,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048701087029914657,
      "loss": 6.0818,
      "mean_token_accuracy": 0.16274466514587402,
      "num_tokens": 26143420.0,
      "step": 10310
    },
    {
      "entropy": 6.385124921798706,
      "epoch": 1.1904212348528562,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004869970492597745,
      "loss": 6.2732,
      "mean_token_accuracy": 0.15256087481975555,
      "num_tokens": 26156041.0,
      "step": 10315
    },
    {
      "entropy": 6.275911903381347,
      "epoch": 1.1909982688978649,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004869832210899871,
      "loss": 6.1044,
      "mean_token_accuracy": 0.16806990206241607,
      "num_tokens": 26168075.0,
      "step": 10320
    },
    {
      "entropy": 6.313507080078125,
      "epoch": 1.1915753029428737,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048696938579024927,
      "loss": 6.1054,
      "mean_token_accuracy": 0.1636284902691841,
      "num_tokens": 26180285.0,
      "step": 10325
    },
    {
      "entropy": 6.27633147239685,
      "epoch": 1.1921523369878824,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048695554336102644,
      "loss": 6.121,
      "mean_token_accuracy": 0.16393174827098847,
      "num_tokens": 26193095.0,
      "step": 10330
    },
    {
      "entropy": 6.305689954757691,
      "epoch": 1.192729371032891,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048694169380278417,
      "loss": 6.1467,
      "mean_token_accuracy": 0.16732767075300217,
      "num_tokens": 26205477.0,
      "step": 10335
    },
    {
      "entropy": 6.285434341430664,
      "epoch": 1.1933064050778996,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004869278371159884,
      "loss": 6.1043,
      "mean_token_accuracy": 0.17125446498394012,
      "num_tokens": 26220057.0,
      "step": 10340
    },
    {
      "entropy": 6.371880960464478,
      "epoch": 1.1938834391229083,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048691397330110503,
      "loss": 6.1986,
      "mean_token_accuracy": 0.15532726496458055,
      "num_tokens": 26232327.0,
      "step": 10345
    },
    {
      "entropy": 6.229359245300293,
      "epoch": 1.194460473167917,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004869001023586005,
      "loss": 6.1477,
      "mean_token_accuracy": 0.1645748496055603,
      "num_tokens": 26245087.0,
      "step": 10350
    },
    {
      "entropy": 6.387897205352783,
      "epoch": 1.1950375072129256,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004868862242889413,
      "loss": 6.2016,
      "mean_token_accuracy": 0.15893673598766328,
      "num_tokens": 26258041.0,
      "step": 10355
    },
    {
      "entropy": 6.408567094802857,
      "epoch": 1.1956145412579342,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000486872339092594,
      "loss": 6.2046,
      "mean_token_accuracy": 0.15730382800102233,
      "num_tokens": 26270443.0,
      "step": 10360
    },
    {
      "entropy": 6.262932157516479,
      "epoch": 1.1961915753029428,
      "grad_norm": 1.375,
      "learning_rate": 0.000486858446770026,
      "loss": 6.1515,
      "mean_token_accuracy": 0.1608057364821434,
      "num_tokens": 26282685.0,
      "step": 10365
    },
    {
      "entropy": 6.319109964370727,
      "epoch": 1.1967686093479515,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004868445473217043,
      "loss": 6.2158,
      "mean_token_accuracy": 0.161103555560112,
      "num_tokens": 26296003.0,
      "step": 10370
    },
    {
      "entropy": 6.2928773880004885,
      "epoch": 1.1973456433929601,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004868306407480965,
      "loss": 6.1029,
      "mean_token_accuracy": 0.16285141408443451,
      "num_tokens": 26309102.0,
      "step": 10375
    },
    {
      "entropy": 6.35385308265686,
      "epoch": 1.1979226774379688,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048681672704967036,
      "loss": 6.1494,
      "mean_token_accuracy": 0.164886237680912,
      "num_tokens": 26321571.0,
      "step": 10380
    },
    {
      "entropy": 6.421018648147583,
      "epoch": 1.1984997114829774,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004868028062268938,
      "loss": 6.2463,
      "mean_token_accuracy": 0.15964062064886092,
      "num_tokens": 26334761.0,
      "step": 10385
    },
    {
      "entropy": 6.310076951980591,
      "epoch": 1.1990767455279863,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048678887828023504,
      "loss": 6.0642,
      "mean_token_accuracy": 0.1698111414909363,
      "num_tokens": 26348815.0,
      "step": 10390
    },
    {
      "entropy": 6.309312057495117,
      "epoch": 1.1996537795729947,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004867749432101626,
      "loss": 6.1605,
      "mean_token_accuracy": 0.1607862025499344,
      "num_tokens": 26361675.0,
      "step": 10395
    },
    {
      "entropy": 6.306365489959717,
      "epoch": 1.2002308136180035,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004867610010171451,
      "loss": 6.2685,
      "mean_token_accuracy": 0.15660690665245056,
      "num_tokens": 26375010.0,
      "step": 10400
    },
    {
      "entropy": 6.455476760864258,
      "epoch": 1.2008078476630122,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048674705170165147,
      "loss": 6.2228,
      "mean_token_accuracy": 0.16162673979997635,
      "num_tokens": 26388217.0,
      "step": 10405
    },
    {
      "entropy": 6.3156548023223875,
      "epoch": 1.2013848817080208,
      "grad_norm": 1.9375,
      "learning_rate": 0.000486733095264151,
      "loss": 6.1713,
      "mean_token_accuracy": 0.1644631251692772,
      "num_tokens": 26402088.0,
      "step": 10410
    },
    {
      "entropy": 6.300989246368408,
      "epoch": 1.2019619157530295,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048671913170511306,
      "loss": 6.1699,
      "mean_token_accuracy": 0.16167462170124053,
      "num_tokens": 26414816.0,
      "step": 10415
    },
    {
      "entropy": 6.333500766754151,
      "epoch": 1.202538949798038,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004867051610250073,
      "loss": 6.1075,
      "mean_token_accuracy": 0.16527679413557053,
      "num_tokens": 26426345.0,
      "step": 10420
    },
    {
      "entropy": 6.232378339767456,
      "epoch": 1.2031159838430467,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004866911832243035,
      "loss": 5.9599,
      "mean_token_accuracy": 0.1687069922685623,
      "num_tokens": 26437982.0,
      "step": 10425
    },
    {
      "entropy": 6.195571947097778,
      "epoch": 1.2036930178880554,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048667719830347203,
      "loss": 6.1396,
      "mean_token_accuracy": 0.15564415901899337,
      "num_tokens": 26449855.0,
      "step": 10430
    },
    {
      "entropy": 6.404849576950073,
      "epoch": 1.204270051933064,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048666320626298307,
      "loss": 6.1891,
      "mean_token_accuracy": 0.16427553296089173,
      "num_tokens": 26462927.0,
      "step": 10435
    },
    {
      "entropy": 6.369866466522216,
      "epoch": 1.2048470859780727,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048664920710330735,
      "loss": 6.1967,
      "mean_token_accuracy": 0.16102144569158555,
      "num_tokens": 26475022.0,
      "step": 10440
    },
    {
      "entropy": 6.272254943847656,
      "epoch": 1.2054241200230813,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048663520082491564,
      "loss": 6.1338,
      "mean_token_accuracy": 0.1697706624865532,
      "num_tokens": 26486959.0,
      "step": 10445
    },
    {
      "entropy": 6.312459659576416,
      "epoch": 1.20600115406809,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004866211874282791,
      "loss": 6.2305,
      "mean_token_accuracy": 0.16129841059446334,
      "num_tokens": 26500345.0,
      "step": 10450
    },
    {
      "entropy": 6.3494813442230225,
      "epoch": 1.2065781881130986,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000486607166913869,
      "loss": 6.2039,
      "mean_token_accuracy": 0.16385735422372819,
      "num_tokens": 26511625.0,
      "step": 10455
    },
    {
      "entropy": 6.3159326076507565,
      "epoch": 1.2071552221581072,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048659313928215705,
      "loss": 6.1765,
      "mean_token_accuracy": 0.16436698138713837,
      "num_tokens": 26523764.0,
      "step": 10460
    },
    {
      "entropy": 6.311506223678589,
      "epoch": 1.207732256203116,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004865791045336149,
      "loss": 6.1735,
      "mean_token_accuracy": 0.1606744185090065,
      "num_tokens": 26536925.0,
      "step": 10465
    },
    {
      "entropy": 6.241578531265259,
      "epoch": 1.2083092902481247,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004865650626687146,
      "loss": 6.0318,
      "mean_token_accuracy": 0.17650363296270372,
      "num_tokens": 26549201.0,
      "step": 10470
    },
    {
      "entropy": 6.305879020690918,
      "epoch": 1.2088863242931334,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048655101368792866,
      "loss": 6.158,
      "mean_token_accuracy": 0.16642459928989412,
      "num_tokens": 26561495.0,
      "step": 10475
    },
    {
      "entropy": 6.3534986019134525,
      "epoch": 1.209463358338142,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004865369575917293,
      "loss": 6.1958,
      "mean_token_accuracy": 0.15989347100257872,
      "num_tokens": 26573820.0,
      "step": 10480
    },
    {
      "entropy": 6.335755109786987,
      "epoch": 1.2100403923831506,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048652289438058953,
      "loss": 6.1832,
      "mean_token_accuracy": 0.16023981124162673,
      "num_tokens": 26587468.0,
      "step": 10485
    },
    {
      "entropy": 6.260964012145996,
      "epoch": 1.2106174264281593,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048650882405498226,
      "loss": 6.2022,
      "mean_token_accuracy": 0.16167952865362167,
      "num_tokens": 26599884.0,
      "step": 10490
    },
    {
      "entropy": 6.255992984771728,
      "epoch": 1.211194460473168,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004864947466153808,
      "loss": 6.1251,
      "mean_token_accuracy": 0.16567855775356294,
      "num_tokens": 26612501.0,
      "step": 10495
    },
    {
      "entropy": 6.383900165557861,
      "epoch": 1.2117714945181766,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004864806620622585,
      "loss": 6.1492,
      "mean_token_accuracy": 0.17049698084592818,
      "num_tokens": 26624559.0,
      "step": 10500
    },
    {
      "entropy": 6.350470304489136,
      "epoch": 1.2123485285631852,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004864665703960892,
      "loss": 6.1593,
      "mean_token_accuracy": 0.1594645231962204,
      "num_tokens": 26636903.0,
      "step": 10505
    },
    {
      "entropy": 6.318430614471436,
      "epoch": 1.2129255626081938,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004864524716173469,
      "loss": 6.1214,
      "mean_token_accuracy": 0.16265250593423844,
      "num_tokens": 26649393.0,
      "step": 10510
    },
    {
      "entropy": 6.26350040435791,
      "epoch": 1.2135025966532025,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004864383657265058,
      "loss": 6.121,
      "mean_token_accuracy": 0.1690155252814293,
      "num_tokens": 26661134.0,
      "step": 10515
    },
    {
      "entropy": 6.280627727508545,
      "epoch": 1.2140796306982111,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048642425272404016,
      "loss": 6.2002,
      "mean_token_accuracy": 0.16160008162260056,
      "num_tokens": 26674451.0,
      "step": 10520
    },
    {
      "entropy": 6.328465127944947,
      "epoch": 1.2146566647432198,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004864101326104248,
      "loss": 6.1461,
      "mean_token_accuracy": 0.1652497887611389,
      "num_tokens": 26686382.0,
      "step": 10525
    },
    {
      "entropy": 6.283163166046142,
      "epoch": 1.2152336987882286,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004863960053861348,
      "loss": 6.1057,
      "mean_token_accuracy": 0.1680065721273422,
      "num_tokens": 26698491.0,
      "step": 10530
    },
    {
      "entropy": 6.267314338684082,
      "epoch": 1.215810732833237,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048638187105164507,
      "loss": 6.1228,
      "mean_token_accuracy": 0.16545756608247758,
      "num_tokens": 26710836.0,
      "step": 10535
    },
    {
      "entropy": 6.323285484313965,
      "epoch": 1.216387766878246,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004863677296074311,
      "loss": 6.1311,
      "mean_token_accuracy": 0.16326241195201874,
      "num_tokens": 26723195.0,
      "step": 10540
    },
    {
      "entropy": 6.319986152648926,
      "epoch": 1.2169648009232545,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004863535810539686,
      "loss": 6.1515,
      "mean_token_accuracy": 0.17019888758659363,
      "num_tokens": 26736048.0,
      "step": 10545
    },
    {
      "entropy": 6.2805334568023685,
      "epoch": 1.2175418349682632,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048633942539173325,
      "loss": 6.2238,
      "mean_token_accuracy": 0.1601146176457405,
      "num_tokens": 26748712.0,
      "step": 10550
    },
    {
      "entropy": 6.339949321746826,
      "epoch": 1.2181188690132718,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048632526262120144,
      "loss": 6.2048,
      "mean_token_accuracy": 0.15875715762376785,
      "num_tokens": 26760546.0,
      "step": 10555
    },
    {
      "entropy": 6.344151830673217,
      "epoch": 1.2186959030582805,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004863110927428493,
      "loss": 6.108,
      "mean_token_accuracy": 0.16397978961467743,
      "num_tokens": 26772792.0,
      "step": 10560
    },
    {
      "entropy": 6.23669023513794,
      "epoch": 1.219272937103289,
      "grad_norm": 2.0625,
      "learning_rate": 0.0004862969157571536,
      "loss": 6.0775,
      "mean_token_accuracy": 0.16881214827299118,
      "num_tokens": 26785321.0,
      "step": 10565
    },
    {
      "entropy": 6.264487791061401,
      "epoch": 1.2198499711482977,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048628273166459105,
      "loss": 6.0608,
      "mean_token_accuracy": 0.165188530087471,
      "num_tokens": 26797481.0,
      "step": 10570
    },
    {
      "entropy": 6.337562322616577,
      "epoch": 1.2204270051933064,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00048626854046563876,
      "loss": 6.2219,
      "mean_token_accuracy": 0.1599225729703903,
      "num_tokens": 26810644.0,
      "step": 10575
    },
    {
      "entropy": 6.375046730041504,
      "epoch": 1.221004039238315,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048625434216077404,
      "loss": 6.1245,
      "mean_token_accuracy": 0.16600160151720048,
      "num_tokens": 26822545.0,
      "step": 10580
    },
    {
      "entropy": 6.209045267105102,
      "epoch": 1.2215810732833237,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048624013675047453,
      "loss": 6.168,
      "mean_token_accuracy": 0.1631156712770462,
      "num_tokens": 26836221.0,
      "step": 10585
    },
    {
      "entropy": 6.313969373703003,
      "epoch": 1.2221581073283323,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048622592423521783,
      "loss": 6.2109,
      "mean_token_accuracy": 0.1565389335155487,
      "num_tokens": 26848543.0,
      "step": 10590
    },
    {
      "entropy": 6.435025596618653,
      "epoch": 1.222735141373341,
      "grad_norm": 1.125,
      "learning_rate": 0.0004862117046154822,
      "loss": 6.177,
      "mean_token_accuracy": 0.16341440230607987,
      "num_tokens": 26859833.0,
      "step": 10595
    },
    {
      "entropy": 6.351103830337524,
      "epoch": 1.2233121754183496,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048619747789174577,
      "loss": 6.1531,
      "mean_token_accuracy": 0.1668863445520401,
      "num_tokens": 26873450.0,
      "step": 10600
    },
    {
      "entropy": 6.2641448974609375,
      "epoch": 1.2238892094633584,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004861832440644871,
      "loss": 6.1458,
      "mean_token_accuracy": 0.1593625470995903,
      "num_tokens": 26886545.0,
      "step": 10605
    },
    {
      "entropy": 6.28742151260376,
      "epoch": 1.224466243508367,
      "grad_norm": 1.5,
      "learning_rate": 0.00048616900313418486,
      "loss": 6.0731,
      "mean_token_accuracy": 0.16876358091831206,
      "num_tokens": 26898394.0,
      "step": 10610
    },
    {
      "entropy": 6.3268989562988285,
      "epoch": 1.2250432775533757,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048615475510131815,
      "loss": 6.1404,
      "mean_token_accuracy": 0.16805960536003112,
      "num_tokens": 26910172.0,
      "step": 10615
    },
    {
      "entropy": 6.253440999984742,
      "epoch": 1.2256203115983844,
      "grad_norm": 2.046875,
      "learning_rate": 0.00048614049996636614,
      "loss": 6.0143,
      "mean_token_accuracy": 0.17637682110071182,
      "num_tokens": 26922496.0,
      "step": 10620
    },
    {
      "entropy": 6.295119285583496,
      "epoch": 1.226197345643393,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048612623772980825,
      "loss": 6.0815,
      "mean_token_accuracy": 0.17071359008550643,
      "num_tokens": 26935264.0,
      "step": 10625
    },
    {
      "entropy": 6.306150913238525,
      "epoch": 1.2267743796884016,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048611196839212426,
      "loss": 6.0947,
      "mean_token_accuracy": 0.1753847971558571,
      "num_tokens": 26949006.0,
      "step": 10630
    },
    {
      "entropy": 6.194625997543335,
      "epoch": 1.2273514137334103,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000486097691953794,
      "loss": 6.0233,
      "mean_token_accuracy": 0.16946956366300583,
      "num_tokens": 26961305.0,
      "step": 10635
    },
    {
      "entropy": 6.303498363494873,
      "epoch": 1.227928447778419,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048608340841529784,
      "loss": 6.1697,
      "mean_token_accuracy": 0.16096686720848083,
      "num_tokens": 26973445.0,
      "step": 10640
    },
    {
      "entropy": 6.2427514553070065,
      "epoch": 1.2285054818234276,
      "grad_norm": 1.4375,
      "learning_rate": 0.000486069117777116,
      "loss": 6.1154,
      "mean_token_accuracy": 0.16666444391012192,
      "num_tokens": 26986286.0,
      "step": 10645
    },
    {
      "entropy": 6.36623592376709,
      "epoch": 1.2290825158684362,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004860548200397293,
      "loss": 6.1561,
      "mean_token_accuracy": 0.16069630533456802,
      "num_tokens": 26999281.0,
      "step": 10650
    },
    {
      "entropy": 6.38177375793457,
      "epoch": 1.2296595499134448,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048604051520361846,
      "loss": 6.2931,
      "mean_token_accuracy": 0.15522295832633973,
      "num_tokens": 27011916.0,
      "step": 10655
    },
    {
      "entropy": 6.38488450050354,
      "epoch": 1.2302365839584535,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048602620326926484,
      "loss": 6.1754,
      "mean_token_accuracy": 0.1675037696957588,
      "num_tokens": 27024442.0,
      "step": 10660
    },
    {
      "entropy": 6.414252138137817,
      "epoch": 1.2308136180034621,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048601188423714965,
      "loss": 6.2689,
      "mean_token_accuracy": 0.15624295622110368,
      "num_tokens": 27037978.0,
      "step": 10665
    },
    {
      "entropy": 6.334872579574585,
      "epoch": 1.231390652048471,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004859975581077545,
      "loss": 6.2598,
      "mean_token_accuracy": 0.15260412395000458,
      "num_tokens": 27049447.0,
      "step": 10670
    },
    {
      "entropy": 6.296382093429566,
      "epoch": 1.2319676860934794,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004859832248815614,
      "loss": 6.1431,
      "mean_token_accuracy": 0.17176230102777482,
      "num_tokens": 27062961.0,
      "step": 10675
    },
    {
      "entropy": 6.392310857772827,
      "epoch": 1.2325447201384883,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004859688845590522,
      "loss": 6.2079,
      "mean_token_accuracy": 0.16266993284225464,
      "num_tokens": 27076297.0,
      "step": 10680
    },
    {
      "entropy": 6.316580009460449,
      "epoch": 1.233121754183497,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048595453714070944,
      "loss": 6.2308,
      "mean_token_accuracy": 0.15837038457393646,
      "num_tokens": 27088496.0,
      "step": 10685
    },
    {
      "entropy": 6.401510953903198,
      "epoch": 1.2336987882285055,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004859401826270156,
      "loss": 6.2055,
      "mean_token_accuracy": 0.15987962186336518,
      "num_tokens": 27100695.0,
      "step": 10690
    },
    {
      "entropy": 6.322301721572876,
      "epoch": 1.2342758222735142,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004859258210184536,
      "loss": 6.208,
      "mean_token_accuracy": 0.16477376520633696,
      "num_tokens": 27113703.0,
      "step": 10695
    },
    {
      "entropy": 6.351233863830567,
      "epoch": 1.2348528563185228,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048591145231550623,
      "loss": 6.2269,
      "mean_token_accuracy": 0.15568701475858687,
      "num_tokens": 27128669.0,
      "step": 10700
    },
    {
      "entropy": 6.387229156494141,
      "epoch": 1.2354298903635315,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048589707651865697,
      "loss": 6.1804,
      "mean_token_accuracy": 0.16083226203918458,
      "num_tokens": 27141311.0,
      "step": 10705
    },
    {
      "entropy": 6.387057733535767,
      "epoch": 1.23600692440854,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004858826936283893,
      "loss": 6.1847,
      "mean_token_accuracy": 0.16173309981822967,
      "num_tokens": 27152935.0,
      "step": 10710
    },
    {
      "entropy": 6.302968740463257,
      "epoch": 1.2365839584535487,
      "grad_norm": 1.75,
      "learning_rate": 0.000485868303645187,
      "loss": 6.0807,
      "mean_token_accuracy": 0.16808868646621705,
      "num_tokens": 27166775.0,
      "step": 10715
    },
    {
      "entropy": 6.193493843078613,
      "epoch": 1.2371609924985574,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048585390656953397,
      "loss": 6.1429,
      "mean_token_accuracy": 0.16636423021554947,
      "num_tokens": 27179406.0,
      "step": 10720
    },
    {
      "entropy": 6.3158278465271,
      "epoch": 1.237738026543566,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004858395024019147,
      "loss": 6.1274,
      "mean_token_accuracy": 0.1683789536356926,
      "num_tokens": 27191562.0,
      "step": 10725
    },
    {
      "entropy": 6.365101718902588,
      "epoch": 1.2383150605885747,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004858250911428133,
      "loss": 6.1614,
      "mean_token_accuracy": 0.16290413737297058,
      "num_tokens": 27203827.0,
      "step": 10730
    },
    {
      "entropy": 6.311492300033569,
      "epoch": 1.2388920946335833,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004858106727927148,
      "loss": 6.1333,
      "mean_token_accuracy": 0.16150906234979628,
      "num_tokens": 27217645.0,
      "step": 10735
    },
    {
      "entropy": 6.31927604675293,
      "epoch": 1.239469128678592,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000485796247352104,
      "loss": 6.1873,
      "mean_token_accuracy": 0.1700813129544258,
      "num_tokens": 27230794.0,
      "step": 10740
    },
    {
      "entropy": 6.310563325881958,
      "epoch": 1.2400461627236008,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004857818148214662,
      "loss": 6.2377,
      "mean_token_accuracy": 0.15220990478992463,
      "num_tokens": 27243965.0,
      "step": 10745
    },
    {
      "entropy": 6.406279706954956,
      "epoch": 1.2406231967686094,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004857673752012866,
      "loss": 6.1923,
      "mean_token_accuracy": 0.16095540374517442,
      "num_tokens": 27256951.0,
      "step": 10750
    },
    {
      "entropy": 6.376666307449341,
      "epoch": 1.241200230813618,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048575292849205114,
      "loss": 6.2496,
      "mean_token_accuracy": 0.1572696790099144,
      "num_tokens": 27268556.0,
      "step": 10755
    },
    {
      "entropy": 6.281139230728149,
      "epoch": 1.2417772648586267,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004857384746942456,
      "loss": 6.0908,
      "mean_token_accuracy": 0.15756804049015044,
      "num_tokens": 27280367.0,
      "step": 10760
    },
    {
      "entropy": 6.36142110824585,
      "epoch": 1.2423542989036354,
      "grad_norm": 1.625,
      "learning_rate": 0.0004857240138083562,
      "loss": 6.1589,
      "mean_token_accuracy": 0.17009591162204743,
      "num_tokens": 27293727.0,
      "step": 10765
    },
    {
      "entropy": 6.4208721160888675,
      "epoch": 1.242931332948644,
      "grad_norm": 1.375,
      "learning_rate": 0.0004857095458348692,
      "loss": 6.206,
      "mean_token_accuracy": 0.16630731970071794,
      "num_tokens": 27306677.0,
      "step": 10770
    },
    {
      "entropy": 6.421437358856201,
      "epoch": 1.2435083669936526,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048569507077427134,
      "loss": 6.1739,
      "mean_token_accuracy": 0.16194156408309937,
      "num_tokens": 27319317.0,
      "step": 10775
    },
    {
      "entropy": 6.349999570846558,
      "epoch": 1.2440854010386613,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004856805886270494,
      "loss": 6.154,
      "mean_token_accuracy": 0.1680565133690834,
      "num_tokens": 27332021.0,
      "step": 10780
    },
    {
      "entropy": 6.234192276000977,
      "epoch": 1.24466243508367,
      "grad_norm": 1.375,
      "learning_rate": 0.0004856660993936905,
      "loss": 6.0452,
      "mean_token_accuracy": 0.17493621557950972,
      "num_tokens": 27343972.0,
      "step": 10785
    },
    {
      "entropy": 6.339166927337646,
      "epoch": 1.2452394691286786,
      "grad_norm": 1.375,
      "learning_rate": 0.000485651603074682,
      "loss": 6.2345,
      "mean_token_accuracy": 0.15476678535342217,
      "num_tokens": 27356712.0,
      "step": 10790
    },
    {
      "entropy": 6.254733991622925,
      "epoch": 1.2458165031736872,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004856370996705115,
      "loss": 6.1038,
      "mean_token_accuracy": 0.16964900642633438,
      "num_tokens": 27369028.0,
      "step": 10795
    },
    {
      "entropy": 6.296901226043701,
      "epoch": 1.2463935372186958,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004856225891816667,
      "loss": 6.1239,
      "mean_token_accuracy": 0.1664896309375763,
      "num_tokens": 27380789.0,
      "step": 10800
    },
    {
      "entropy": 6.33326268196106,
      "epoch": 1.2469705712637045,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004856080716086358,
      "loss": 6.0895,
      "mean_token_accuracy": 0.1629282146692276,
      "num_tokens": 27394098.0,
      "step": 10805
    },
    {
      "entropy": 6.316369342803955,
      "epoch": 1.2475476053087133,
      "grad_norm": 1.1875,
      "learning_rate": 0.000485593546951907,
      "loss": 6.187,
      "mean_token_accuracy": 0.16884668171405792,
      "num_tokens": 27407270.0,
      "step": 10810
    },
    {
      "entropy": 6.313498497009277,
      "epoch": 1.2481246393537218,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004855790152119688,
      "loss": 6.1572,
      "mean_token_accuracy": 0.16334105879068375,
      "num_tokens": 27420397.0,
      "step": 10815
    },
    {
      "entropy": 6.3574518203735355,
      "epoch": 1.2487016733987306,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004855644763893102,
      "loss": 6.1963,
      "mean_token_accuracy": 0.16616567373275756,
      "num_tokens": 27433007.0,
      "step": 10820
    },
    {
      "entropy": 6.296497392654419,
      "epoch": 1.2492787074437393,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048554993048441987,
      "loss": 6.177,
      "mean_token_accuracy": 0.15731592178344728,
      "num_tokens": 27446683.0,
      "step": 10825
    },
    {
      "entropy": 6.3187668800354,
      "epoch": 1.249855741488748,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004855353774977873,
      "loss": 6.1995,
      "mean_token_accuracy": 0.15644395351409912,
      "num_tokens": 27459298.0,
      "step": 10830
    },
    {
      "entropy": 6.342638778686523,
      "epoch": 1.2504327755337565,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048552081742990184,
      "loss": 6.1643,
      "mean_token_accuracy": 0.1627936080098152,
      "num_tokens": 27472287.0,
      "step": 10835
    },
    {
      "entropy": 6.380952024459839,
      "epoch": 1.2510098095787652,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048550625028125327,
      "loss": 6.1657,
      "mean_token_accuracy": 0.16333349347114562,
      "num_tokens": 27484808.0,
      "step": 10840
    },
    {
      "entropy": 6.348484134674072,
      "epoch": 1.2515868436237738,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004854916760523315,
      "loss": 6.1697,
      "mean_token_accuracy": 0.16139520555734635,
      "num_tokens": 27496980.0,
      "step": 10845
    },
    {
      "entropy": 6.282923889160156,
      "epoch": 1.2521638776687825,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048547709474362684,
      "loss": 6.1984,
      "mean_token_accuracy": 0.15698642879724503,
      "num_tokens": 27509969.0,
      "step": 10850
    },
    {
      "entropy": 6.34557409286499,
      "epoch": 1.252740911713791,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004854625063556296,
      "loss": 6.1609,
      "mean_token_accuracy": 0.16385020911693574,
      "num_tokens": 27523410.0,
      "step": 10855
    },
    {
      "entropy": 6.387068557739258,
      "epoch": 1.2533179457587997,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004854479108888305,
      "loss": 6.1361,
      "mean_token_accuracy": 0.1640760526061058,
      "num_tokens": 27535183.0,
      "step": 10860
    },
    {
      "entropy": 6.349454069137574,
      "epoch": 1.2538949798038084,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048543330834372047,
      "loss": 6.2117,
      "mean_token_accuracy": 0.1553926631808281,
      "num_tokens": 27548028.0,
      "step": 10865
    },
    {
      "entropy": 6.322198343276978,
      "epoch": 1.254472013848817,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048541869872079064,
      "loss": 6.1347,
      "mean_token_accuracy": 0.16066425293684006,
      "num_tokens": 27561302.0,
      "step": 10870
    },
    {
      "entropy": 6.269003582000733,
      "epoch": 1.2550490478938257,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004854040820205324,
      "loss": 6.042,
      "mean_token_accuracy": 0.1691621169447899,
      "num_tokens": 27572933.0,
      "step": 10875
    },
    {
      "entropy": 6.346168088912964,
      "epoch": 1.2556260819388343,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004853894582434373,
      "loss": 6.2337,
      "mean_token_accuracy": 0.1564677447080612,
      "num_tokens": 27586373.0,
      "step": 10880
    },
    {
      "entropy": 6.346892642974853,
      "epoch": 1.2562031159838432,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004853748273899974,
      "loss": 6.0814,
      "mean_token_accuracy": 0.1748109444975853,
      "num_tokens": 27599236.0,
      "step": 10885
    },
    {
      "entropy": 6.302681541442871,
      "epoch": 1.2567801500288516,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004853601894607046,
      "loss": 6.0979,
      "mean_token_accuracy": 0.16508190482854843,
      "num_tokens": 27611904.0,
      "step": 10890
    },
    {
      "entropy": 6.214655542373658,
      "epoch": 1.2573571840738604,
      "grad_norm": 2.84375,
      "learning_rate": 0.0004853455444560514,
      "loss": 6.056,
      "mean_token_accuracy": 0.16355132162570954,
      "num_tokens": 27625778.0,
      "step": 10895
    },
    {
      "entropy": 6.32364501953125,
      "epoch": 1.257934218118869,
      "grad_norm": 3.515625,
      "learning_rate": 0.0004853308923765302,
      "loss": 6.1785,
      "mean_token_accuracy": 0.17193194329738617,
      "num_tokens": 27638359.0,
      "step": 10900
    },
    {
      "entropy": 6.329800653457641,
      "epoch": 1.2585112521638777,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000485316233222634,
      "loss": 6.1046,
      "mean_token_accuracy": 0.16803177893161775,
      "num_tokens": 27651107.0,
      "step": 10905
    },
    {
      "entropy": 6.320554113388061,
      "epoch": 1.2590882862088864,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004853015669948557,
      "loss": 6.2178,
      "mean_token_accuracy": 0.16287010610103608,
      "num_tokens": 27664060.0,
      "step": 10910
    },
    {
      "entropy": 6.312152290344239,
      "epoch": 1.259665320253895,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048528689369368863,
      "loss": 6.1404,
      "mean_token_accuracy": 0.16302927285432817,
      "num_tokens": 27676678.0,
      "step": 10915
    },
    {
      "entropy": 6.390170764923096,
      "epoch": 1.2602423542989036,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004852722133196264,
      "loss": 6.2006,
      "mean_token_accuracy": 0.15728381127119065,
      "num_tokens": 27689696.0,
      "step": 10920
    },
    {
      "entropy": 6.302979755401611,
      "epoch": 1.2608193883439123,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004852575258731627,
      "loss": 6.0878,
      "mean_token_accuracy": 0.16694856733083724,
      "num_tokens": 27702144.0,
      "step": 10925
    },
    {
      "entropy": 6.359341430664062,
      "epoch": 1.261396422388921,
      "grad_norm": 1.625,
      "learning_rate": 0.0004852428313547916,
      "loss": 6.2068,
      "mean_token_accuracy": 0.1572053074836731,
      "num_tokens": 27714618.0,
      "step": 10930
    },
    {
      "entropy": 6.324438142776489,
      "epoch": 1.2619734564339296,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048522812976500726,
      "loss": 6.1403,
      "mean_token_accuracy": 0.16266688853502273,
      "num_tokens": 27727002.0,
      "step": 10935
    },
    {
      "entropy": 6.310562229156494,
      "epoch": 1.2625504904789382,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048521342110430417,
      "loss": 6.2034,
      "mean_token_accuracy": 0.1592251852154732,
      "num_tokens": 27739506.0,
      "step": 10940
    },
    {
      "entropy": 6.297001934051513,
      "epoch": 1.2631275245239468,
      "grad_norm": 1.734375,
      "learning_rate": 0.00048519870537317713,
      "loss": 6.0444,
      "mean_token_accuracy": 0.16531992256641387,
      "num_tokens": 27751959.0,
      "step": 10945
    },
    {
      "entropy": 6.2629921436309814,
      "epoch": 1.2637045585689557,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048518398257212103,
      "loss": 6.0776,
      "mean_token_accuracy": 0.16900534927845,
      "num_tokens": 27762938.0,
      "step": 10950
    },
    {
      "entropy": 6.365654325485229,
      "epoch": 1.2642815926139641,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004851692527016311,
      "loss": 6.2081,
      "mean_token_accuracy": 0.15454971194267272,
      "num_tokens": 27776288.0,
      "step": 10955
    },
    {
      "entropy": 6.331540727615357,
      "epoch": 1.264858626658973,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004851545157622027,
      "loss": 6.2372,
      "mean_token_accuracy": 0.15178136229515077,
      "num_tokens": 27788311.0,
      "step": 10960
    },
    {
      "entropy": 6.3813409328460695,
      "epoch": 1.2654356607039816,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004851397717543316,
      "loss": 6.1609,
      "mean_token_accuracy": 0.15812337547540664,
      "num_tokens": 27799589.0,
      "step": 10965
    },
    {
      "entropy": 6.274319410324097,
      "epoch": 1.2660126947489903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004851250206785137,
      "loss": 6.0588,
      "mean_token_accuracy": 0.1667439490556717,
      "num_tokens": 27813057.0,
      "step": 10970
    },
    {
      "entropy": 6.277303171157837,
      "epoch": 1.266589728793999,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048511026253524503,
      "loss": 6.1465,
      "mean_token_accuracy": 0.16044363081455232,
      "num_tokens": 27826465.0,
      "step": 10975
    },
    {
      "entropy": 6.309837818145752,
      "epoch": 1.2671667628390075,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004850954973250221,
      "loss": 6.1432,
      "mean_token_accuracy": 0.16380210518836974,
      "num_tokens": 27837157.0,
      "step": 10980
    },
    {
      "entropy": 6.386126470565796,
      "epoch": 1.2677437968840162,
      "grad_norm": 2.03125,
      "learning_rate": 0.00048508072504834144,
      "loss": 6.2431,
      "mean_token_accuracy": 0.16418514400720596,
      "num_tokens": 27850101.0,
      "step": 10985
    },
    {
      "entropy": 6.272424221038818,
      "epoch": 1.2683208309290248,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048506594570569997,
      "loss": 6.1588,
      "mean_token_accuracy": 0.16430216729640962,
      "num_tokens": 27862786.0,
      "step": 10990
    },
    {
      "entropy": 6.425026226043701,
      "epoch": 1.2688978649740335,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004850511592975947,
      "loss": 6.1663,
      "mean_token_accuracy": 0.16461049318313598,
      "num_tokens": 27875196.0,
      "step": 10995
    },
    {
      "entropy": 6.349469900131226,
      "epoch": 1.269474899019042,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004850363658245231,
      "loss": 6.2633,
      "mean_token_accuracy": 0.1542629912495613,
      "num_tokens": 27888348.0,
      "step": 11000
    },
    {
      "entropy": 6.35367865562439,
      "epoch": 1.2700519330640507,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004850215652869826,
      "loss": 6.1909,
      "mean_token_accuracy": 0.16688053011894227,
      "num_tokens": 27901301.0,
      "step": 11005
    },
    {
      "entropy": 6.3888359546661375,
      "epoch": 1.2706289671090594,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004850067576854711,
      "loss": 6.2115,
      "mean_token_accuracy": 0.15654693841934203,
      "num_tokens": 27915665.0,
      "step": 11010
    },
    {
      "entropy": 6.297394275665283,
      "epoch": 1.271206001154068,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004849919430204867,
      "loss": 6.1552,
      "mean_token_accuracy": 0.16020799279212952,
      "num_tokens": 27928333.0,
      "step": 11015
    },
    {
      "entropy": 6.3494425296783445,
      "epoch": 1.2717830351990767,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004849771212925275,
      "loss": 6.0888,
      "mean_token_accuracy": 0.16375294476747512,
      "num_tokens": 27940429.0,
      "step": 11020
    },
    {
      "entropy": 6.313184547424316,
      "epoch": 1.2723600692440855,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00048496229250209223,
      "loss": 6.2169,
      "mean_token_accuracy": 0.1601503983139992,
      "num_tokens": 27953846.0,
      "step": 11025
    },
    {
      "entropy": 6.332328128814697,
      "epoch": 1.272937103289094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004849474566496795,
      "loss": 6.0899,
      "mean_token_accuracy": 0.16779401600360871,
      "num_tokens": 27966393.0,
      "step": 11030
    },
    {
      "entropy": 6.264288711547851,
      "epoch": 1.2735141373341028,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004849326137357883,
      "loss": 6.0361,
      "mean_token_accuracy": 0.1776391953229904,
      "num_tokens": 27979928.0,
      "step": 11035
    },
    {
      "entropy": 6.353927135467529,
      "epoch": 1.2740911713791114,
      "grad_norm": 1.359375,
      "learning_rate": 0.000484917763760918,
      "loss": 6.1312,
      "mean_token_accuracy": 0.16302897185087203,
      "num_tokens": 27991450.0,
      "step": 11040
    },
    {
      "entropy": 6.341830396652222,
      "epoch": 1.27466820542412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048490290672556784,
      "loss": 6.1625,
      "mean_token_accuracy": 0.16186770498752595,
      "num_tokens": 28003696.0,
      "step": 11045
    },
    {
      "entropy": 6.251212215423584,
      "epoch": 1.2752452394691287,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004848880426302378,
      "loss": 6.1039,
      "mean_token_accuracy": 0.167728927731514,
      "num_tokens": 28015886.0,
      "step": 11050
    },
    {
      "entropy": 6.331211757659912,
      "epoch": 1.2758222735141374,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004848731714754277,
      "loss": 6.1559,
      "mean_token_accuracy": 0.16511764675378798,
      "num_tokens": 28029552.0,
      "step": 11055
    },
    {
      "entropy": 6.295982503890992,
      "epoch": 1.276399307559146,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004848582932616377,
      "loss": 6.2124,
      "mean_token_accuracy": 0.16154432594776152,
      "num_tokens": 28042275.0,
      "step": 11060
    },
    {
      "entropy": 6.2871945858001705,
      "epoch": 1.2769763416041546,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004848434079893682,
      "loss": 6.101,
      "mean_token_accuracy": 0.17026014775037765,
      "num_tokens": 28054840.0,
      "step": 11065
    },
    {
      "entropy": 6.31745138168335,
      "epoch": 1.2775533756491633,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004848285156591201,
      "loss": 6.2333,
      "mean_token_accuracy": 0.1593561977148056,
      "num_tokens": 28068052.0,
      "step": 11070
    },
    {
      "entropy": 6.420611810684204,
      "epoch": 1.278130409694172,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048481361627139384,
      "loss": 6.2162,
      "mean_token_accuracy": 0.15895957499742508,
      "num_tokens": 28081157.0,
      "step": 11075
    },
    {
      "entropy": 6.396226978302002,
      "epoch": 1.2787074437391805,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048479870982669096,
      "loss": 6.257,
      "mean_token_accuracy": 0.16313387006521224,
      "num_tokens": 28093842.0,
      "step": 11080
    },
    {
      "entropy": 6.40015058517456,
      "epoch": 1.2792844777841892,
      "grad_norm": 1.375,
      "learning_rate": 0.00048478379632551266,
      "loss": 6.1188,
      "mean_token_accuracy": 0.17054009437561035,
      "num_tokens": 28107529.0,
      "step": 11085
    },
    {
      "entropy": 6.291486501693726,
      "epoch": 1.279861511829198,
      "grad_norm": 1.125,
      "learning_rate": 0.0004847688757683606,
      "loss": 6.2009,
      "mean_token_accuracy": 0.16051672995090485,
      "num_tokens": 28119650.0,
      "step": 11090
    },
    {
      "entropy": 6.270731353759766,
      "epoch": 1.2804385458742065,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004847539481557366,
      "loss": 6.1499,
      "mean_token_accuracy": 0.17015368193387986,
      "num_tokens": 28131209.0,
      "step": 11095
    },
    {
      "entropy": 6.332563591003418,
      "epoch": 1.2810155799192153,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048473901348814277,
      "loss": 6.1685,
      "mean_token_accuracy": 0.1653625950217247,
      "num_tokens": 28143624.0,
      "step": 11100
    },
    {
      "entropy": 6.43041090965271,
      "epoch": 1.281592613964224,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004847240717660814,
      "loss": 6.1952,
      "mean_token_accuracy": 0.16168740540742874,
      "num_tokens": 28156752.0,
      "step": 11105
    },
    {
      "entropy": 6.225031948089599,
      "epoch": 1.2821696480092326,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048470912299005493,
      "loss": 6.0443,
      "mean_token_accuracy": 0.17290171086788178,
      "num_tokens": 28169288.0,
      "step": 11110
    },
    {
      "entropy": 6.259183263778686,
      "epoch": 1.2827466820542412,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048469416716056635,
      "loss": 6.1553,
      "mean_token_accuracy": 0.16693965941667557,
      "num_tokens": 28181739.0,
      "step": 11115
    },
    {
      "entropy": 6.299275255203247,
      "epoch": 1.28332371609925,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004846792042781187,
      "loss": 6.1047,
      "mean_token_accuracy": 0.16790862381458282,
      "num_tokens": 28193861.0,
      "step": 11120
    },
    {
      "entropy": 6.28367805480957,
      "epoch": 1.2839007501442585,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048466423434321513,
      "loss": 6.03,
      "mean_token_accuracy": 0.1750580132007599,
      "num_tokens": 28205453.0,
      "step": 11125
    },
    {
      "entropy": 6.2496013164520265,
      "epoch": 1.2844777841892672,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048464925735635907,
      "loss": 6.1232,
      "mean_token_accuracy": 0.16223495900630952,
      "num_tokens": 28217554.0,
      "step": 11130
    },
    {
      "entropy": 6.396459007263184,
      "epoch": 1.2850548182342758,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048463427331805445,
      "loss": 6.2473,
      "mean_token_accuracy": 0.158101187646389,
      "num_tokens": 28230091.0,
      "step": 11135
    },
    {
      "entropy": 6.279246997833252,
      "epoch": 1.2856318522792844,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004846192822288052,
      "loss": 6.1181,
      "mean_token_accuracy": 0.16716209203004836,
      "num_tokens": 28243285.0,
      "step": 11140
    },
    {
      "entropy": 6.256467676162719,
      "epoch": 1.286208886324293,
      "grad_norm": 1.25,
      "learning_rate": 0.0004846042840891155,
      "loss": 6.0446,
      "mean_token_accuracy": 0.17056648582220077,
      "num_tokens": 28256210.0,
      "step": 11145
    },
    {
      "entropy": 6.294454431533813,
      "epoch": 1.2867859203693017,
      "grad_norm": 1.875,
      "learning_rate": 0.0004845892788994899,
      "loss": 6.0873,
      "mean_token_accuracy": 0.1733658105134964,
      "num_tokens": 28268291.0,
      "step": 11150
    },
    {
      "entropy": 6.206016921997071,
      "epoch": 1.2873629544143104,
      "grad_norm": 1.25,
      "learning_rate": 0.0004845742666604329,
      "loss": 6.0748,
      "mean_token_accuracy": 0.17029971033334732,
      "num_tokens": 28281462.0,
      "step": 11155
    },
    {
      "entropy": 6.264611148834229,
      "epoch": 1.287939988459319,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048455924737244953,
      "loss": 6.2065,
      "mean_token_accuracy": 0.1581377938389778,
      "num_tokens": 28292766.0,
      "step": 11160
    },
    {
      "entropy": 6.273323631286621,
      "epoch": 1.2885170225043279,
      "grad_norm": 1.6875,
      "learning_rate": 0.00048454422103604505,
      "loss": 6.093,
      "mean_token_accuracy": 0.17237376421689987,
      "num_tokens": 28304500.0,
      "step": 11165
    },
    {
      "entropy": 6.362926387786866,
      "epoch": 1.2890940565493363,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004845291876517247,
      "loss": 6.1684,
      "mean_token_accuracy": 0.16565798670053483,
      "num_tokens": 28317262.0,
      "step": 11170
    },
    {
      "entropy": 6.215715932846069,
      "epoch": 1.2896710905943451,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048451414721999424,
      "loss": 6.0599,
      "mean_token_accuracy": 0.17366328686475754,
      "num_tokens": 28330313.0,
      "step": 11175
    },
    {
      "entropy": 6.207749509811402,
      "epoch": 1.2902481246393538,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048449909974135954,
      "loss": 6.0742,
      "mean_token_accuracy": 0.16498281359672545,
      "num_tokens": 28344058.0,
      "step": 11180
    },
    {
      "entropy": 6.276063632965088,
      "epoch": 1.2908251586843624,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004844840452163267,
      "loss": 6.1343,
      "mean_token_accuracy": 0.1621875062584877,
      "num_tokens": 28356664.0,
      "step": 11185
    },
    {
      "entropy": 6.368418121337891,
      "epoch": 1.291402192729371,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000484468983645402,
      "loss": 6.1943,
      "mean_token_accuracy": 0.1614609658718109,
      "num_tokens": 28369040.0,
      "step": 11190
    },
    {
      "entropy": 6.265853071212769,
      "epoch": 1.2919792267743797,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048445391502909213,
      "loss": 6.159,
      "mean_token_accuracy": 0.15824418812990187,
      "num_tokens": 28381788.0,
      "step": 11195
    },
    {
      "entropy": 6.321166610717773,
      "epoch": 1.2925562608193883,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048443883936790386,
      "loss": 6.1928,
      "mean_token_accuracy": 0.16284841299057007,
      "num_tokens": 28394531.0,
      "step": 11200
    },
    {
      "entropy": 6.334868097305298,
      "epoch": 1.293133294864397,
      "grad_norm": 1.859375,
      "learning_rate": 0.00048442375666234427,
      "loss": 6.1199,
      "mean_token_accuracy": 0.16238492727279663,
      "num_tokens": 28407766.0,
      "step": 11205
    },
    {
      "entropy": 6.2941638946533205,
      "epoch": 1.2937103289094056,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004844086669129206,
      "loss": 6.1924,
      "mean_token_accuracy": 0.16062195599079132,
      "num_tokens": 28420185.0,
      "step": 11210
    },
    {
      "entropy": 6.388847303390503,
      "epoch": 1.2942873629544143,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048439357012014045,
      "loss": 6.1364,
      "mean_token_accuracy": 0.16039270013570786,
      "num_tokens": 28432781.0,
      "step": 11215
    },
    {
      "entropy": 6.268585920333862,
      "epoch": 1.294864396999423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004843784662845116,
      "loss": 6.1237,
      "mean_token_accuracy": 0.16846691071987152,
      "num_tokens": 28446181.0,
      "step": 11220
    },
    {
      "entropy": 6.288886260986328,
      "epoch": 1.2954414310444315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000484363355406542,
      "loss": 6.0599,
      "mean_token_accuracy": 0.16910330504179,
      "num_tokens": 28457959.0,
      "step": 11225
    },
    {
      "entropy": 6.279118585586548,
      "epoch": 1.2960184650894404,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048434823748674,
      "loss": 6.1589,
      "mean_token_accuracy": 0.16250982731580735,
      "num_tokens": 28470656.0,
      "step": 11230
    },
    {
      "entropy": 6.280978965759277,
      "epoch": 1.2965954991344488,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048433311252561403,
      "loss": 6.181,
      "mean_token_accuracy": 0.1643814906477928,
      "num_tokens": 28483359.0,
      "step": 11235
    },
    {
      "entropy": 6.279931116104126,
      "epoch": 1.2971725331794577,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004843179805236728,
      "loss": 6.1298,
      "mean_token_accuracy": 0.16412907987833023,
      "num_tokens": 28495984.0,
      "step": 11240
    },
    {
      "entropy": 6.2893836975097654,
      "epoch": 1.297749567224466,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048430284148142516,
      "loss": 6.188,
      "mean_token_accuracy": 0.1599840596318245,
      "num_tokens": 28508629.0,
      "step": 11245
    },
    {
      "entropy": 6.3177934169769285,
      "epoch": 1.298326601269475,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004842876953993805,
      "loss": 6.2082,
      "mean_token_accuracy": 0.16367824673652648,
      "num_tokens": 28521952.0,
      "step": 11250
    },
    {
      "entropy": 6.315539360046387,
      "epoch": 1.2989036353144836,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004842725422780482,
      "loss": 6.1008,
      "mean_token_accuracy": 0.1647193670272827,
      "num_tokens": 28535210.0,
      "step": 11255
    },
    {
      "entropy": 6.339979457855224,
      "epoch": 1.2994806693594922,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004842573821179378,
      "loss": 6.216,
      "mean_token_accuracy": 0.15640299618244172,
      "num_tokens": 28547294.0,
      "step": 11260
    },
    {
      "entropy": 6.408569622039795,
      "epoch": 1.3000577034045009,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004842422149195593,
      "loss": 6.2359,
      "mean_token_accuracy": 0.15522149056196213,
      "num_tokens": 28558934.0,
      "step": 11265
    },
    {
      "entropy": 6.27472767829895,
      "epoch": 1.3006347374495095,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048422704068342294,
      "loss": 6.1151,
      "mean_token_accuracy": 0.15930722355842591,
      "num_tokens": 28571897.0,
      "step": 11270
    },
    {
      "entropy": 6.294620418548584,
      "epoch": 1.3012117714945182,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004842118594100389,
      "loss": 6.155,
      "mean_token_accuracy": 0.16027718931436538,
      "num_tokens": 28584320.0,
      "step": 11275
    },
    {
      "entropy": 6.250262880325318,
      "epoch": 1.3017888055395268,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048419667109991793,
      "loss": 6.0886,
      "mean_token_accuracy": 0.16860387623310089,
      "num_tokens": 28597018.0,
      "step": 11280
    },
    {
      "entropy": 6.248487567901611,
      "epoch": 1.3023658395845354,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048418147575357085,
      "loss": 6.0987,
      "mean_token_accuracy": 0.16950415819883347,
      "num_tokens": 28609145.0,
      "step": 11285
    },
    {
      "entropy": 6.32426905632019,
      "epoch": 1.302942873629544,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004841662733715087,
      "loss": 6.0984,
      "mean_token_accuracy": 0.16909146755933763,
      "num_tokens": 28622208.0,
      "step": 11290
    },
    {
      "entropy": 6.208732604980469,
      "epoch": 1.3035199076745527,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048415106395424294,
      "loss": 6.11,
      "mean_token_accuracy": 0.1594777375459671,
      "num_tokens": 28634719.0,
      "step": 11295
    },
    {
      "entropy": 6.29807767868042,
      "epoch": 1.3040969417195614,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048413584750228494,
      "loss": 6.164,
      "mean_token_accuracy": 0.1640383556485176,
      "num_tokens": 28647138.0,
      "step": 11300
    },
    {
      "entropy": 6.19341287612915,
      "epoch": 1.3046739757645702,
      "grad_norm": 1.25,
      "learning_rate": 0.00048412062401614653,
      "loss": 6.0934,
      "mean_token_accuracy": 0.16587843000888824,
      "num_tokens": 28660898.0,
      "step": 11305
    },
    {
      "entropy": 6.344057846069336,
      "epoch": 1.3052510098095786,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048410539349634,
      "loss": 6.2212,
      "mean_token_accuracy": 0.16715804785490035,
      "num_tokens": 28672898.0,
      "step": 11310
    },
    {
      "entropy": 6.31997652053833,
      "epoch": 1.3058280438545875,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048409015594337725,
      "loss": 6.2326,
      "mean_token_accuracy": 0.16915102154016495,
      "num_tokens": 28684608.0,
      "step": 11315
    },
    {
      "entropy": 6.339677190780639,
      "epoch": 1.3064050778995961,
      "grad_norm": 1.109375,
      "learning_rate": 0.000484074911357771,
      "loss": 6.1071,
      "mean_token_accuracy": 0.17026219218969346,
      "num_tokens": 28696867.0,
      "step": 11320
    },
    {
      "entropy": 6.284737062454224,
      "epoch": 1.3069821119446048,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048405965974003404,
      "loss": 6.1815,
      "mean_token_accuracy": 0.1607493944466114,
      "num_tokens": 28709533.0,
      "step": 11325
    },
    {
      "entropy": 6.273362398147583,
      "epoch": 1.3075591459896134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048404440109067927,
      "loss": 6.1495,
      "mean_token_accuracy": 0.15757483690977098,
      "num_tokens": 28721647.0,
      "step": 11330
    },
    {
      "entropy": 6.358157205581665,
      "epoch": 1.308136180034622,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004840291354102198,
      "loss": 6.1643,
      "mean_token_accuracy": 0.1579637810587883,
      "num_tokens": 28735261.0,
      "step": 11335
    },
    {
      "entropy": 6.355404090881348,
      "epoch": 1.3087132140796307,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004840138626991693,
      "loss": 6.1813,
      "mean_token_accuracy": 0.16192290782928467,
      "num_tokens": 28747930.0,
      "step": 11340
    },
    {
      "entropy": 6.281838083267212,
      "epoch": 1.3092902481246393,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004839985829580413,
      "loss": 6.1725,
      "mean_token_accuracy": 0.16236553341150284,
      "num_tokens": 28760497.0,
      "step": 11345
    },
    {
      "entropy": 6.378918027877807,
      "epoch": 1.309867282169648,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048398329618734977,
      "loss": 6.1821,
      "mean_token_accuracy": 0.16250278949737548,
      "num_tokens": 28772787.0,
      "step": 11350
    },
    {
      "entropy": 6.279808425903321,
      "epoch": 1.3104443162146566,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004839680023876089,
      "loss": 6.1225,
      "mean_token_accuracy": 0.16467889100313188,
      "num_tokens": 28784901.0,
      "step": 11355
    },
    {
      "entropy": 6.267178392410278,
      "epoch": 1.3110213502596653,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004839527015593331,
      "loss": 6.1847,
      "mean_token_accuracy": 0.155230500549078,
      "num_tokens": 28797376.0,
      "step": 11360
    },
    {
      "entropy": 6.296201229095459,
      "epoch": 1.311598384304674,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048393739370303684,
      "loss": 6.0965,
      "mean_token_accuracy": 0.16835566163063048,
      "num_tokens": 28808805.0,
      "step": 11365
    },
    {
      "entropy": 6.336475372314453,
      "epoch": 1.3121754183496828,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004839220788192353,
      "loss": 6.236,
      "mean_token_accuracy": 0.15595455914735795,
      "num_tokens": 28821734.0,
      "step": 11370
    },
    {
      "entropy": 6.313124322891236,
      "epoch": 1.3127524523946912,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048390675690844335,
      "loss": 6.1081,
      "mean_token_accuracy": 0.15779682695865632,
      "num_tokens": 28834864.0,
      "step": 11375
    },
    {
      "entropy": 6.213588857650757,
      "epoch": 1.3133294864397,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004838914279711764,
      "loss": 6.1031,
      "mean_token_accuracy": 0.16101833879947663,
      "num_tokens": 28848154.0,
      "step": 11380
    },
    {
      "entropy": 6.278816747665405,
      "epoch": 1.3139065204847085,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048387609200795003,
      "loss": 6.1566,
      "mean_token_accuracy": 0.16542939245700836,
      "num_tokens": 28861562.0,
      "step": 11385
    },
    {
      "entropy": 6.3024074077606205,
      "epoch": 1.3144835545297173,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048386074901928,
      "loss": 6.1507,
      "mean_token_accuracy": 0.1619145765900612,
      "num_tokens": 28873921.0,
      "step": 11390
    },
    {
      "entropy": 6.294253396987915,
      "epoch": 1.315060588574726,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004838453990056825,
      "loss": 6.2042,
      "mean_token_accuracy": 0.15534769594669343,
      "num_tokens": 28886126.0,
      "step": 11395
    },
    {
      "entropy": 6.3025377750396725,
      "epoch": 1.3156376226197346,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048383004196767373,
      "loss": 6.0669,
      "mean_token_accuracy": 0.16867079287767411,
      "num_tokens": 28897980.0,
      "step": 11400
    },
    {
      "entropy": 6.36027307510376,
      "epoch": 1.3162146566647432,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004838146779057702,
      "loss": 6.1161,
      "mean_token_accuracy": 0.1627037927508354,
      "num_tokens": 28910105.0,
      "step": 11405
    },
    {
      "entropy": 6.22172999382019,
      "epoch": 1.3167916907097519,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004837993068204887,
      "loss": 6.1778,
      "mean_token_accuracy": 0.16126993000507356,
      "num_tokens": 28922004.0,
      "step": 11410
    },
    {
      "entropy": 6.219480228424072,
      "epoch": 1.3173687247547605,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048378392871234634,
      "loss": 6.0341,
      "mean_token_accuracy": 0.17347660958766936,
      "num_tokens": 28935990.0,
      "step": 11415
    },
    {
      "entropy": 6.357144784927368,
      "epoch": 1.3179457587997692,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004837685435818601,
      "loss": 6.1271,
      "mean_token_accuracy": 0.1658763289451599,
      "num_tokens": 28948274.0,
      "step": 11420
    },
    {
      "entropy": 6.311680173873901,
      "epoch": 1.3185227928447778,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004837531514295477,
      "loss": 6.228,
      "mean_token_accuracy": 0.155182783305645,
      "num_tokens": 28960101.0,
      "step": 11425
    },
    {
      "entropy": 6.284453725814819,
      "epoch": 1.3190998268897864,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004837377522559267,
      "loss": 6.1773,
      "mean_token_accuracy": 0.16812565177679062,
      "num_tokens": 28972408.0,
      "step": 11430
    },
    {
      "entropy": 6.337407922744751,
      "epoch": 1.319676860934795,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048372234606151507,
      "loss": 6.1459,
      "mean_token_accuracy": 0.16081255227327346,
      "num_tokens": 28985025.0,
      "step": 11435
    },
    {
      "entropy": 6.349707365036011,
      "epoch": 1.3202538949798037,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048370693284683106,
      "loss": 6.2225,
      "mean_token_accuracy": 0.16389241069555283,
      "num_tokens": 28997568.0,
      "step": 11440
    },
    {
      "entropy": 6.323847770690918,
      "epoch": 1.3208309290248126,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048369151261239303,
      "loss": 6.2415,
      "mean_token_accuracy": 0.15582263171672822,
      "num_tokens": 29009567.0,
      "step": 11445
    },
    {
      "entropy": 6.289740896224975,
      "epoch": 1.321407963069821,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004836760853587196,
      "loss": 6.1286,
      "mean_token_accuracy": 0.16066635698080062,
      "num_tokens": 29021403.0,
      "step": 11450
    },
    {
      "entropy": 6.277720832824707,
      "epoch": 1.3219849971148299,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048366065108632967,
      "loss": 6.1198,
      "mean_token_accuracy": 0.16584020256996154,
      "num_tokens": 29033430.0,
      "step": 11455
    },
    {
      "entropy": 6.319123935699463,
      "epoch": 1.3225620311598385,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048364520979574246,
      "loss": 6.1489,
      "mean_token_accuracy": 0.1609363466501236,
      "num_tokens": 29045659.0,
      "step": 11460
    },
    {
      "entropy": 6.234172105789185,
      "epoch": 1.3231390652048471,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048362976148747715,
      "loss": 6.1121,
      "mean_token_accuracy": 0.1620397835969925,
      "num_tokens": 29058962.0,
      "step": 11465
    },
    {
      "entropy": 6.299294567108154,
      "epoch": 1.3237160992498558,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004836143061620536,
      "loss": 6.0509,
      "mean_token_accuracy": 0.165127757191658,
      "num_tokens": 29072009.0,
      "step": 11470
    },
    {
      "entropy": 6.163762044906616,
      "epoch": 1.3242931332948644,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004835988438199914,
      "loss": 6.0393,
      "mean_token_accuracy": 0.17196400761604308,
      "num_tokens": 29084174.0,
      "step": 11475
    },
    {
      "entropy": 6.279996156692505,
      "epoch": 1.324870167339873,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004835833744618107,
      "loss": 6.1329,
      "mean_token_accuracy": 0.16213522255420684,
      "num_tokens": 29096189.0,
      "step": 11480
    },
    {
      "entropy": 6.312683248519898,
      "epoch": 1.3254472013848817,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004835678980880318,
      "loss": 6.0979,
      "mean_token_accuracy": 0.16390204429626465,
      "num_tokens": 29108212.0,
      "step": 11485
    },
    {
      "entropy": 6.309130334854126,
      "epoch": 1.3260242354298903,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004835524146991753,
      "loss": 6.1004,
      "mean_token_accuracy": 0.16586587578058243,
      "num_tokens": 29121058.0,
      "step": 11490
    },
    {
      "entropy": 6.241234064102173,
      "epoch": 1.326601269474899,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048353692429576185,
      "loss": 6.188,
      "mean_token_accuracy": 0.16660408526659012,
      "num_tokens": 29132981.0,
      "step": 11495
    },
    {
      "entropy": 6.332604503631591,
      "epoch": 1.3271783035199076,
      "grad_norm": 1.25,
      "learning_rate": 0.0004835214268783126,
      "loss": 6.1632,
      "mean_token_accuracy": 0.15822002291679382,
      "num_tokens": 29145143.0,
      "step": 11500
    },
    {
      "entropy": 6.333203220367432,
      "epoch": 1.3277553375649163,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048350592244734866,
      "loss": 6.0906,
      "mean_token_accuracy": 0.16814986616373062,
      "num_tokens": 29157548.0,
      "step": 11505
    },
    {
      "entropy": 6.203010225296021,
      "epoch": 1.3283323716099251,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004834904110033917,
      "loss": 6.0297,
      "mean_token_accuracy": 0.16850072741508484,
      "num_tokens": 29170243.0,
      "step": 11510
    },
    {
      "entropy": 6.297244215011597,
      "epoch": 1.3289094056549335,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048347489254696327,
      "loss": 6.1258,
      "mean_token_accuracy": 0.16527609378099442,
      "num_tokens": 29181769.0,
      "step": 11515
    },
    {
      "entropy": 6.33937406539917,
      "epoch": 1.3294864396999424,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004834593670785854,
      "loss": 6.2181,
      "mean_token_accuracy": 0.15585944056510925,
      "num_tokens": 29194196.0,
      "step": 11520
    },
    {
      "entropy": 6.325269174575806,
      "epoch": 1.3300634737449508,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048344383459878024,
      "loss": 6.1271,
      "mean_token_accuracy": 0.1619534283876419,
      "num_tokens": 29206101.0,
      "step": 11525
    },
    {
      "entropy": 6.217943239212036,
      "epoch": 1.3306405077899597,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00048342829510807024,
      "loss": 6.0847,
      "mean_token_accuracy": 0.16871753782033921,
      "num_tokens": 29217851.0,
      "step": 11530
    },
    {
      "entropy": 6.189042425155639,
      "epoch": 1.3312175418349683,
      "grad_norm": 1.234375,
      "learning_rate": 0.000483412748606978,
      "loss": 6.0601,
      "mean_token_accuracy": 0.17444779425859452,
      "num_tokens": 29231497.0,
      "step": 11535
    },
    {
      "entropy": 6.353026437759399,
      "epoch": 1.331794575879977,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004833971950960266,
      "loss": 6.208,
      "mean_token_accuracy": 0.1626723363995552,
      "num_tokens": 29245227.0,
      "step": 11540
    },
    {
      "entropy": 6.34786024093628,
      "epoch": 1.3323716099249856,
      "grad_norm": 2.578125,
      "learning_rate": 0.0004833816345757391,
      "loss": 6.1389,
      "mean_token_accuracy": 0.159043388068676,
      "num_tokens": 29259403.0,
      "step": 11545
    },
    {
      "entropy": 6.323790502548218,
      "epoch": 1.3329486439699942,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004833660670466387,
      "loss": 6.1127,
      "mean_token_accuracy": 0.16409413069486617,
      "num_tokens": 29270938.0,
      "step": 11550
    },
    {
      "entropy": 6.297755861282349,
      "epoch": 1.3335256780150029,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004833504925092492,
      "loss": 6.1452,
      "mean_token_accuracy": 0.16037169843912125,
      "num_tokens": 29284078.0,
      "step": 11555
    },
    {
      "entropy": 6.286835145950318,
      "epoch": 1.3341027120600115,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004833349109640944,
      "loss": 6.1047,
      "mean_token_accuracy": 0.16069846004247665,
      "num_tokens": 29295698.0,
      "step": 11560
    },
    {
      "entropy": 6.289292192459106,
      "epoch": 1.3346797461050202,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004833193224116983,
      "loss": 6.1341,
      "mean_token_accuracy": 0.16754496693611146,
      "num_tokens": 29307585.0,
      "step": 11565
    },
    {
      "entropy": 6.301826333999633,
      "epoch": 1.3352567801500288,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048330372685258526,
      "loss": 6.1567,
      "mean_token_accuracy": 0.16277503669261933,
      "num_tokens": 29321027.0,
      "step": 11570
    },
    {
      "entropy": 6.293691682815552,
      "epoch": 1.3358338141950374,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004832881242872799,
      "loss": 6.1722,
      "mean_token_accuracy": 0.1613366723060608,
      "num_tokens": 29333121.0,
      "step": 11575
    },
    {
      "entropy": 6.3491006851196286,
      "epoch": 1.336410848240046,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004832725147163069,
      "loss": 6.1535,
      "mean_token_accuracy": 0.15993758887052537,
      "num_tokens": 29346120.0,
      "step": 11580
    },
    {
      "entropy": 6.324691343307495,
      "epoch": 1.336987882285055,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048325689814019134,
      "loss": 6.1097,
      "mean_token_accuracy": 0.16394416689872743,
      "num_tokens": 29357872.0,
      "step": 11585
    },
    {
      "entropy": 6.338229036331176,
      "epoch": 1.3375649163300634,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004832412745594585,
      "loss": 6.1739,
      "mean_token_accuracy": 0.1639093592762947,
      "num_tokens": 29370657.0,
      "step": 11590
    },
    {
      "entropy": 6.240398502349853,
      "epoch": 1.3381419503750722,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048322564397463376,
      "loss": 6.0945,
      "mean_token_accuracy": 0.17317451536655426,
      "num_tokens": 29384880.0,
      "step": 11595
    },
    {
      "entropy": 6.3124298572540285,
      "epoch": 1.3387189844200809,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048321000638624296,
      "loss": 6.1144,
      "mean_token_accuracy": 0.16322875767946243,
      "num_tokens": 29397215.0,
      "step": 11600
    },
    {
      "entropy": 6.282052659988404,
      "epoch": 1.3392960184650895,
      "grad_norm": 2.125,
      "learning_rate": 0.000483194361794812,
      "loss": 6.1033,
      "mean_token_accuracy": 0.16651321500539779,
      "num_tokens": 29409634.0,
      "step": 11605
    },
    {
      "entropy": 6.286690282821655,
      "epoch": 1.3398730525100981,
      "grad_norm": 1.546875,
      "learning_rate": 0.000483178710200867,
      "loss": 6.0536,
      "mean_token_accuracy": 0.17319692820310592,
      "num_tokens": 29422404.0,
      "step": 11610
    },
    {
      "entropy": 6.323809862136841,
      "epoch": 1.3404500865551068,
      "grad_norm": 1.25,
      "learning_rate": 0.0004831630516049346,
      "loss": 6.2113,
      "mean_token_accuracy": 0.16090573668479918,
      "num_tokens": 29434727.0,
      "step": 11615
    },
    {
      "entropy": 6.316535472869873,
      "epoch": 1.3410271206001154,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004831473860075414,
      "loss": 6.136,
      "mean_token_accuracy": 0.15914638340473175,
      "num_tokens": 29447176.0,
      "step": 11620
    },
    {
      "entropy": 6.277497625350952,
      "epoch": 1.341604154645124,
      "grad_norm": 2.5625,
      "learning_rate": 0.00048313171340921417,
      "loss": 6.0976,
      "mean_token_accuracy": 0.17469708025455474,
      "num_tokens": 29460080.0,
      "step": 11625
    },
    {
      "entropy": 6.240986299514771,
      "epoch": 1.3421811886901327,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048311603381048025,
      "loss": 6.1296,
      "mean_token_accuracy": 0.1626189112663269,
      "num_tokens": 29472503.0,
      "step": 11630
    },
    {
      "entropy": 6.272552442550659,
      "epoch": 1.3427582227351413,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004831003472118668,
      "loss": 6.0962,
      "mean_token_accuracy": 0.16392599642276764,
      "num_tokens": 29485139.0,
      "step": 11635
    },
    {
      "entropy": 6.398651361465454,
      "epoch": 1.34333525678015,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004830846536139016,
      "loss": 6.1837,
      "mean_token_accuracy": 0.16744163930416106,
      "num_tokens": 29497455.0,
      "step": 11640
    },
    {
      "entropy": 6.3032959461212155,
      "epoch": 1.3439122908251586,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004830689530171124,
      "loss": 6.1284,
      "mean_token_accuracy": 0.1687514752149582,
      "num_tokens": 29509859.0,
      "step": 11645
    },
    {
      "entropy": 6.352106809616089,
      "epoch": 1.3444893248701673,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004830532454220273,
      "loss": 6.1896,
      "mean_token_accuracy": 0.1648508906364441,
      "num_tokens": 29523117.0,
      "step": 11650
    },
    {
      "entropy": 6.255600261688232,
      "epoch": 1.345066358915176,
      "grad_norm": 1.765625,
      "learning_rate": 0.00048303753082917474,
      "loss": 6.0831,
      "mean_token_accuracy": 0.16598070561885833,
      "num_tokens": 29535943.0,
      "step": 11655
    },
    {
      "entropy": 6.291255092620849,
      "epoch": 1.3456433929601848,
      "grad_norm": 1.5,
      "learning_rate": 0.00048302180923908306,
      "loss": 6.1696,
      "mean_token_accuracy": 0.16465048789978026,
      "num_tokens": 29549036.0,
      "step": 11660
    },
    {
      "entropy": 6.290594291687012,
      "epoch": 1.3462204270051932,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048300608065228126,
      "loss": 6.1377,
      "mean_token_accuracy": 0.1643916040658951,
      "num_tokens": 29563465.0,
      "step": 11665
    },
    {
      "entropy": 6.274702882766723,
      "epoch": 1.346797461050202,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048299034506929815,
      "loss": 6.0228,
      "mean_token_accuracy": 0.17369044125080108,
      "num_tokens": 29575891.0,
      "step": 11670
    },
    {
      "entropy": 6.33291220664978,
      "epoch": 1.3473744950952107,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00048297460249066315,
      "loss": 6.2557,
      "mean_token_accuracy": 0.15672912448644638,
      "num_tokens": 29589162.0,
      "step": 11675
    },
    {
      "entropy": 6.331118774414063,
      "epoch": 1.3479515291402193,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004829588529169057,
      "loss": 6.0916,
      "mean_token_accuracy": 0.1665223941206932,
      "num_tokens": 29602735.0,
      "step": 11680
    },
    {
      "entropy": 6.220363473892212,
      "epoch": 1.348528563185228,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004829430963485555,
      "loss": 6.0742,
      "mean_token_accuracy": 0.1701531946659088,
      "num_tokens": 29615683.0,
      "step": 11685
    },
    {
      "entropy": 6.316802501678467,
      "epoch": 1.3491055972302366,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004829273327861426,
      "loss": 6.1707,
      "mean_token_accuracy": 0.16423047631978988,
      "num_tokens": 29629606.0,
      "step": 11690
    },
    {
      "entropy": 6.291276979446411,
      "epoch": 1.3496826312752452,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004829115622301971,
      "loss": 6.0758,
      "mean_token_accuracy": 0.17152973264455795,
      "num_tokens": 29642250.0,
      "step": 11695
    },
    {
      "entropy": 6.3764811038970945,
      "epoch": 1.3502596653202539,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048289578468124956,
      "loss": 6.1999,
      "mean_token_accuracy": 0.16525790095329285,
      "num_tokens": 29655689.0,
      "step": 11700
    },
    {
      "entropy": 6.153978443145752,
      "epoch": 1.3508366993652625,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048288000013983046,
      "loss": 6.0987,
      "mean_token_accuracy": 0.16323120892047882,
      "num_tokens": 29670084.0,
      "step": 11705
    },
    {
      "entropy": 6.329030132293701,
      "epoch": 1.3514137334102712,
      "grad_norm": 1.90625,
      "learning_rate": 0.00048286420860647086,
      "loss": 6.1822,
      "mean_token_accuracy": 0.1604864478111267,
      "num_tokens": 29683453.0,
      "step": 11710
    },
    {
      "entropy": 6.313185548782348,
      "epoch": 1.3519907674552798,
      "grad_norm": 1.953125,
      "learning_rate": 0.00048284841008170185,
      "loss": 6.0545,
      "mean_token_accuracy": 0.16783878356218337,
      "num_tokens": 29695514.0,
      "step": 11715
    },
    {
      "entropy": 6.30164647102356,
      "epoch": 1.3525678015002884,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00048283260456605487,
      "loss": 6.1151,
      "mean_token_accuracy": 0.1725817874073982,
      "num_tokens": 29708544.0,
      "step": 11720
    },
    {
      "entropy": 6.364648389816284,
      "epoch": 1.3531448355452973,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004828167920600614,
      "loss": 6.0963,
      "mean_token_accuracy": 0.16400440484285356,
      "num_tokens": 29721310.0,
      "step": 11725
    },
    {
      "entropy": 6.348964786529541,
      "epoch": 1.3537218695903057,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004828009725642534,
      "loss": 6.1322,
      "mean_token_accuracy": 0.16307055801153184,
      "num_tokens": 29733995.0,
      "step": 11730
    },
    {
      "entropy": 6.314385557174683,
      "epoch": 1.3542989036353146,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004827851460791628,
      "loss": 6.1939,
      "mean_token_accuracy": 0.15663280338048935,
      "num_tokens": 29745896.0,
      "step": 11735
    },
    {
      "entropy": 6.3260167121887205,
      "epoch": 1.3548759376803232,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048276931260532213,
      "loss": 6.0957,
      "mean_token_accuracy": 0.16611895859241485,
      "num_tokens": 29758967.0,
      "step": 11740
    },
    {
      "entropy": 6.299111270904541,
      "epoch": 1.3554529717253319,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004827534721432639,
      "loss": 6.1453,
      "mean_token_accuracy": 0.15984882563352584,
      "num_tokens": 29772167.0,
      "step": 11745
    },
    {
      "entropy": 6.265276765823364,
      "epoch": 1.3560300057703405,
      "grad_norm": 1.25,
      "learning_rate": 0.0004827376246935207,
      "loss": 6.0998,
      "mean_token_accuracy": 0.16514657139778138,
      "num_tokens": 29783841.0,
      "step": 11750
    },
    {
      "entropy": 6.3505817413330075,
      "epoch": 1.3566070398153491,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004827217702566257,
      "loss": 6.1709,
      "mean_token_accuracy": 0.16244319677352906,
      "num_tokens": 29795993.0,
      "step": 11755
    },
    {
      "entropy": 6.288106441497803,
      "epoch": 1.3571840738603578,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048270590883311217,
      "loss": 6.1015,
      "mean_token_accuracy": 0.17155284732580184,
      "num_tokens": 29807692.0,
      "step": 11760
    },
    {
      "entropy": 6.287062501907348,
      "epoch": 1.3577611079053664,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048269004042351363,
      "loss": 6.1851,
      "mean_token_accuracy": 0.16180962175130845,
      "num_tokens": 29820778.0,
      "step": 11765
    },
    {
      "entropy": 6.317913389205932,
      "epoch": 1.358338141950375,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004826741650283637,
      "loss": 6.1186,
      "mean_token_accuracy": 0.1640876680612564,
      "num_tokens": 29833976.0,
      "step": 11770
    },
    {
      "entropy": 6.251207447052002,
      "epoch": 1.3589151759953837,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004826582826481963,
      "loss": 6.0767,
      "mean_token_accuracy": 0.17455727756023406,
      "num_tokens": 29846226.0,
      "step": 11775
    },
    {
      "entropy": 6.219823408126831,
      "epoch": 1.3594922100403923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004826423932835458,
      "loss": 6.0798,
      "mean_token_accuracy": 0.16895576119422911,
      "num_tokens": 29858483.0,
      "step": 11780
    },
    {
      "entropy": 6.314107465744018,
      "epoch": 1.360069244085401,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048262649693494653,
      "loss": 6.1396,
      "mean_token_accuracy": 0.16297580003738404,
      "num_tokens": 29871178.0,
      "step": 11785
    },
    {
      "entropy": 6.4023756980896,
      "epoch": 1.3606462781304096,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004826105936029332,
      "loss": 6.2018,
      "mean_token_accuracy": 0.15794295221567153,
      "num_tokens": 29885477.0,
      "step": 11790
    },
    {
      "entropy": 6.298975467681885,
      "epoch": 1.3612233121754183,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004825946832880406,
      "loss": 6.1311,
      "mean_token_accuracy": 0.16285300999879837,
      "num_tokens": 29897917.0,
      "step": 11795
    },
    {
      "entropy": 6.278939914703369,
      "epoch": 1.3618003462204271,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048257876599080404,
      "loss": 6.1943,
      "mean_token_accuracy": 0.16165986955165862,
      "num_tokens": 29911321.0,
      "step": 11800
    },
    {
      "entropy": 6.305810499191284,
      "epoch": 1.3623773802654355,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048256284171175874,
      "loss": 6.0903,
      "mean_token_accuracy": 0.17238346338272095,
      "num_tokens": 29924651.0,
      "step": 11805
    },
    {
      "entropy": 6.1934586524963375,
      "epoch": 1.3629544143104444,
      "grad_norm": 2.40625,
      "learning_rate": 0.00048254691045144035,
      "loss": 6.0301,
      "mean_token_accuracy": 0.16324946731328965,
      "num_tokens": 29937334.0,
      "step": 11810
    },
    {
      "entropy": 6.265475225448609,
      "epoch": 1.363531448355453,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004825309722103848,
      "loss": 6.073,
      "mean_token_accuracy": 0.17350615113973616,
      "num_tokens": 29949751.0,
      "step": 11815
    },
    {
      "entropy": 6.2313230514526365,
      "epoch": 1.3641084824004617,
      "grad_norm": 1.390625,
      "learning_rate": 0.000482515026989128,
      "loss": 6.081,
      "mean_token_accuracy": 0.1747325286269188,
      "num_tokens": 29961675.0,
      "step": 11820
    },
    {
      "entropy": 6.3038712501525875,
      "epoch": 1.3646855164454703,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048249907478820634,
      "loss": 6.2137,
      "mean_token_accuracy": 0.1612464517354965,
      "num_tokens": 29973222.0,
      "step": 11825
    },
    {
      "entropy": 6.353469610214233,
      "epoch": 1.365262550490479,
      "grad_norm": 1.828125,
      "learning_rate": 0.00048248311560815637,
      "loss": 6.1997,
      "mean_token_accuracy": 0.16309360787272453,
      "num_tokens": 29985845.0,
      "step": 11830
    },
    {
      "entropy": 6.320858573913574,
      "epoch": 1.3658395845354876,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004824671494495149,
      "loss": 6.1476,
      "mean_token_accuracy": 0.16458719223737717,
      "num_tokens": 29998252.0,
      "step": 11835
    },
    {
      "entropy": 6.346536493301391,
      "epoch": 1.3664166185804962,
      "grad_norm": 3.3125,
      "learning_rate": 0.0004824511763128189,
      "loss": 6.0951,
      "mean_token_accuracy": 0.17218410074710847,
      "num_tokens": 30011150.0,
      "step": 11840
    },
    {
      "entropy": 6.357008361816407,
      "epoch": 1.3669936526255049,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048243519619860567,
      "loss": 6.2157,
      "mean_token_accuracy": 0.1605319306254387,
      "num_tokens": 30024750.0,
      "step": 11845
    },
    {
      "entropy": 6.286159610748291,
      "epoch": 1.3675706866705135,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004824192091074126,
      "loss": 6.0616,
      "mean_token_accuracy": 0.1729810953140259,
      "num_tokens": 30038317.0,
      "step": 11850
    },
    {
      "entropy": 6.283872461318969,
      "epoch": 1.3681477207155222,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004824032150397775,
      "loss": 6.192,
      "mean_token_accuracy": 0.17048034220933914,
      "num_tokens": 30052156.0,
      "step": 11855
    },
    {
      "entropy": 6.305463886260986,
      "epoch": 1.3687247547605308,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048238721399623824,
      "loss": 6.1517,
      "mean_token_accuracy": 0.16549787670373917,
      "num_tokens": 30063463.0,
      "step": 11860
    },
    {
      "entropy": 6.304179716110229,
      "epoch": 1.3693017888055397,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048237120597733316,
      "loss": 6.0799,
      "mean_token_accuracy": 0.17041393965482712,
      "num_tokens": 30075090.0,
      "step": 11865
    },
    {
      "entropy": 6.236459493637085,
      "epoch": 1.369878822850548,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004823551909836005,
      "loss": 6.0931,
      "mean_token_accuracy": 0.16820509731769562,
      "num_tokens": 30087618.0,
      "step": 11870
    },
    {
      "entropy": 6.319930219650269,
      "epoch": 1.370455856895557,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048233916901557896,
      "loss": 6.08,
      "mean_token_accuracy": 0.16947691589593888,
      "num_tokens": 30098781.0,
      "step": 11875
    },
    {
      "entropy": 6.3354767799377445,
      "epoch": 1.3710328909405656,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048232314007380753,
      "loss": 6.1378,
      "mean_token_accuracy": 0.1753944605588913,
      "num_tokens": 30111163.0,
      "step": 11880
    },
    {
      "entropy": 6.294744157791138,
      "epoch": 1.3716099249855742,
      "grad_norm": 1.703125,
      "learning_rate": 0.00048230710415882524,
      "loss": 6.1374,
      "mean_token_accuracy": 0.16997584998607634,
      "num_tokens": 30124609.0,
      "step": 11885
    },
    {
      "entropy": 6.316244983673096,
      "epoch": 1.3721869590305829,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048229106127117144,
      "loss": 6.1872,
      "mean_token_accuracy": 0.16650330871343613,
      "num_tokens": 30137823.0,
      "step": 11890
    },
    {
      "entropy": 6.389352178573608,
      "epoch": 1.3727639930755915,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004822750114113858,
      "loss": 6.1348,
      "mean_token_accuracy": 0.16203177869319915,
      "num_tokens": 30150295.0,
      "step": 11895
    },
    {
      "entropy": 6.263393402099609,
      "epoch": 1.3733410271206001,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004822589545800081,
      "loss": 6.072,
      "mean_token_accuracy": 0.1705651491880417,
      "num_tokens": 30163533.0,
      "step": 11900
    },
    {
      "entropy": 6.308654689788819,
      "epoch": 1.3739180611656088,
      "grad_norm": 2.171875,
      "learning_rate": 0.0004822428907775784,
      "loss": 6.1621,
      "mean_token_accuracy": 0.15814722031354905,
      "num_tokens": 30176437.0,
      "step": 11905
    },
    {
      "entropy": 6.218947982788086,
      "epoch": 1.3744950952106174,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048222682000463704,
      "loss": 6.0643,
      "mean_token_accuracy": 0.17379665672779082,
      "num_tokens": 30190181.0,
      "step": 11910
    },
    {
      "entropy": 6.212062406539917,
      "epoch": 1.375072129255626,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004822107422617245,
      "loss": 6.1404,
      "mean_token_accuracy": 0.16834926456212998,
      "num_tokens": 30202374.0,
      "step": 11915
    },
    {
      "entropy": 6.24512882232666,
      "epoch": 1.3756491633006347,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048219465754938156,
      "loss": 6.0318,
      "mean_token_accuracy": 0.16531560868024825,
      "num_tokens": 30215009.0,
      "step": 11920
    },
    {
      "entropy": 6.35242805480957,
      "epoch": 1.3762261973456433,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048217856586814926,
      "loss": 6.1363,
      "mean_token_accuracy": 0.16434847116470336,
      "num_tokens": 30226688.0,
      "step": 11925
    },
    {
      "entropy": 6.3303159236907955,
      "epoch": 1.376803231390652,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048216246721856875,
      "loss": 6.1683,
      "mean_token_accuracy": 0.16100845187902452,
      "num_tokens": 30238501.0,
      "step": 11930
    },
    {
      "entropy": 6.335440397262573,
      "epoch": 1.3773802654356606,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048214636160118164,
      "loss": 6.107,
      "mean_token_accuracy": 0.16803493201732636,
      "num_tokens": 30250216.0,
      "step": 11935
    },
    {
      "entropy": 6.3147039890289305,
      "epoch": 1.3779572994806695,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004821302490165295,
      "loss": 6.1213,
      "mean_token_accuracy": 0.16672479808330537,
      "num_tokens": 30262555.0,
      "step": 11940
    },
    {
      "entropy": 6.233082962036133,
      "epoch": 1.378534333525678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004821141294651544,
      "loss": 6.0262,
      "mean_token_accuracy": 0.1772979751229286,
      "num_tokens": 30274593.0,
      "step": 11945
    },
    {
      "entropy": 6.325069808959961,
      "epoch": 1.3791113675706868,
      "grad_norm": 1.71875,
      "learning_rate": 0.00048209800294759836,
      "loss": 6.0518,
      "mean_token_accuracy": 0.17313085049390792,
      "num_tokens": 30286579.0,
      "step": 11950
    },
    {
      "entropy": 6.351945686340332,
      "epoch": 1.3796884016156954,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004820818694644039,
      "loss": 6.1369,
      "mean_token_accuracy": 0.16453606635332108,
      "num_tokens": 30299186.0,
      "step": 11955
    },
    {
      "entropy": 6.3624310970306395,
      "epoch": 1.380265435660704,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048206572901611364,
      "loss": 6.1881,
      "mean_token_accuracy": 0.15905419141054153,
      "num_tokens": 30311429.0,
      "step": 11960
    },
    {
      "entropy": 6.3475316047668455,
      "epoch": 1.3808424697057127,
      "grad_norm": 1.875,
      "learning_rate": 0.00048204958160327034,
      "loss": 6.1504,
      "mean_token_accuracy": 0.1662835657596588,
      "num_tokens": 30323323.0,
      "step": 11965
    },
    {
      "entropy": 6.276279354095459,
      "epoch": 1.3814195037507213,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048203342722641726,
      "loss": 6.1237,
      "mean_token_accuracy": 0.16957689821720123,
      "num_tokens": 30335815.0,
      "step": 11970
    },
    {
      "entropy": 6.24206919670105,
      "epoch": 1.38199653779573,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048201726588609776,
      "loss": 6.0503,
      "mean_token_accuracy": 0.1701612576842308,
      "num_tokens": 30347426.0,
      "step": 11975
    },
    {
      "entropy": 6.314512062072754,
      "epoch": 1.3825735718407386,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048200109758285534,
      "loss": 6.1244,
      "mean_token_accuracy": 0.16389783024787902,
      "num_tokens": 30359781.0,
      "step": 11980
    },
    {
      "entropy": 6.318188142776489,
      "epoch": 1.3831506058857472,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004819849223172337,
      "loss": 6.133,
      "mean_token_accuracy": 0.16730300784111024,
      "num_tokens": 30373399.0,
      "step": 11985
    },
    {
      "entropy": 6.202302885055542,
      "epoch": 1.3837276399307559,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048196874008977716,
      "loss": 6.0891,
      "mean_token_accuracy": 0.17317621558904647,
      "num_tokens": 30386296.0,
      "step": 11990
    },
    {
      "entropy": 6.289775609970093,
      "epoch": 1.3843046739757645,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004819525509010298,
      "loss": 6.131,
      "mean_token_accuracy": 0.17061251550912857,
      "num_tokens": 30399496.0,
      "step": 11995
    },
    {
      "entropy": 6.286764621734619,
      "epoch": 1.3848817080207732,
      "grad_norm": 1.984375,
      "learning_rate": 0.0004819363547515361,
      "loss": 6.0803,
      "mean_token_accuracy": 0.16182739287614822,
      "num_tokens": 30412491.0,
      "step": 12000
    },
    {
      "epoch": 1.3848817080207732,
      "eval_entropy": 6.224000952614618,
      "eval_loss": 6.247929573059082,
      "eval_mean_token_accuracy": 0.16458880992324,
      "eval_num_tokens": 30412491.0,
      "eval_runtime": 22.6792,
      "eval_samples_per_second": 1240.607,
      "eval_steps_per_second": 155.076,
      "step": 12000
    },
    {
      "entropy": 6.369675827026367,
      "epoch": 1.385458742065782,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000481920151641841,
      "loss": 6.2147,
      "mean_token_accuracy": 0.16291867345571517,
      "num_tokens": 30427600.0,
      "step": 12005
    },
    {
      "entropy": 6.355693054199219,
      "epoch": 1.3860357761107904,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048190394157248935,
      "loss": 6.0842,
      "mean_token_accuracy": 0.171800796687603,
      "num_tokens": 30440534.0,
      "step": 12010
    },
    {
      "entropy": 6.275745153427124,
      "epoch": 1.3866128101557993,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004818877245440264,
      "loss": 6.0815,
      "mean_token_accuracy": 0.16982559263706207,
      "num_tokens": 30453996.0,
      "step": 12015
    },
    {
      "entropy": 6.300146865844726,
      "epoch": 1.3871898442008077,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048187150055699763,
      "loss": 6.1495,
      "mean_token_accuracy": 0.16266092360019685,
      "num_tokens": 30465667.0,
      "step": 12020
    },
    {
      "entropy": 6.351484680175782,
      "epoch": 1.3877668782458166,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004818552696119487,
      "loss": 6.1958,
      "mean_token_accuracy": 0.1609213873744011,
      "num_tokens": 30478950.0,
      "step": 12025
    },
    {
      "entropy": 6.2422269821167,
      "epoch": 1.3883439122908252,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004818390317094255,
      "loss": 6.0572,
      "mean_token_accuracy": 0.17113839089870453,
      "num_tokens": 30490979.0,
      "step": 12030
    },
    {
      "entropy": 6.311171054840088,
      "epoch": 1.3889209463358339,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004818227868499742,
      "loss": 6.0909,
      "mean_token_accuracy": 0.17111729681491852,
      "num_tokens": 30504242.0,
      "step": 12035
    },
    {
      "entropy": 6.25700569152832,
      "epoch": 1.3894979803808425,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004818065350341412,
      "loss": 6.1619,
      "mean_token_accuracy": 0.16251288056373597,
      "num_tokens": 30515739.0,
      "step": 12040
    },
    {
      "entropy": 6.387627744674683,
      "epoch": 1.3900750144258511,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048179027626247325,
      "loss": 6.1466,
      "mean_token_accuracy": 0.15438068956136702,
      "num_tokens": 30528338.0,
      "step": 12045
    },
    {
      "entropy": 6.408785104751587,
      "epoch": 1.3906520484708598,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004817740105355169,
      "loss": 6.1735,
      "mean_token_accuracy": 0.16251056343317033,
      "num_tokens": 30540572.0,
      "step": 12050
    },
    {
      "entropy": 6.360489416122436,
      "epoch": 1.3912290825158684,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048175773785381947,
      "loss": 6.196,
      "mean_token_accuracy": 0.1643637403845787,
      "num_tokens": 30554317.0,
      "step": 12055
    },
    {
      "entropy": 6.341495227813721,
      "epoch": 1.391806116560877,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048174145821792833,
      "loss": 6.1534,
      "mean_token_accuracy": 0.16100704222917556,
      "num_tokens": 30567177.0,
      "step": 12060
    },
    {
      "entropy": 6.35797700881958,
      "epoch": 1.3923831506058857,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004817251716283908,
      "loss": 6.1489,
      "mean_token_accuracy": 0.1654942288994789,
      "num_tokens": 30577948.0,
      "step": 12065
    },
    {
      "entropy": 6.228287220001221,
      "epoch": 1.3929601846508943,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004817088780857548,
      "loss": 6.1331,
      "mean_token_accuracy": 0.16523568630218505,
      "num_tokens": 30590548.0,
      "step": 12070
    },
    {
      "entropy": 6.303491497039795,
      "epoch": 1.393537218695903,
      "grad_norm": 1.5,
      "learning_rate": 0.00048169257759056845,
      "loss": 6.049,
      "mean_token_accuracy": 0.17548049986362457,
      "num_tokens": 30602735.0,
      "step": 12075
    },
    {
      "entropy": 6.357599115371704,
      "epoch": 1.3941142527409118,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048167627014337994,
      "loss": 6.1888,
      "mean_token_accuracy": 0.16183228939771652,
      "num_tokens": 30615112.0,
      "step": 12080
    },
    {
      "entropy": 6.253079414367676,
      "epoch": 1.3946912867859202,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048165995574473764,
      "loss": 6.1103,
      "mean_token_accuracy": 0.1625622555613518,
      "num_tokens": 30627406.0,
      "step": 12085
    },
    {
      "entropy": 6.312600135803223,
      "epoch": 1.395268320830929,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00048164363439519043,
      "loss": 6.1182,
      "mean_token_accuracy": 0.1654726192355156,
      "num_tokens": 30639764.0,
      "step": 12090
    },
    {
      "entropy": 6.347026824951172,
      "epoch": 1.3958453548759377,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004816273060952873,
      "loss": 6.1022,
      "mean_token_accuracy": 0.16689932197332383,
      "num_tokens": 30652181.0,
      "step": 12095
    },
    {
      "entropy": 6.306456422805786,
      "epoch": 1.3964223889209464,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048161097084557726,
      "loss": 6.1286,
      "mean_token_accuracy": 0.1631346881389618,
      "num_tokens": 30663586.0,
      "step": 12100
    },
    {
      "entropy": 6.374965047836303,
      "epoch": 1.396999422965955,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048159462864660993,
      "loss": 6.1952,
      "mean_token_accuracy": 0.15762596130371093,
      "num_tokens": 30676218.0,
      "step": 12105
    },
    {
      "entropy": 6.339052629470825,
      "epoch": 1.3975764570109637,
      "grad_norm": 1.90625,
      "learning_rate": 0.0004815782794989348,
      "loss": 6.1864,
      "mean_token_accuracy": 0.1606118932366371,
      "num_tokens": 30688532.0,
      "step": 12110
    },
    {
      "entropy": 6.2383440971374515,
      "epoch": 1.3981534910559723,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004815619234031019,
      "loss": 6.1052,
      "mean_token_accuracy": 0.16475950181484222,
      "num_tokens": 30700698.0,
      "step": 12115
    },
    {
      "entropy": 6.2884252071380615,
      "epoch": 1.398730525100981,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004815455603596613,
      "loss": 6.1533,
      "mean_token_accuracy": 0.16549549996852875,
      "num_tokens": 30713032.0,
      "step": 12120
    },
    {
      "entropy": 6.294129419326782,
      "epoch": 1.3993075591459896,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004815291903691634,
      "loss": 6.0719,
      "mean_token_accuracy": 0.1696466714143753,
      "num_tokens": 30726829.0,
      "step": 12125
    },
    {
      "entropy": 6.337021923065185,
      "epoch": 1.3998845931909982,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048151281343215873,
      "loss": 6.1566,
      "mean_token_accuracy": 0.1605507642030716,
      "num_tokens": 30740280.0,
      "step": 12130
    },
    {
      "entropy": 6.21012511253357,
      "epoch": 1.4004616272360069,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004814964295491982,
      "loss": 6.0093,
      "mean_token_accuracy": 0.17915696054697036,
      "num_tokens": 30753467.0,
      "step": 12135
    },
    {
      "entropy": 6.279628944396973,
      "epoch": 1.4010386612810155,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048148003872083286,
      "loss": 6.0944,
      "mean_token_accuracy": 0.17217430621385574,
      "num_tokens": 30766958.0,
      "step": 12140
    },
    {
      "entropy": 6.305673313140869,
      "epoch": 1.4016156953260244,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048146364094761384,
      "loss": 6.1684,
      "mean_token_accuracy": 0.1595797434449196,
      "num_tokens": 30779114.0,
      "step": 12145
    },
    {
      "entropy": 6.294431924819946,
      "epoch": 1.4021927293710328,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048144723623009297,
      "loss": 6.0811,
      "mean_token_accuracy": 0.16946638971567154,
      "num_tokens": 30792063.0,
      "step": 12150
    },
    {
      "entropy": 6.3064147472381595,
      "epoch": 1.4027697634160416,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004814308245688218,
      "loss": 6.1343,
      "mean_token_accuracy": 0.16511736810207367,
      "num_tokens": 30804404.0,
      "step": 12155
    },
    {
      "entropy": 6.235158681869507,
      "epoch": 1.40334679746105,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048141440596435235,
      "loss": 6.0719,
      "mean_token_accuracy": 0.17058005332946777,
      "num_tokens": 30817607.0,
      "step": 12160
    },
    {
      "entropy": 6.2955762386322025,
      "epoch": 1.403923831506059,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004813979804172369,
      "loss": 6.1501,
      "mean_token_accuracy": 0.1665330022573471,
      "num_tokens": 30831077.0,
      "step": 12165
    },
    {
      "entropy": 6.28407392501831,
      "epoch": 1.4045008655510676,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048138154792802795,
      "loss": 6.2252,
      "mean_token_accuracy": 0.1606106773018837,
      "num_tokens": 30844096.0,
      "step": 12170
    },
    {
      "entropy": 6.323961544036865,
      "epoch": 1.4050778995960762,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004813651084972781,
      "loss": 6.1439,
      "mean_token_accuracy": 0.1660185769200325,
      "num_tokens": 30856147.0,
      "step": 12175
    },
    {
      "entropy": 6.302383995056152,
      "epoch": 1.4056549336410848,
      "grad_norm": 1.125,
      "learning_rate": 0.00048134866212554036,
      "loss": 6.1361,
      "mean_token_accuracy": 0.15867800861597062,
      "num_tokens": 30869436.0,
      "step": 12180
    },
    {
      "entropy": 6.245109987258911,
      "epoch": 1.4062319676860935,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004813322088133679,
      "loss": 6.1151,
      "mean_token_accuracy": 0.16820482015609742,
      "num_tokens": 30882313.0,
      "step": 12185
    },
    {
      "entropy": 6.295883369445801,
      "epoch": 1.4068090017311021,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048131574856131407,
      "loss": 6.1159,
      "mean_token_accuracy": 0.16215179339051247,
      "num_tokens": 30895062.0,
      "step": 12190
    },
    {
      "entropy": 6.299847841262817,
      "epoch": 1.4073860357761108,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004812992813699324,
      "loss": 6.1053,
      "mean_token_accuracy": 0.1652792364358902,
      "num_tokens": 30907736.0,
      "step": 12195
    },
    {
      "entropy": 6.322113800048828,
      "epoch": 1.4079630698211194,
      "grad_norm": 1.484375,
      "learning_rate": 0.000481282807239777,
      "loss": 6.1166,
      "mean_token_accuracy": 0.16788258105516435,
      "num_tokens": 30920462.0,
      "step": 12200
    },
    {
      "entropy": 6.28887448310852,
      "epoch": 1.408540103866128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004812663261714019,
      "loss": 6.0392,
      "mean_token_accuracy": 0.17208631336688995,
      "num_tokens": 30933501.0,
      "step": 12205
    },
    {
      "entropy": 6.288034152984619,
      "epoch": 1.4091171379111367,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004812498381653613,
      "loss": 6.0682,
      "mean_token_accuracy": 0.17028167694807053,
      "num_tokens": 30944780.0,
      "step": 12210
    },
    {
      "entropy": 6.348947048187256,
      "epoch": 1.4096941719561453,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004812333432222098,
      "loss": 6.1574,
      "mean_token_accuracy": 0.16152927428483962,
      "num_tokens": 30957730.0,
      "step": 12215
    },
    {
      "entropy": 6.068280220031738,
      "epoch": 1.4102712060011542,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004812168413425023,
      "loss": 5.9705,
      "mean_token_accuracy": 0.181918466091156,
      "num_tokens": 30969371.0,
      "step": 12220
    },
    {
      "entropy": 6.326417303085327,
      "epoch": 1.4108482400461626,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048120033252679374,
      "loss": 6.1889,
      "mean_token_accuracy": 0.16231568902730942,
      "num_tokens": 30983662.0,
      "step": 12225
    },
    {
      "entropy": 6.430174541473389,
      "epoch": 1.4114252740911715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048118381677563946,
      "loss": 6.1601,
      "mean_token_accuracy": 0.1615968704223633,
      "num_tokens": 30995641.0,
      "step": 12230
    },
    {
      "entropy": 6.362520742416382,
      "epoch": 1.41200230813618,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004811672940895949,
      "loss": 6.2089,
      "mean_token_accuracy": 0.16106947362422944,
      "num_tokens": 31008508.0,
      "step": 12235
    },
    {
      "entropy": 6.3622204780578615,
      "epoch": 1.4125793421811887,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004811507644692158,
      "loss": 6.1686,
      "mean_token_accuracy": 0.16129661202430726,
      "num_tokens": 31020903.0,
      "step": 12240
    },
    {
      "entropy": 6.325216770172119,
      "epoch": 1.4131563762261974,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004811342279150581,
      "loss": 6.1443,
      "mean_token_accuracy": 0.16731538027524948,
      "num_tokens": 31033668.0,
      "step": 12245
    },
    {
      "entropy": 6.345950746536255,
      "epoch": 1.413733410271206,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004811176844276781,
      "loss": 6.1887,
      "mean_token_accuracy": 0.1661725699901581,
      "num_tokens": 31046980.0,
      "step": 12250
    },
    {
      "entropy": 6.335570049285889,
      "epoch": 1.4143104443162147,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004811011340076322,
      "loss": 6.053,
      "mean_token_accuracy": 0.16766103059053422,
      "num_tokens": 31059023.0,
      "step": 12255
    },
    {
      "entropy": 6.250242805480957,
      "epoch": 1.4148874783612233,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048108457665547695,
      "loss": 6.0954,
      "mean_token_accuracy": 0.17201029807329177,
      "num_tokens": 31070449.0,
      "step": 12260
    },
    {
      "entropy": 6.257473611831665,
      "epoch": 1.415464512406232,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004810680123717694,
      "loss": 6.06,
      "mean_token_accuracy": 0.17232086062431334,
      "num_tokens": 31081974.0,
      "step": 12265
    },
    {
      "entropy": 6.306641006469727,
      "epoch": 1.4160415464512406,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004810514411570666,
      "loss": 6.1572,
      "mean_token_accuracy": 0.1588302806019783,
      "num_tokens": 31094825.0,
      "step": 12270
    },
    {
      "entropy": 6.242742490768433,
      "epoch": 1.4166185804962492,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004810348630119259,
      "loss": 6.0835,
      "mean_token_accuracy": 0.16637780368328095,
      "num_tokens": 31106589.0,
      "step": 12275
    },
    {
      "entropy": 6.383942031860352,
      "epoch": 1.4171956145412579,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048101827793690493,
      "loss": 6.1439,
      "mean_token_accuracy": 0.157748444378376,
      "num_tokens": 31118228.0,
      "step": 12280
    },
    {
      "entropy": 6.387330436706543,
      "epoch": 1.4177726485862667,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004810016859325615,
      "loss": 6.208,
      "mean_token_accuracy": 0.16304440647363663,
      "num_tokens": 31131396.0,
      "step": 12285
    },
    {
      "entropy": 6.327767467498779,
      "epoch": 1.4183496826312751,
      "grad_norm": 1.25,
      "learning_rate": 0.0004809850869994537,
      "loss": 6.1342,
      "mean_token_accuracy": 0.16658884882926941,
      "num_tokens": 31145524.0,
      "step": 12290
    },
    {
      "entropy": 6.349434852600098,
      "epoch": 1.418926716676284,
      "grad_norm": 1.25,
      "learning_rate": 0.0004809684811381398,
      "loss": 6.0971,
      "mean_token_accuracy": 0.17099672704935073,
      "num_tokens": 31157866.0,
      "step": 12295
    },
    {
      "entropy": 6.213384199142456,
      "epoch": 1.4195037507212924,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004809518683491785,
      "loss": 6.0452,
      "mean_token_accuracy": 0.17853738218545914,
      "num_tokens": 31170634.0,
      "step": 12300
    },
    {
      "entropy": 6.281233596801758,
      "epoch": 1.4200807847663013,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048093524863312823,
      "loss": 6.0843,
      "mean_token_accuracy": 0.16574939489364623,
      "num_tokens": 31183585.0,
      "step": 12305
    },
    {
      "entropy": 6.327755784988403,
      "epoch": 1.42065781881131,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004809186219905482,
      "loss": 6.1063,
      "mean_token_accuracy": 0.16455602496862412,
      "num_tokens": 31196375.0,
      "step": 12310
    },
    {
      "entropy": 6.266958808898925,
      "epoch": 1.4212348528563186,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004809019884219976,
      "loss": 6.0963,
      "mean_token_accuracy": 0.17540175914764405,
      "num_tokens": 31209797.0,
      "step": 12315
    },
    {
      "entropy": 6.20970196723938,
      "epoch": 1.4218118869013272,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048088534792803595,
      "loss": 5.9975,
      "mean_token_accuracy": 0.17160076051950454,
      "num_tokens": 31221196.0,
      "step": 12320
    },
    {
      "entropy": 6.305485630035401,
      "epoch": 1.4223889209463358,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00048086870050922286,
      "loss": 6.1375,
      "mean_token_accuracy": 0.1657287821173668,
      "num_tokens": 31234126.0,
      "step": 12325
    },
    {
      "entropy": 6.313985776901245,
      "epoch": 1.4229659549913445,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00048085204616611833,
      "loss": 6.1394,
      "mean_token_accuracy": 0.16098834723234176,
      "num_tokens": 31245933.0,
      "step": 12330
    },
    {
      "entropy": 6.293331670761108,
      "epoch": 1.4235429890363531,
      "grad_norm": 1.125,
      "learning_rate": 0.00048083538489928246,
      "loss": 6.1354,
      "mean_token_accuracy": 0.15946199297904967,
      "num_tokens": 31258734.0,
      "step": 12335
    },
    {
      "entropy": 6.274984502792359,
      "epoch": 1.4241200230813618,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004808187167092757,
      "loss": 6.0379,
      "mean_token_accuracy": 0.16822846084833146,
      "num_tokens": 31272487.0,
      "step": 12340
    },
    {
      "entropy": 6.23299126625061,
      "epoch": 1.4246970571263704,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004808020415966586,
      "loss": 6.1069,
      "mean_token_accuracy": 0.16606823056936265,
      "num_tokens": 31285117.0,
      "step": 12345
    },
    {
      "entropy": 6.254212093353272,
      "epoch": 1.425274091171379,
      "grad_norm": 1.453125,
      "learning_rate": 0.000480785359561992,
      "loss": 6.0913,
      "mean_token_accuracy": 0.1718271940946579,
      "num_tokens": 31297772.0,
      "step": 12350
    },
    {
      "entropy": 6.278341388702392,
      "epoch": 1.4258511252163877,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048076867060583704,
      "loss": 6.0946,
      "mean_token_accuracy": 0.1666000247001648,
      "num_tokens": 31310901.0,
      "step": 12355
    },
    {
      "entropy": 6.279241514205933,
      "epoch": 1.4264281592613965,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004807519747287551,
      "loss": 6.1194,
      "mean_token_accuracy": 0.16237711459398269,
      "num_tokens": 31323090.0,
      "step": 12360
    },
    {
      "entropy": 6.275631761550903,
      "epoch": 1.427005193306405,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004807352719313078,
      "loss": 6.0664,
      "mean_token_accuracy": 0.16846050322055817,
      "num_tokens": 31335883.0,
      "step": 12365
    },
    {
      "entropy": 6.350681781768799,
      "epoch": 1.4275822273514138,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004807185622140567,
      "loss": 6.1694,
      "mean_token_accuracy": 0.16167529821395873,
      "num_tokens": 31348237.0,
      "step": 12370
    },
    {
      "entropy": 6.219129180908203,
      "epoch": 1.4281592613964225,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048070184557756396,
      "loss": 6.0196,
      "mean_token_accuracy": 0.17124342024326325,
      "num_tokens": 31360414.0,
      "step": 12375
    },
    {
      "entropy": 6.2988440990448,
      "epoch": 1.428736295441431,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00048068512202239177,
      "loss": 6.1451,
      "mean_token_accuracy": 0.16669683009386063,
      "num_tokens": 31372765.0,
      "step": 12380
    },
    {
      "entropy": 6.306262874603272,
      "epoch": 1.4293133294864397,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004806683915491028,
      "loss": 6.1163,
      "mean_token_accuracy": 0.16952491849660872,
      "num_tokens": 31386176.0,
      "step": 12385
    },
    {
      "entropy": 6.350312519073486,
      "epoch": 1.4298903635314484,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004806516541582596,
      "loss": 6.1894,
      "mean_token_accuracy": 0.15940093100070954,
      "num_tokens": 31398385.0,
      "step": 12390
    },
    {
      "entropy": 6.2261138439178465,
      "epoch": 1.430467397576457,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048063490985042506,
      "loss": 6.0459,
      "mean_token_accuracy": 0.16982655823230744,
      "num_tokens": 31410841.0,
      "step": 12395
    },
    {
      "entropy": 6.241001892089844,
      "epoch": 1.4310444316214657,
      "grad_norm": 1.375,
      "learning_rate": 0.0004806181586261626,
      "loss": 6.1138,
      "mean_token_accuracy": 0.1683602899312973,
      "num_tokens": 31425586.0,
      "step": 12400
    },
    {
      "entropy": 6.321263217926026,
      "epoch": 1.4316214656664743,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048060140048603544,
      "loss": 6.0572,
      "mean_token_accuracy": 0.17428047508001326,
      "num_tokens": 31438943.0,
      "step": 12405
    },
    {
      "entropy": 6.241196346282959,
      "epoch": 1.432198499711483,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004805846354306073,
      "loss": 6.0361,
      "mean_token_accuracy": 0.1756817027926445,
      "num_tokens": 31452284.0,
      "step": 12410
    },
    {
      "entropy": 6.28016357421875,
      "epoch": 1.4327755337564916,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048056786346044214,
      "loss": 6.2197,
      "mean_token_accuracy": 0.1551727384328842,
      "num_tokens": 31466311.0,
      "step": 12415
    },
    {
      "entropy": 6.351090955734253,
      "epoch": 1.4333525678015002,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048055108457610395,
      "loss": 6.0679,
      "mean_token_accuracy": 0.1721023514866829,
      "num_tokens": 31478916.0,
      "step": 12420
    },
    {
      "entropy": 6.31014575958252,
      "epoch": 1.4339296018465089,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004805342987781571,
      "loss": 6.1017,
      "mean_token_accuracy": 0.16851240694522857,
      "num_tokens": 31490337.0,
      "step": 12425
    },
    {
      "entropy": 6.290010833740235,
      "epoch": 1.4345066358915175,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004805175060671663,
      "loss": 6.1066,
      "mean_token_accuracy": 0.16243394911289216,
      "num_tokens": 31502625.0,
      "step": 12430
    },
    {
      "entropy": 6.323895978927612,
      "epoch": 1.4350836699365264,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004805007064436962,
      "loss": 6.1662,
      "mean_token_accuracy": 0.1605980172753334,
      "num_tokens": 31515440.0,
      "step": 12435
    },
    {
      "entropy": 6.297709703445435,
      "epoch": 1.4356607039815348,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004804838999083119,
      "loss": 6.1129,
      "mean_token_accuracy": 0.16135017722845077,
      "num_tokens": 31527812.0,
      "step": 12440
    },
    {
      "entropy": 6.344575595855713,
      "epoch": 1.4362377380265436,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004804670864615787,
      "loss": 6.1149,
      "mean_token_accuracy": 0.16526193022727967,
      "num_tokens": 31540249.0,
      "step": 12445
    },
    {
      "entropy": 6.285081386566162,
      "epoch": 1.4368147720715523,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048045026610406223,
      "loss": 6.065,
      "mean_token_accuracy": 0.17280279397964476,
      "num_tokens": 31552312.0,
      "step": 12450
    },
    {
      "entropy": 6.2848773956298825,
      "epoch": 1.437391806116561,
      "grad_norm": 1.671875,
      "learning_rate": 0.000480433438836328,
      "loss": 6.1143,
      "mean_token_accuracy": 0.16405625715851785,
      "num_tokens": 31564496.0,
      "step": 12455
    },
    {
      "entropy": 6.360372257232666,
      "epoch": 1.4379688401615696,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048041660465894206,
      "loss": 6.1226,
      "mean_token_accuracy": 0.16145581156015396,
      "num_tokens": 31577705.0,
      "step": 12460
    },
    {
      "entropy": 6.3164557933807375,
      "epoch": 1.4385458742065782,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004803997635724707,
      "loss": 6.1298,
      "mean_token_accuracy": 0.16466890573501586,
      "num_tokens": 31589566.0,
      "step": 12465
    },
    {
      "entropy": 6.304163312911987,
      "epoch": 1.4391229082515868,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004803829155774804,
      "loss": 6.1433,
      "mean_token_accuracy": 0.16303829699754716,
      "num_tokens": 31602379.0,
      "step": 12470
    },
    {
      "entropy": 6.294868755340576,
      "epoch": 1.4396999422965955,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004803660606745377,
      "loss": 6.0728,
      "mean_token_accuracy": 0.17045496106147767,
      "num_tokens": 31615469.0,
      "step": 12475
    },
    {
      "entropy": 6.249364948272705,
      "epoch": 1.4402769763416041,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048034919886420953,
      "loss": 6.198,
      "mean_token_accuracy": 0.1614298328757286,
      "num_tokens": 31627129.0,
      "step": 12480
    },
    {
      "entropy": 6.392160129547119,
      "epoch": 1.4408540103866128,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048033233014706304,
      "loss": 6.1553,
      "mean_token_accuracy": 0.16891637742519378,
      "num_tokens": 31639564.0,
      "step": 12485
    },
    {
      "entropy": 6.259813594818115,
      "epoch": 1.4414310444316214,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048031545452366565,
      "loss": 6.0918,
      "mean_token_accuracy": 0.16519918888807297,
      "num_tokens": 31651984.0,
      "step": 12490
    },
    {
      "entropy": 6.27579026222229,
      "epoch": 1.44200807847663,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048029857199458493,
      "loss": 6.1202,
      "mean_token_accuracy": 0.17073165774345397,
      "num_tokens": 31664850.0,
      "step": 12495
    },
    {
      "entropy": 6.299057912826538,
      "epoch": 1.442585112521639,
      "grad_norm": 1.671875,
      "learning_rate": 0.00048028168256038873,
      "loss": 6.0618,
      "mean_token_accuracy": 0.17626596838235856,
      "num_tokens": 31676993.0,
      "step": 12500
    },
    {
      "entropy": 6.314305353164673,
      "epoch": 1.4431621465666473,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048026478622164513,
      "loss": 6.199,
      "mean_token_accuracy": 0.1673912525177002,
      "num_tokens": 31689632.0,
      "step": 12505
    },
    {
      "entropy": 6.3189842224121096,
      "epoch": 1.4437391806116562,
      "grad_norm": 1.25,
      "learning_rate": 0.00048024788297892234,
      "loss": 6.1337,
      "mean_token_accuracy": 0.16439658850431443,
      "num_tokens": 31701610.0,
      "step": 12510
    },
    {
      "entropy": 6.324826955795288,
      "epoch": 1.4443162146566648,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000480230972832789,
      "loss": 6.0846,
      "mean_token_accuracy": 0.16393847912549972,
      "num_tokens": 31713838.0,
      "step": 12515
    },
    {
      "entropy": 6.257188940048218,
      "epoch": 1.4448932487016735,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048021405578381384,
      "loss": 6.0549,
      "mean_token_accuracy": 0.16817671060562134,
      "num_tokens": 31726097.0,
      "step": 12520
    },
    {
      "entropy": 6.286736917495728,
      "epoch": 1.445470282746682,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004801971318325659,
      "loss": 6.0761,
      "mean_token_accuracy": 0.1660412147641182,
      "num_tokens": 31739646.0,
      "step": 12525
    },
    {
      "entropy": 6.279177141189575,
      "epoch": 1.4460473167916907,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004801802009796142,
      "loss": 6.0792,
      "mean_token_accuracy": 0.16675680130720139,
      "num_tokens": 31752161.0,
      "step": 12530
    },
    {
      "entropy": 6.275993585586548,
      "epoch": 1.4466243508366994,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004801632632255285,
      "loss": 6.0858,
      "mean_token_accuracy": 0.1660511910915375,
      "num_tokens": 31764498.0,
      "step": 12535
    },
    {
      "entropy": 6.336894416809082,
      "epoch": 1.447201384881708,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004801463185708783,
      "loss": 6.1314,
      "mean_token_accuracy": 0.16206191182136537,
      "num_tokens": 31776453.0,
      "step": 12540
    },
    {
      "entropy": 6.299965858459473,
      "epoch": 1.4477784189267167,
      "grad_norm": 2.0625,
      "learning_rate": 0.00048012936701623363,
      "loss": 6.1735,
      "mean_token_accuracy": 0.16327776312828063,
      "num_tokens": 31789761.0,
      "step": 12545
    },
    {
      "entropy": 6.394418525695801,
      "epoch": 1.4483554529717253,
      "grad_norm": 2.21875,
      "learning_rate": 0.00048011240856216456,
      "loss": 6.1803,
      "mean_token_accuracy": 0.16258420944213867,
      "num_tokens": 31802689.0,
      "step": 12550
    },
    {
      "entropy": 6.29752893447876,
      "epoch": 1.448932487016734,
      "grad_norm": 2.0,
      "learning_rate": 0.00048009544320924154,
      "loss": 6.169,
      "mean_token_accuracy": 0.16543462723493577,
      "num_tokens": 31816768.0,
      "step": 12555
    },
    {
      "entropy": 6.260969018936157,
      "epoch": 1.4495095210617426,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004800784709580351,
      "loss": 6.1266,
      "mean_token_accuracy": 0.16824788302183152,
      "num_tokens": 31828920.0,
      "step": 12560
    },
    {
      "entropy": 6.357678985595703,
      "epoch": 1.4500865551067512,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004800614918091161,
      "loss": 6.1638,
      "mean_token_accuracy": 0.1669384077191353,
      "num_tokens": 31841467.0,
      "step": 12565
    },
    {
      "entropy": 6.349787187576294,
      "epoch": 1.4506635891517599,
      "grad_norm": 3.34375,
      "learning_rate": 0.0004800445057630558,
      "loss": 6.1629,
      "mean_token_accuracy": 0.1676136553287506,
      "num_tokens": 31855254.0,
      "step": 12570
    },
    {
      "entropy": 6.27349066734314,
      "epoch": 1.4512406231967687,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004800275128204254,
      "loss": 6.1044,
      "mean_token_accuracy": 0.1662316232919693,
      "num_tokens": 31868234.0,
      "step": 12575
    },
    {
      "entropy": 6.35194959640503,
      "epoch": 1.4518176572417771,
      "grad_norm": 2.015625,
      "learning_rate": 0.00048001051298179637,
      "loss": 6.1405,
      "mean_token_accuracy": 0.17040657103061677,
      "num_tokens": 31880014.0,
      "step": 12580
    },
    {
      "entropy": 6.3423114776611325,
      "epoch": 1.452394691286786,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004799935062477407,
      "loss": 6.0906,
      "mean_token_accuracy": 0.16711047291755676,
      "num_tokens": 31892019.0,
      "step": 12585
    },
    {
      "entropy": 6.354385280609131,
      "epoch": 1.4529717253317946,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047997649261883013,
      "loss": 6.1275,
      "mean_token_accuracy": 0.15655057430267333,
      "num_tokens": 31905837.0,
      "step": 12590
    },
    {
      "entropy": 6.31659779548645,
      "epoch": 1.4535487593768033,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004799594720956371,
      "loss": 6.0835,
      "mean_token_accuracy": 0.16454610973596573,
      "num_tokens": 31917823.0,
      "step": 12595
    },
    {
      "entropy": 6.348938417434693,
      "epoch": 1.454125793421812,
      "grad_norm": 3.34375,
      "learning_rate": 0.00047994244467873407,
      "loss": 6.1813,
      "mean_token_accuracy": 0.1576891675591469,
      "num_tokens": 31931221.0,
      "step": 12600
    },
    {
      "entropy": 6.339001035690307,
      "epoch": 1.4547028274668206,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047992541036869364,
      "loss": 6.1319,
      "mean_token_accuracy": 0.16306519508361816,
      "num_tokens": 31943739.0,
      "step": 12605
    },
    {
      "entropy": 6.3493794918060305,
      "epoch": 1.4552798615118292,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004799083691660889,
      "loss": 6.1816,
      "mean_token_accuracy": 0.16041698455810546,
      "num_tokens": 31957805.0,
      "step": 12610
    },
    {
      "entropy": 6.352484798431396,
      "epoch": 1.4558568955568378,
      "grad_norm": 3.078125,
      "learning_rate": 0.0004798913210714929,
      "loss": 6.1328,
      "mean_token_accuracy": 0.17066436260938644,
      "num_tokens": 31970983.0,
      "step": 12615
    },
    {
      "entropy": 6.307969427108764,
      "epoch": 1.4564339296018465,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00047987426608547915,
      "loss": 6.1348,
      "mean_token_accuracy": 0.1637205332517624,
      "num_tokens": 31982985.0,
      "step": 12620
    },
    {
      "entropy": 6.25712218284607,
      "epoch": 1.4570109636468551,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004798572042086212,
      "loss": 6.0702,
      "mean_token_accuracy": 0.16790583580732346,
      "num_tokens": 31994854.0,
      "step": 12625
    },
    {
      "entropy": 6.328677368164063,
      "epoch": 1.4575879976918638,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047984013544149286,
      "loss": 6.2202,
      "mean_token_accuracy": 0.15574894845485687,
      "num_tokens": 32007476.0,
      "step": 12630
    },
    {
      "entropy": 6.4281259059906,
      "epoch": 1.4581650317368724,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047982305978466836,
      "loss": 6.201,
      "mean_token_accuracy": 0.16399176120758058,
      "num_tokens": 32021417.0,
      "step": 12635
    },
    {
      "entropy": 6.320074844360351,
      "epoch": 1.4587420657818813,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047980597723872205,
      "loss": 6.1079,
      "mean_token_accuracy": 0.1619710475206375,
      "num_tokens": 32033997.0,
      "step": 12640
    },
    {
      "entropy": 6.289112663269043,
      "epoch": 1.4593190998268897,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004797888878042283,
      "loss": 6.1041,
      "mean_token_accuracy": 0.1678844302892685,
      "num_tokens": 32046797.0,
      "step": 12645
    },
    {
      "entropy": 6.276427221298218,
      "epoch": 1.4598961338718985,
      "grad_norm": 2.0,
      "learning_rate": 0.00047977179148176217,
      "loss": 6.1286,
      "mean_token_accuracy": 0.1663985624909401,
      "num_tokens": 32060797.0,
      "step": 12650
    },
    {
      "entropy": 6.30776014328003,
      "epoch": 1.4604731679169072,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004797546882718985,
      "loss": 6.2171,
      "mean_token_accuracy": 0.15766320675611495,
      "num_tokens": 32073936.0,
      "step": 12655
    },
    {
      "entropy": 6.3672514915466305,
      "epoch": 1.4610502019619158,
      "grad_norm": 2.40625,
      "learning_rate": 0.00047973757817521256,
      "loss": 6.0737,
      "mean_token_accuracy": 0.1655890852212906,
      "num_tokens": 32086730.0,
      "step": 12660
    },
    {
      "entropy": 6.404605484008789,
      "epoch": 1.4616272360069245,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004797204611922799,
      "loss": 6.1674,
      "mean_token_accuracy": 0.16363464891910554,
      "num_tokens": 32099201.0,
      "step": 12665
    },
    {
      "entropy": 6.250916290283203,
      "epoch": 1.462204270051933,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047970333732367626,
      "loss": 6.1245,
      "mean_token_accuracy": 0.16768237501382827,
      "num_tokens": 32112475.0,
      "step": 12670
    },
    {
      "entropy": 6.351142501831054,
      "epoch": 1.4627813040969417,
      "grad_norm": 2.34375,
      "learning_rate": 0.00047968620656997754,
      "loss": 6.1456,
      "mean_token_accuracy": 0.16002534329891205,
      "num_tokens": 32124411.0,
      "step": 12675
    },
    {
      "entropy": 6.312311601638794,
      "epoch": 1.4633583381419504,
      "grad_norm": 1.25,
      "learning_rate": 0.00047966906893175994,
      "loss": 6.1315,
      "mean_token_accuracy": 0.17081220149993898,
      "num_tokens": 32136665.0,
      "step": 12680
    },
    {
      "entropy": 6.348316621780396,
      "epoch": 1.463935372186959,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004796519244095998,
      "loss": 6.1543,
      "mean_token_accuracy": 0.16716521829366685,
      "num_tokens": 32149546.0,
      "step": 12685
    },
    {
      "entropy": 6.273695564270019,
      "epoch": 1.4645124062319677,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047963477300407394,
      "loss": 6.0739,
      "mean_token_accuracy": 0.16554967164993287,
      "num_tokens": 32161327.0,
      "step": 12690
    },
    {
      "entropy": 6.307106685638428,
      "epoch": 1.4650894402769763,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047961761471575903,
      "loss": 6.0782,
      "mean_token_accuracy": 0.16557173281908036,
      "num_tokens": 32173678.0,
      "step": 12695
    },
    {
      "entropy": 6.207534408569336,
      "epoch": 1.465666474321985,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047960044954523237,
      "loss": 6.0288,
      "mean_token_accuracy": 0.1673967197537422,
      "num_tokens": 32186834.0,
      "step": 12700
    },
    {
      "entropy": 6.304496192932129,
      "epoch": 1.4662435083669936,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047958327749307117,
      "loss": 6.1396,
      "mean_token_accuracy": 0.17024445086717604,
      "num_tokens": 32198876.0,
      "step": 12705
    },
    {
      "entropy": 6.291701412200927,
      "epoch": 1.4668205424120022,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047956609855985305,
      "loss": 6.0548,
      "mean_token_accuracy": 0.16752492636442184,
      "num_tokens": 32212190.0,
      "step": 12710
    },
    {
      "entropy": 6.335958576202392,
      "epoch": 1.467397576457011,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004795489127461559,
      "loss": 6.1112,
      "mean_token_accuracy": 0.165318101644516,
      "num_tokens": 32226129.0,
      "step": 12715
    },
    {
      "entropy": 6.2873485565185545,
      "epoch": 1.4679746105020195,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047953172005255756,
      "loss": 6.1574,
      "mean_token_accuracy": 0.15961160510778427,
      "num_tokens": 32240116.0,
      "step": 12720
    },
    {
      "entropy": 6.333165073394776,
      "epoch": 1.4685516445470284,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004795145204796365,
      "loss": 6.0513,
      "mean_token_accuracy": 0.16849917322397232,
      "num_tokens": 32252781.0,
      "step": 12725
    },
    {
      "entropy": 6.2766618728637695,
      "epoch": 1.469128678592037,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004794973140279711,
      "loss": 6.0605,
      "mean_token_accuracy": 0.17072835117578505,
      "num_tokens": 32265625.0,
      "step": 12730
    },
    {
      "entropy": 6.324501991271973,
      "epoch": 1.4697057126370456,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047948010069814005,
      "loss": 6.1673,
      "mean_token_accuracy": 0.16713138967752456,
      "num_tokens": 32278942.0,
      "step": 12735
    },
    {
      "entropy": 6.299522256851196,
      "epoch": 1.4702827466820543,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004794628804907225,
      "loss": 6.1279,
      "mean_token_accuracy": 0.16506323367357253,
      "num_tokens": 32292085.0,
      "step": 12740
    },
    {
      "entropy": 6.298580837249756,
      "epoch": 1.470859780727063,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047944565340629755,
      "loss": 6.1307,
      "mean_token_accuracy": 0.16882706582546234,
      "num_tokens": 32303880.0,
      "step": 12745
    },
    {
      "entropy": 6.234088325500489,
      "epoch": 1.4714368147720716,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047942841944544453,
      "loss": 6.064,
      "mean_token_accuracy": 0.17083367854356765,
      "num_tokens": 32315711.0,
      "step": 12750
    },
    {
      "entropy": 6.344754409790039,
      "epoch": 1.4720138488170802,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004794111786087431,
      "loss": 6.1696,
      "mean_token_accuracy": 0.1633049726486206,
      "num_tokens": 32329344.0,
      "step": 12755
    },
    {
      "entropy": 6.303945016860962,
      "epoch": 1.4725908828620888,
      "grad_norm": 4.125,
      "learning_rate": 0.0004793939308967733,
      "loss": 6.1848,
      "mean_token_accuracy": 0.16809794306755066,
      "num_tokens": 32343335.0,
      "step": 12760
    },
    {
      "entropy": 6.258049535751343,
      "epoch": 1.4731679169070975,
      "grad_norm": 6.78125,
      "learning_rate": 0.0004793766763101152,
      "loss": 6.0793,
      "mean_token_accuracy": 0.16625706404447554,
      "num_tokens": 32357663.0,
      "step": 12765
    },
    {
      "entropy": 6.259440612792969,
      "epoch": 1.4737449509521061,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047935941484934905,
      "loss": 6.0974,
      "mean_token_accuracy": 0.1702799528837204,
      "num_tokens": 32369310.0,
      "step": 12770
    },
    {
      "entropy": 6.247601699829102,
      "epoch": 1.4743219849971148,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047934214651505547,
      "loss": 6.1111,
      "mean_token_accuracy": 0.16637591719627381,
      "num_tokens": 32381478.0,
      "step": 12775
    },
    {
      "entropy": 6.2855452537536625,
      "epoch": 1.4748990190421236,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047932487130781533,
      "loss": 6.0526,
      "mean_token_accuracy": 0.17474593371152877,
      "num_tokens": 32394697.0,
      "step": 12780
    },
    {
      "entropy": 6.227901601791382,
      "epoch": 1.475476053087132,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004793075892282097,
      "loss": 5.9675,
      "mean_token_accuracy": 0.17708137333393098,
      "num_tokens": 32405456.0,
      "step": 12785
    },
    {
      "entropy": 6.235746669769287,
      "epoch": 1.476053087132141,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004792903002768197,
      "loss": 6.0669,
      "mean_token_accuracy": 0.1676441565155983,
      "num_tokens": 32416772.0,
      "step": 12790
    },
    {
      "entropy": 6.221759462356568,
      "epoch": 1.4766301211771495,
      "grad_norm": 2.28125,
      "learning_rate": 0.00047927300445422687,
      "loss": 6.0199,
      "mean_token_accuracy": 0.17797380387783052,
      "num_tokens": 32429319.0,
      "step": 12795
    },
    {
      "entropy": 6.217854356765747,
      "epoch": 1.4772071552221582,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004792557017610131,
      "loss": 5.9956,
      "mean_token_accuracy": 0.18017663508653642,
      "num_tokens": 32442017.0,
      "step": 12800
    },
    {
      "entropy": 6.242416000366211,
      "epoch": 1.4777841892671668,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047923839219776027,
      "loss": 6.0055,
      "mean_token_accuracy": 0.17175121158361434,
      "num_tokens": 32454255.0,
      "step": 12805
    },
    {
      "entropy": 6.161380672454834,
      "epoch": 1.4783612233121755,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004792210757650506,
      "loss": 6.062,
      "mean_token_accuracy": 0.1737545147538185,
      "num_tokens": 32465471.0,
      "step": 12810
    },
    {
      "entropy": 6.248135375976562,
      "epoch": 1.478938257357184,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047920375246346636,
      "loss": 6.1001,
      "mean_token_accuracy": 0.16704308986663818,
      "num_tokens": 32478271.0,
      "step": 12815
    },
    {
      "entropy": 6.362189865112304,
      "epoch": 1.4795152914021927,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047918642229359044,
      "loss": 6.0901,
      "mean_token_accuracy": 0.16624097675085067,
      "num_tokens": 32490239.0,
      "step": 12820
    },
    {
      "entropy": 6.296052885055542,
      "epoch": 1.4800923254472014,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004791690852560056,
      "loss": 6.1359,
      "mean_token_accuracy": 0.16350751668214797,
      "num_tokens": 32503511.0,
      "step": 12825
    },
    {
      "entropy": 6.24784746170044,
      "epoch": 1.48066935949221,
      "grad_norm": 1.375,
      "learning_rate": 0.000479151741351295,
      "loss": 6.1049,
      "mean_token_accuracy": 0.16343062818050386,
      "num_tokens": 32515652.0,
      "step": 12830
    },
    {
      "entropy": 6.277792072296142,
      "epoch": 1.4812463935372187,
      "grad_norm": 1.453125,
      "learning_rate": 0.000479134390580042,
      "loss": 6.1176,
      "mean_token_accuracy": 0.16031728088855743,
      "num_tokens": 32527620.0,
      "step": 12835
    },
    {
      "entropy": 6.39299464225769,
      "epoch": 1.4818234275822273,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047911703294283015,
      "loss": 6.127,
      "mean_token_accuracy": 0.16917325854301452,
      "num_tokens": 32540948.0,
      "step": 12840
    },
    {
      "entropy": 6.256704425811767,
      "epoch": 1.482400461627236,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047909966844024334,
      "loss": 6.1118,
      "mean_token_accuracy": 0.1677668496966362,
      "num_tokens": 32553311.0,
      "step": 12845
    },
    {
      "entropy": 6.193738842010498,
      "epoch": 1.4829774956722446,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047908229707286547,
      "loss": 5.9834,
      "mean_token_accuracy": 0.17342813462018966,
      "num_tokens": 32564639.0,
      "step": 12850
    },
    {
      "entropy": 6.245530033111573,
      "epoch": 1.4835545297172534,
      "grad_norm": 1.4375,
      "learning_rate": 0.000479064918841281,
      "loss": 6.1341,
      "mean_token_accuracy": 0.164619517326355,
      "num_tokens": 32577255.0,
      "step": 12855
    },
    {
      "entropy": 6.2433384418487545,
      "epoch": 1.4841315637622619,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047904753374607427,
      "loss": 6.0864,
      "mean_token_accuracy": 0.16741010099649428,
      "num_tokens": 32590201.0,
      "step": 12860
    },
    {
      "entropy": 6.279276657104492,
      "epoch": 1.4847085978072707,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047903014178783015,
      "loss": 6.0701,
      "mean_token_accuracy": 0.16273524165153502,
      "num_tokens": 32603010.0,
      "step": 12865
    },
    {
      "entropy": 6.354335880279541,
      "epoch": 1.4852856318522794,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004790127429671335,
      "loss": 6.0863,
      "mean_token_accuracy": 0.1656494528055191,
      "num_tokens": 32614579.0,
      "step": 12870
    },
    {
      "entropy": 6.246316576004029,
      "epoch": 1.485862665897288,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004789953372845697,
      "loss": 6.064,
      "mean_token_accuracy": 0.1642582133412361,
      "num_tokens": 32627696.0,
      "step": 12875
    },
    {
      "entropy": 6.24684886932373,
      "epoch": 1.4864396999422966,
      "grad_norm": 1.375,
      "learning_rate": 0.000478977924740724,
      "loss": 6.171,
      "mean_token_accuracy": 0.15860685259103774,
      "num_tokens": 32640824.0,
      "step": 12880
    },
    {
      "entropy": 6.320673418045044,
      "epoch": 1.4870167339873053,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004789605053361821,
      "loss": 6.0864,
      "mean_token_accuracy": 0.16608539447188378,
      "num_tokens": 32654373.0,
      "step": 12885
    },
    {
      "entropy": 6.3553577899932865,
      "epoch": 1.487593768032314,
      "grad_norm": 1.25,
      "learning_rate": 0.0004789430790715299,
      "loss": 6.206,
      "mean_token_accuracy": 0.1578859195113182,
      "num_tokens": 32667717.0,
      "step": 12890
    },
    {
      "entropy": 6.260083961486816,
      "epoch": 1.4881708020773226,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047892564594735355,
      "loss": 6.1414,
      "mean_token_accuracy": 0.16327641457319259,
      "num_tokens": 32681529.0,
      "step": 12895
    },
    {
      "entropy": 6.357323503494262,
      "epoch": 1.4887478361223312,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047890820596423943,
      "loss": 6.1894,
      "mean_token_accuracy": 0.16188809275627136,
      "num_tokens": 32694883.0,
      "step": 12900
    },
    {
      "entropy": 6.320400857925415,
      "epoch": 1.4893248701673398,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000478890759122774,
      "loss": 6.0706,
      "mean_token_accuracy": 0.1698184132575989,
      "num_tokens": 32706908.0,
      "step": 12905
    },
    {
      "entropy": 6.324786186218262,
      "epoch": 1.4899019042123485,
      "grad_norm": 1.875,
      "learning_rate": 0.0004788733054235443,
      "loss": 6.1161,
      "mean_token_accuracy": 0.1591644510626793,
      "num_tokens": 32720057.0,
      "step": 12910
    },
    {
      "entropy": 6.208461332321167,
      "epoch": 1.4904789382573571,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047885584486713717,
      "loss": 6.0907,
      "mean_token_accuracy": 0.17133676409721374,
      "num_tokens": 32732200.0,
      "step": 12915
    },
    {
      "entropy": 6.282963752746582,
      "epoch": 1.491055972302366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004788383774541399,
      "loss": 6.0879,
      "mean_token_accuracy": 0.16833843439817428,
      "num_tokens": 32745144.0,
      "step": 12920
    },
    {
      "entropy": 6.295819664001465,
      "epoch": 1.4916330063473744,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004788209031851402,
      "loss": 6.122,
      "mean_token_accuracy": 0.1617732271552086,
      "num_tokens": 32757495.0,
      "step": 12925
    },
    {
      "entropy": 6.223090553283692,
      "epoch": 1.4922100403923833,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004788034220607256,
      "loss": 5.9892,
      "mean_token_accuracy": 0.1723991960287094,
      "num_tokens": 32769562.0,
      "step": 12930
    },
    {
      "entropy": 6.260346746444702,
      "epoch": 1.4927870744373917,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00047878593408148405,
      "loss": 6.0792,
      "mean_token_accuracy": 0.16787715405225753,
      "num_tokens": 32782234.0,
      "step": 12935
    },
    {
      "entropy": 6.21394419670105,
      "epoch": 1.4933641084824005,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047876843924800393,
      "loss": 6.0872,
      "mean_token_accuracy": 0.16682145297527312,
      "num_tokens": 32795274.0,
      "step": 12940
    },
    {
      "entropy": 6.271086740493774,
      "epoch": 1.4939411425274092,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004787509375608735,
      "loss": 6.0843,
      "mean_token_accuracy": 0.15776642337441443,
      "num_tokens": 32808338.0,
      "step": 12945
    },
    {
      "entropy": 6.294681644439697,
      "epoch": 1.4945181765724178,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047873342902068157,
      "loss": 6.0906,
      "mean_token_accuracy": 0.16260672509670257,
      "num_tokens": 32820204.0,
      "step": 12950
    },
    {
      "entropy": 6.337964868545532,
      "epoch": 1.4950952106174265,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047871591362801694,
      "loss": 6.139,
      "mean_token_accuracy": 0.1639711320400238,
      "num_tokens": 32832998.0,
      "step": 12955
    },
    {
      "entropy": 6.223607349395752,
      "epoch": 1.495672244662435,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004786983913834687,
      "loss": 6.1281,
      "mean_token_accuracy": 0.16615548431873323,
      "num_tokens": 32845484.0,
      "step": 12960
    },
    {
      "entropy": 6.414913702011108,
      "epoch": 1.4962492787074437,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047868086228762633,
      "loss": 6.1956,
      "mean_token_accuracy": 0.16196768283843993,
      "num_tokens": 32857718.0,
      "step": 12965
    },
    {
      "entropy": 6.343223285675049,
      "epoch": 1.4968263127524524,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047866332634107926,
      "loss": 6.1177,
      "mean_token_accuracy": 0.16523806303739547,
      "num_tokens": 32869908.0,
      "step": 12970
    },
    {
      "entropy": 6.271601438522339,
      "epoch": 1.497403346797461,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047864578354441743,
      "loss": 6.1172,
      "mean_token_accuracy": 0.16479604095220565,
      "num_tokens": 32882549.0,
      "step": 12975
    },
    {
      "entropy": 6.2289995670318605,
      "epoch": 1.4979803808424696,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004786282338982308,
      "loss": 6.0305,
      "mean_token_accuracy": 0.1769894078373909,
      "num_tokens": 32894587.0,
      "step": 12980
    },
    {
      "entropy": 6.339492082595825,
      "epoch": 1.4985574148874783,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004786106774031096,
      "loss": 6.1258,
      "mean_token_accuracy": 0.16558513790369034,
      "num_tokens": 32907001.0,
      "step": 12985
    },
    {
      "entropy": 6.32451868057251,
      "epoch": 1.499134448932487,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004785931140596445,
      "loss": 6.059,
      "mean_token_accuracy": 0.167315773665905,
      "num_tokens": 32919498.0,
      "step": 12990
    },
    {
      "entropy": 6.3400568008422855,
      "epoch": 1.4997114829774958,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047857554386842606,
      "loss": 6.1788,
      "mean_token_accuracy": 0.1587929204106331,
      "num_tokens": 32932611.0,
      "step": 12995
    },
    {
      "entropy": 6.306359529495239,
      "epoch": 1.5002885170225042,
      "grad_norm": 2.0,
      "learning_rate": 0.00047855796683004534,
      "loss": 6.0872,
      "mean_token_accuracy": 0.16214493662118912,
      "num_tokens": 32944845.0,
      "step": 13000
    },
    {
      "entropy": 6.368089056015014,
      "epoch": 1.500865551067513,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047854038294509356,
      "loss": 6.1238,
      "mean_token_accuracy": 0.1523377075791359,
      "num_tokens": 32958076.0,
      "step": 13005
    },
    {
      "entropy": 6.357781219482422,
      "epoch": 1.5014425851125215,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004785227922141621,
      "loss": 6.0859,
      "mean_token_accuracy": 0.1753346249461174,
      "num_tokens": 32968984.0,
      "step": 13010
    },
    {
      "entropy": 6.359879016876221,
      "epoch": 1.5020196191575304,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047850519463784263,
      "loss": 6.228,
      "mean_token_accuracy": 0.1593037635087967,
      "num_tokens": 32981488.0,
      "step": 13015
    },
    {
      "entropy": 6.3068126201629635,
      "epoch": 1.502596653202539,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047848759021672693,
      "loss": 6.1443,
      "mean_token_accuracy": 0.16349516659975052,
      "num_tokens": 32994156.0,
      "step": 13020
    },
    {
      "entropy": 6.335974407196045,
      "epoch": 1.5031736872475476,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004784699789514073,
      "loss": 6.0814,
      "mean_token_accuracy": 0.16241802722215654,
      "num_tokens": 33007135.0,
      "step": 13025
    },
    {
      "entropy": 6.35773777961731,
      "epoch": 1.5037507212925563,
      "grad_norm": 2.0625,
      "learning_rate": 0.000478452360842476,
      "loss": 6.1742,
      "mean_token_accuracy": 0.15925257354974748,
      "num_tokens": 33021259.0,
      "step": 13030
    },
    {
      "entropy": 6.268192148208618,
      "epoch": 1.504327755337565,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047843473589052557,
      "loss": 6.1114,
      "mean_token_accuracy": 0.16721211671829223,
      "num_tokens": 33033690.0,
      "step": 13035
    },
    {
      "entropy": 6.382220840454101,
      "epoch": 1.5049047893825735,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047841710409614893,
      "loss": 6.1256,
      "mean_token_accuracy": 0.16039337813854218,
      "num_tokens": 33045744.0,
      "step": 13040
    },
    {
      "entropy": 6.359376525878906,
      "epoch": 1.5054818234275822,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004783994654599389,
      "loss": 6.1675,
      "mean_token_accuracy": 0.15905430912971497,
      "num_tokens": 33058307.0,
      "step": 13045
    },
    {
      "entropy": 6.391716718673706,
      "epoch": 1.5060588574725908,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047838181998248897,
      "loss": 6.2093,
      "mean_token_accuracy": 0.15556481406092643,
      "num_tokens": 33073622.0,
      "step": 13050
    },
    {
      "entropy": 6.295185232162476,
      "epoch": 1.5066358915175995,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004783641676643925,
      "loss": 6.0359,
      "mean_token_accuracy": 0.16836757212877274,
      "num_tokens": 33087286.0,
      "step": 13055
    },
    {
      "entropy": 6.236673784255982,
      "epoch": 1.5072129255626083,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047834650850624334,
      "loss": 6.109,
      "mean_token_accuracy": 0.1679818168282509,
      "num_tokens": 33100333.0,
      "step": 13060
    },
    {
      "entropy": 6.2044658184051515,
      "epoch": 1.5077899596076167,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004783288425086353,
      "loss": 6.1113,
      "mean_token_accuracy": 0.1649872049689293,
      "num_tokens": 33111318.0,
      "step": 13065
    },
    {
      "entropy": 6.406934833526611,
      "epoch": 1.5083669936526256,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004783111696721627,
      "loss": 6.1825,
      "mean_token_accuracy": 0.16137512922286987,
      "num_tokens": 33123766.0,
      "step": 13070
    },
    {
      "entropy": 6.325654792785644,
      "epoch": 1.508944027697634,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004782934899974199,
      "loss": 6.0832,
      "mean_token_accuracy": 0.17172017246484755,
      "num_tokens": 33136042.0,
      "step": 13075
    },
    {
      "entropy": 6.260492467880249,
      "epoch": 1.5095210617426429,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047827580348500147,
      "loss": 6.0174,
      "mean_token_accuracy": 0.17097728699445724,
      "num_tokens": 33148451.0,
      "step": 13080
    },
    {
      "entropy": 6.268058347702026,
      "epoch": 1.5100980957876513,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047825811013550246,
      "loss": 6.0869,
      "mean_token_accuracy": 0.1745191514492035,
      "num_tokens": 33161886.0,
      "step": 13085
    },
    {
      "entropy": 6.308102035522461,
      "epoch": 1.5106751298326602,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047824040994951786,
      "loss": 6.1993,
      "mean_token_accuracy": 0.16559694707393646,
      "num_tokens": 33174812.0,
      "step": 13090
    },
    {
      "entropy": 6.3163261890411375,
      "epoch": 1.5112521638776688,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004782227029276429,
      "loss": 6.1028,
      "mean_token_accuracy": 0.17115537822246552,
      "num_tokens": 33187389.0,
      "step": 13095
    },
    {
      "entropy": 6.218478345870972,
      "epoch": 1.5118291979226774,
      "grad_norm": 1.5,
      "learning_rate": 0.00047820498907047345,
      "loss": 6.0571,
      "mean_token_accuracy": 0.16387352347373962,
      "num_tokens": 33200106.0,
      "step": 13100
    },
    {
      "entropy": 6.287623453140259,
      "epoch": 1.512406231967686,
      "grad_norm": 3.109375,
      "learning_rate": 0.000478187268378605,
      "loss": 6.081,
      "mean_token_accuracy": 0.16405822485685348,
      "num_tokens": 33212929.0,
      "step": 13105
    },
    {
      "entropy": 6.3526145935058596,
      "epoch": 1.5129832660126947,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047816954085263375,
      "loss": 6.1577,
      "mean_token_accuracy": 0.1622457042336464,
      "num_tokens": 33226068.0,
      "step": 13110
    },
    {
      "entropy": 6.265248584747314,
      "epoch": 1.5135603000577034,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004781518064931559,
      "loss": 6.0269,
      "mean_token_accuracy": 0.16928547471761704,
      "num_tokens": 33237677.0,
      "step": 13115
    },
    {
      "entropy": 6.333219432830811,
      "epoch": 1.514137334102712,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000478134065300768,
      "loss": 6.1473,
      "mean_token_accuracy": 0.16705526113510133,
      "num_tokens": 33250640.0,
      "step": 13120
    },
    {
      "entropy": 6.280381155014038,
      "epoch": 1.5147143681477209,
      "grad_norm": 2.125,
      "learning_rate": 0.0004781163172760667,
      "loss": 6.0739,
      "mean_token_accuracy": 0.17093583345413207,
      "num_tokens": 33262899.0,
      "step": 13125
    },
    {
      "entropy": 6.378185701370239,
      "epoch": 1.5152914021927293,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047809856241964893,
      "loss": 6.08,
      "mean_token_accuracy": 0.16901710629463196,
      "num_tokens": 33275073.0,
      "step": 13130
    },
    {
      "entropy": 6.205471515655518,
      "epoch": 1.5158684362377381,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004780808007321119,
      "loss": 6.0516,
      "mean_token_accuracy": 0.16522097885608672,
      "num_tokens": 33286285.0,
      "step": 13135
    },
    {
      "entropy": 6.32527985572815,
      "epoch": 1.5164454702827466,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004780630322140531,
      "loss": 6.049,
      "mean_token_accuracy": 0.1732782319188118,
      "num_tokens": 33297745.0,
      "step": 13140
    },
    {
      "entropy": 6.356187438964843,
      "epoch": 1.5170225043277554,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047804525686607,
      "loss": 6.1214,
      "mean_token_accuracy": 0.15876830518245696,
      "num_tokens": 33309463.0,
      "step": 13145
    },
    {
      "entropy": 6.308989381790161,
      "epoch": 1.5175995383727638,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004780274746887606,
      "loss": 6.147,
      "mean_token_accuracy": 0.15899598598480225,
      "num_tokens": 33322352.0,
      "step": 13150
    },
    {
      "entropy": 6.343785285949707,
      "epoch": 1.5181765724177727,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047800968568272284,
      "loss": 6.1776,
      "mean_token_accuracy": 0.16358117014169693,
      "num_tokens": 33336051.0,
      "step": 13155
    },
    {
      "entropy": 6.3434113502502445,
      "epoch": 1.5187536064627813,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004779918898485551,
      "loss": 6.1022,
      "mean_token_accuracy": 0.16069099232554435,
      "num_tokens": 33350321.0,
      "step": 13160
    },
    {
      "entropy": 6.337099504470825,
      "epoch": 1.51933064050779,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047797408718685614,
      "loss": 6.0725,
      "mean_token_accuracy": 0.16448751240968704,
      "num_tokens": 33363637.0,
      "step": 13165
    },
    {
      "entropy": 6.224223947525024,
      "epoch": 1.5199076745527986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047795627769822447,
      "loss": 6.0496,
      "mean_token_accuracy": 0.17334780991077423,
      "num_tokens": 33376257.0,
      "step": 13170
    },
    {
      "entropy": 6.165194988250732,
      "epoch": 1.5204847085978073,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047793846138325925,
      "loss": 6.0099,
      "mean_token_accuracy": 0.1717988818883896,
      "num_tokens": 33388812.0,
      "step": 13175
    },
    {
      "entropy": 6.248253870010376,
      "epoch": 1.521061742642816,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004779206382425598,
      "loss": 6.1,
      "mean_token_accuracy": 0.1664251208305359,
      "num_tokens": 33401141.0,
      "step": 13180
    },
    {
      "entropy": 6.297205305099487,
      "epoch": 1.5216387766878245,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004779028082767253,
      "loss": 6.0997,
      "mean_token_accuracy": 0.16420046240091324,
      "num_tokens": 33412821.0,
      "step": 13185
    },
    {
      "entropy": 6.258190584182739,
      "epoch": 1.5222158107328332,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004778849714863557,
      "loss": 6.1259,
      "mean_token_accuracy": 0.1637619987130165,
      "num_tokens": 33424709.0,
      "step": 13190
    },
    {
      "entropy": 6.308904504776001,
      "epoch": 1.5227928447778418,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004778671278720508,
      "loss": 6.1683,
      "mean_token_accuracy": 0.1601940304040909,
      "num_tokens": 33437069.0,
      "step": 13195
    },
    {
      "entropy": 6.263119840621949,
      "epoch": 1.5233698788228507,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004778492774344109,
      "loss": 5.978,
      "mean_token_accuracy": 0.1729772388935089,
      "num_tokens": 33449441.0,
      "step": 13200
    },
    {
      "entropy": 6.2551123142242435,
      "epoch": 1.523946912867859,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004778314201740363,
      "loss": 6.1833,
      "mean_token_accuracy": 0.15870602279901505,
      "num_tokens": 33463184.0,
      "step": 13205
    },
    {
      "entropy": 6.308429145812989,
      "epoch": 1.524523946912868,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047781355609152764,
      "loss": 6.0285,
      "mean_token_accuracy": 0.17487031370401382,
      "num_tokens": 33476118.0,
      "step": 13210
    },
    {
      "entropy": 6.277584075927734,
      "epoch": 1.5251009809578764,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004777956851874858,
      "loss": 6.0615,
      "mean_token_accuracy": 0.16608939617872237,
      "num_tokens": 33487575.0,
      "step": 13215
    },
    {
      "entropy": 6.237762451171875,
      "epoch": 1.5256780150028852,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047777780746251176,
      "loss": 6.1003,
      "mean_token_accuracy": 0.1641905725002289,
      "num_tokens": 33499507.0,
      "step": 13220
    },
    {
      "entropy": 6.2443382263183596,
      "epoch": 1.5262550490478937,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047775992291720687,
      "loss": 6.083,
      "mean_token_accuracy": 0.16951340883970262,
      "num_tokens": 33513818.0,
      "step": 13225
    },
    {
      "entropy": 6.284352970123291,
      "epoch": 1.5268320830929025,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004777420315521728,
      "loss": 6.0771,
      "mean_token_accuracy": 0.16708099693059922,
      "num_tokens": 33525762.0,
      "step": 13230
    },
    {
      "entropy": 6.191858911514283,
      "epoch": 1.5274091171379112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004777241333680111,
      "loss": 5.9738,
      "mean_token_accuracy": 0.1767125442624092,
      "num_tokens": 33537469.0,
      "step": 13235
    },
    {
      "entropy": 6.221562433242798,
      "epoch": 1.5279861511829198,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004777062283653239,
      "loss": 6.0999,
      "mean_token_accuracy": 0.17175007611513138,
      "num_tokens": 33550226.0,
      "step": 13240
    },
    {
      "entropy": 6.314879369735718,
      "epoch": 1.5285631852279284,
      "grad_norm": 2.515625,
      "learning_rate": 0.00047768831654471333,
      "loss": 6.1739,
      "mean_token_accuracy": 0.16319517195224761,
      "num_tokens": 33564164.0,
      "step": 13245
    },
    {
      "entropy": 6.277708244323731,
      "epoch": 1.529140219272937,
      "grad_norm": 1.25,
      "learning_rate": 0.00047767039790678204,
      "loss": 5.9983,
      "mean_token_accuracy": 0.16776868253946303,
      "num_tokens": 33576295.0,
      "step": 13250
    },
    {
      "entropy": 6.210209226608276,
      "epoch": 1.5297172533179457,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047765247245213255,
      "loss": 6.055,
      "mean_token_accuracy": 0.17595264613628386,
      "num_tokens": 33587950.0,
      "step": 13255
    },
    {
      "entropy": 6.216668558120728,
      "epoch": 1.5302942873629544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004776345401813678,
      "loss": 6.0519,
      "mean_token_accuracy": 0.17725253403186797,
      "num_tokens": 33600201.0,
      "step": 13260
    },
    {
      "entropy": 6.2638702392578125,
      "epoch": 1.5308713214079632,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047761660109509095,
      "loss": 6.1106,
      "mean_token_accuracy": 0.1715979754924774,
      "num_tokens": 33612355.0,
      "step": 13265
    },
    {
      "entropy": 6.342147922515869,
      "epoch": 1.5314483554529716,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004775986551939054,
      "loss": 6.1352,
      "mean_token_accuracy": 0.16366921961307526,
      "num_tokens": 33626322.0,
      "step": 13270
    },
    {
      "entropy": 6.246899604797363,
      "epoch": 1.5320253894979805,
      "grad_norm": 1.25,
      "learning_rate": 0.00047758070247841465,
      "loss": 6.0532,
      "mean_token_accuracy": 0.16893962174654006,
      "num_tokens": 33638433.0,
      "step": 13275
    },
    {
      "entropy": 6.2625486850738525,
      "epoch": 1.532602423542989,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047756274294922266,
      "loss": 6.0969,
      "mean_token_accuracy": 0.1695482537150383,
      "num_tokens": 33651743.0,
      "step": 13280
    },
    {
      "entropy": 6.351377201080322,
      "epoch": 1.5331794575879978,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004775447766069334,
      "loss": 6.1451,
      "mean_token_accuracy": 0.1712990552186966,
      "num_tokens": 33663546.0,
      "step": 13285
    },
    {
      "entropy": 6.283286142349243,
      "epoch": 1.5337564916330062,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047752680345215115,
      "loss": 6.101,
      "mean_token_accuracy": 0.16872385144233704,
      "num_tokens": 33675358.0,
      "step": 13290
    },
    {
      "entropy": 6.26472840309143,
      "epoch": 1.534333525678015,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004775088234854805,
      "loss": 6.0767,
      "mean_token_accuracy": 0.16253549009561538,
      "num_tokens": 33688066.0,
      "step": 13295
    },
    {
      "entropy": 6.2303605556488035,
      "epoch": 1.5349105597230237,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004774908367075262,
      "loss": 6.087,
      "mean_token_accuracy": 0.17206478714942933,
      "num_tokens": 33700829.0,
      "step": 13300
    },
    {
      "entropy": 6.2332827091217045,
      "epoch": 1.5354875937680323,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004774728431188931,
      "loss": 6.0588,
      "mean_token_accuracy": 0.1641959458589554,
      "num_tokens": 33714049.0,
      "step": 13305
    },
    {
      "entropy": 6.240980577468872,
      "epoch": 1.536064627813041,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047745484272018664,
      "loss": 6.0188,
      "mean_token_accuracy": 0.17436153143644334,
      "num_tokens": 33727026.0,
      "step": 13310
    },
    {
      "entropy": 6.289119005203247,
      "epoch": 1.5366416618580496,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004774368355120119,
      "loss": 6.057,
      "mean_token_accuracy": 0.17154422402381897,
      "num_tokens": 33740400.0,
      "step": 13315
    },
    {
      "entropy": 6.1061982154846195,
      "epoch": 1.5372186959030583,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000477418821494975,
      "loss": 5.8766,
      "mean_token_accuracy": 0.1809625193476677,
      "num_tokens": 33753621.0,
      "step": 13320
    },
    {
      "entropy": 6.258587503433228,
      "epoch": 1.537795729948067,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004774008006696814,
      "loss": 6.1166,
      "mean_token_accuracy": 0.1636723607778549,
      "num_tokens": 33765458.0,
      "step": 13325
    },
    {
      "entropy": 6.32735013961792,
      "epoch": 1.5383727639930755,
      "grad_norm": 2.90625,
      "learning_rate": 0.0004773827730367375,
      "loss": 6.144,
      "mean_token_accuracy": 0.16260947287082672,
      "num_tokens": 33777264.0,
      "step": 13330
    },
    {
      "entropy": 6.311409664154053,
      "epoch": 1.5389497980380842,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047736473859674955,
      "loss": 6.0977,
      "mean_token_accuracy": 0.17010506838560105,
      "num_tokens": 33789837.0,
      "step": 13335
    },
    {
      "entropy": 6.123340892791748,
      "epoch": 1.539526832083093,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047734669735032404,
      "loss": 5.9457,
      "mean_token_accuracy": 0.17906959354877472,
      "num_tokens": 33801666.0,
      "step": 13340
    },
    {
      "entropy": 6.2019556045532225,
      "epoch": 1.5401038661281015,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047732864929806796,
      "loss": 6.0901,
      "mean_token_accuracy": 0.1611937940120697,
      "num_tokens": 33813909.0,
      "step": 13345
    },
    {
      "entropy": 6.256455135345459,
      "epoch": 1.5406809001731103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004773105944405883,
      "loss": 6.0876,
      "mean_token_accuracy": 0.17180472016334533,
      "num_tokens": 33827081.0,
      "step": 13350
    },
    {
      "entropy": 6.224641418457031,
      "epoch": 1.5412579342181187,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047729253277849226,
      "loss": 6.0217,
      "mean_token_accuracy": 0.17465359568595887,
      "num_tokens": 33838261.0,
      "step": 13355
    },
    {
      "entropy": 6.285533332824707,
      "epoch": 1.5418349682631276,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047727446431238734,
      "loss": 6.1732,
      "mean_token_accuracy": 0.16222093552350997,
      "num_tokens": 33850189.0,
      "step": 13360
    },
    {
      "entropy": 6.394769525527954,
      "epoch": 1.542412002308136,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004772563890428813,
      "loss": 6.1958,
      "mean_token_accuracy": 0.16250339150428772,
      "num_tokens": 33863570.0,
      "step": 13365
    },
    {
      "entropy": 6.269595718383789,
      "epoch": 1.5429890363531449,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004772383069705821,
      "loss": 6.1162,
      "mean_token_accuracy": 0.1639903500676155,
      "num_tokens": 33876583.0,
      "step": 13370
    },
    {
      "entropy": 6.305620288848877,
      "epoch": 1.5435660703981535,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004772202180960978,
      "loss": 6.1124,
      "mean_token_accuracy": 0.16432901620864868,
      "num_tokens": 33889744.0,
      "step": 13375
    },
    {
      "entropy": 6.382720422744751,
      "epoch": 1.5441431044431622,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047720212242003703,
      "loss": 6.1648,
      "mean_token_accuracy": 0.16586382687091827,
      "num_tokens": 33904129.0,
      "step": 13380
    },
    {
      "entropy": 6.2075355052948,
      "epoch": 1.5447201384881708,
      "grad_norm": 1.25,
      "learning_rate": 0.00047718401994300825,
      "loss": 6.0131,
      "mean_token_accuracy": 0.173862124979496,
      "num_tokens": 33916277.0,
      "step": 13385
    },
    {
      "entropy": 6.262147378921509,
      "epoch": 1.5452971725331794,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047716591066562043,
      "loss": 6.1458,
      "mean_token_accuracy": 0.167170250415802,
      "num_tokens": 33929434.0,
      "step": 13390
    },
    {
      "entropy": 6.284850931167602,
      "epoch": 1.545874206578188,
      "grad_norm": 1.375,
      "learning_rate": 0.00047714779458848255,
      "loss": 6.1036,
      "mean_token_accuracy": 0.16530850380659104,
      "num_tokens": 33942376.0,
      "step": 13395
    },
    {
      "entropy": 6.329345703125,
      "epoch": 1.5464512406231967,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004771296717122041,
      "loss": 6.1289,
      "mean_token_accuracy": 0.16672763228416443,
      "num_tokens": 33955013.0,
      "step": 13400
    },
    {
      "entropy": 6.318344497680664,
      "epoch": 1.5470282746682056,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004771115420373945,
      "loss": 6.1355,
      "mean_token_accuracy": 0.1615862876176834,
      "num_tokens": 33966673.0,
      "step": 13405
    },
    {
      "entropy": 6.227231454849243,
      "epoch": 1.547605308713214,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047709340556466366,
      "loss": 6.0523,
      "mean_token_accuracy": 0.1722373753786087,
      "num_tokens": 33978400.0,
      "step": 13410
    },
    {
      "entropy": 6.331748580932617,
      "epoch": 1.5481823427582229,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047707526229462144,
      "loss": 6.1714,
      "mean_token_accuracy": 0.1605384096503258,
      "num_tokens": 33990690.0,
      "step": 13415
    },
    {
      "entropy": 6.308060550689698,
      "epoch": 1.5487593768032313,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047705711222787816,
      "loss": 6.0964,
      "mean_token_accuracy": 0.16687259823083878,
      "num_tokens": 34003992.0,
      "step": 13420
    },
    {
      "entropy": 6.286748695373535,
      "epoch": 1.5493364108482401,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047703895536504423,
      "loss": 6.0837,
      "mean_token_accuracy": 0.1681692436337471,
      "num_tokens": 34016075.0,
      "step": 13425
    },
    {
      "entropy": 6.280511522293091,
      "epoch": 1.5499134448932486,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004770207917067305,
      "loss": 6.0552,
      "mean_token_accuracy": 0.16778141260147095,
      "num_tokens": 34029143.0,
      "step": 13430
    },
    {
      "entropy": 6.256841468811035,
      "epoch": 1.5504904789382574,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047700262125354765,
      "loss": 6.1263,
      "mean_token_accuracy": 0.17276279330253602,
      "num_tokens": 34041269.0,
      "step": 13435
    },
    {
      "entropy": 6.292127561569214,
      "epoch": 1.551067512983266,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047698444400610707,
      "loss": 6.1481,
      "mean_token_accuracy": 0.16719116121530533,
      "num_tokens": 34054561.0,
      "step": 13440
    },
    {
      "entropy": 6.344380474090576,
      "epoch": 1.5516445470282747,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047696625996502,
      "loss": 6.0748,
      "mean_token_accuracy": 0.1604306936264038,
      "num_tokens": 34068009.0,
      "step": 13445
    },
    {
      "entropy": 6.280494642257691,
      "epoch": 1.5522215810732833,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047694806913089815,
      "loss": 6.1908,
      "mean_token_accuracy": 0.16584864556789397,
      "num_tokens": 34081217.0,
      "step": 13450
    },
    {
      "entropy": 6.321868991851806,
      "epoch": 1.552798615118292,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004769298715043533,
      "loss": 6.1148,
      "mean_token_accuracy": 0.1603361338376999,
      "num_tokens": 34093903.0,
      "step": 13455
    },
    {
      "entropy": 6.34748387336731,
      "epoch": 1.5533756491633006,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004769116670859975,
      "loss": 6.1551,
      "mean_token_accuracy": 0.16427400708198547,
      "num_tokens": 34105891.0,
      "step": 13460
    },
    {
      "entropy": 6.295809173583985,
      "epoch": 1.5539526832083093,
      "grad_norm": 1.25,
      "learning_rate": 0.00047689345587644314,
      "loss": 6.1441,
      "mean_token_accuracy": 0.16782208383083344,
      "num_tokens": 34118438.0,
      "step": 13465
    },
    {
      "entropy": 6.269343376159668,
      "epoch": 1.554529717253318,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047687523787630256,
      "loss": 6.0041,
      "mean_token_accuracy": 0.17534742504358292,
      "num_tokens": 34130383.0,
      "step": 13470
    },
    {
      "entropy": 6.214257144927979,
      "epoch": 1.5551067512983265,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004768570130861887,
      "loss": 6.1249,
      "mean_token_accuracy": 0.16669410318136216,
      "num_tokens": 34143805.0,
      "step": 13475
    },
    {
      "entropy": 6.295845174789429,
      "epoch": 1.5556837853433354,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004768387815067144,
      "loss": 6.0329,
      "mean_token_accuracy": 0.1684045284986496,
      "num_tokens": 34157053.0,
      "step": 13480
    },
    {
      "entropy": 6.367346620559692,
      "epoch": 1.5562608193883438,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047682054313849304,
      "loss": 6.1567,
      "mean_token_accuracy": 0.16520717442035676,
      "num_tokens": 34168804.0,
      "step": 13485
    },
    {
      "entropy": 6.327089548110962,
      "epoch": 1.5568378534333527,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004768022979821379,
      "loss": 6.142,
      "mean_token_accuracy": 0.16089055836200714,
      "num_tokens": 34181201.0,
      "step": 13490
    },
    {
      "entropy": 6.327263450622558,
      "epoch": 1.557414887478361,
      "grad_norm": 1.625,
      "learning_rate": 0.0004767840460382628,
      "loss": 6.1518,
      "mean_token_accuracy": 0.16688635051250458,
      "num_tokens": 34193906.0,
      "step": 13495
    },
    {
      "entropy": 6.298494577407837,
      "epoch": 1.55799192152337,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004767657873074815,
      "loss": 6.148,
      "mean_token_accuracy": 0.16546330600976944,
      "num_tokens": 34207462.0,
      "step": 13500
    },
    {
      "entropy": 6.263404703140258,
      "epoch": 1.5585689555683784,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004767475217904081,
      "loss": 6.0006,
      "mean_token_accuracy": 0.1721050813794136,
      "num_tokens": 34219696.0,
      "step": 13505
    },
    {
      "entropy": 6.25515718460083,
      "epoch": 1.5591459896133872,
      "grad_norm": 2.828125,
      "learning_rate": 0.00047672924948765705,
      "loss": 6.0677,
      "mean_token_accuracy": 0.17219026386737823,
      "num_tokens": 34232413.0,
      "step": 13510
    },
    {
      "entropy": 6.330908012390137,
      "epoch": 1.5597230236583959,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047671097039984293,
      "loss": 6.1494,
      "mean_token_accuracy": 0.16325631588697434,
      "num_tokens": 34245150.0,
      "step": 13515
    },
    {
      "entropy": 6.307193422317505,
      "epoch": 1.5603000577034045,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047669268452758053,
      "loss": 6.0021,
      "mean_token_accuracy": 0.1737485185265541,
      "num_tokens": 34257416.0,
      "step": 13520
    },
    {
      "entropy": 6.241674613952637,
      "epoch": 1.5608770917484132,
      "grad_norm": 3.03125,
      "learning_rate": 0.00047667439187148476,
      "loss": 6.0485,
      "mean_token_accuracy": 0.17202006727457048,
      "num_tokens": 34269518.0,
      "step": 13525
    },
    {
      "entropy": 6.244004392623902,
      "epoch": 1.5614541257934218,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004766560924321711,
      "loss": 6.0468,
      "mean_token_accuracy": 0.1732081651687622,
      "num_tokens": 34280965.0,
      "step": 13530
    },
    {
      "entropy": 6.3121312141418455,
      "epoch": 1.5620311598384304,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047663778621025496,
      "loss": 6.1132,
      "mean_token_accuracy": 0.16684572100639344,
      "num_tokens": 34294051.0,
      "step": 13535
    },
    {
      "entropy": 6.288514518737793,
      "epoch": 1.562608193883439,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004766194732063519,
      "loss": 6.1479,
      "mean_token_accuracy": 0.15813490897417068,
      "num_tokens": 34308052.0,
      "step": 13540
    },
    {
      "entropy": 6.302996253967285,
      "epoch": 1.563185227928448,
      "grad_norm": 2.859375,
      "learning_rate": 0.00047660115342107814,
      "loss": 6.1005,
      "mean_token_accuracy": 0.16520090252161027,
      "num_tokens": 34321802.0,
      "step": 13545
    },
    {
      "entropy": 6.25574312210083,
      "epoch": 1.5637622619734564,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047658282685504973,
      "loss": 6.0311,
      "mean_token_accuracy": 0.17565805912017823,
      "num_tokens": 34334789.0,
      "step": 13550
    },
    {
      "entropy": 6.323505210876465,
      "epoch": 1.5643392960184652,
      "grad_norm": 1.28125,
      "learning_rate": 0.000476564493508883,
      "loss": 6.0697,
      "mean_token_accuracy": 0.1631614997982979,
      "num_tokens": 34347167.0,
      "step": 13555
    },
    {
      "entropy": 6.392896747589111,
      "epoch": 1.5649163300634736,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047654615338319466,
      "loss": 6.2075,
      "mean_token_accuracy": 0.15777474641799927,
      "num_tokens": 34360468.0,
      "step": 13560
    },
    {
      "entropy": 6.371495723724365,
      "epoch": 1.5654933641084825,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004765278064786016,
      "loss": 6.1411,
      "mean_token_accuracy": 0.16007252633571625,
      "num_tokens": 34373835.0,
      "step": 13565
    },
    {
      "entropy": 6.312303352355957,
      "epoch": 1.566070398153491,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00047650945279572085,
      "loss": 6.1172,
      "mean_token_accuracy": 0.16909689903259278,
      "num_tokens": 34385591.0,
      "step": 13570
    },
    {
      "entropy": 6.34000563621521,
      "epoch": 1.5666474321984998,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004764910923351698,
      "loss": 6.1949,
      "mean_token_accuracy": 0.15573060512542725,
      "num_tokens": 34398980.0,
      "step": 13575
    },
    {
      "entropy": 6.33928861618042,
      "epoch": 1.5672244662435084,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047647272509756584,
      "loss": 6.1339,
      "mean_token_accuracy": 0.1675115406513214,
      "num_tokens": 34412274.0,
      "step": 13580
    },
    {
      "entropy": 6.297660255432129,
      "epoch": 1.567801500288517,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004764543510835269,
      "loss": 6.0849,
      "mean_token_accuracy": 0.1719527691602707,
      "num_tokens": 34425043.0,
      "step": 13585
    },
    {
      "entropy": 6.312202835083008,
      "epoch": 1.5683785343335257,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000476435970293671,
      "loss": 6.154,
      "mean_token_accuracy": 0.16889333426952363,
      "num_tokens": 34437022.0,
      "step": 13590
    },
    {
      "entropy": 6.268210458755493,
      "epoch": 1.5689555683785343,
      "grad_norm": 4.0625,
      "learning_rate": 0.00047641758272861626,
      "loss": 5.9941,
      "mean_token_accuracy": 0.17029512375593187,
      "num_tokens": 34449668.0,
      "step": 13595
    },
    {
      "entropy": 6.270573329925537,
      "epoch": 1.569532602423543,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004763991883889811,
      "loss": 6.0436,
      "mean_token_accuracy": 0.16294163167476655,
      "num_tokens": 34461588.0,
      "step": 13600
    },
    {
      "entropy": 6.220047664642334,
      "epoch": 1.5701096364685516,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004763807872753843,
      "loss": 6.0681,
      "mean_token_accuracy": 0.16750042587518693,
      "num_tokens": 34474696.0,
      "step": 13605
    },
    {
      "entropy": 6.280921459197998,
      "epoch": 1.5706866705135603,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047636237938844484,
      "loss": 6.1026,
      "mean_token_accuracy": 0.16470820903778077,
      "num_tokens": 34488477.0,
      "step": 13610
    },
    {
      "entropy": 6.263820552825928,
      "epoch": 1.571263704558569,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004763439647287817,
      "loss": 6.0493,
      "mean_token_accuracy": 0.16895479708909988,
      "num_tokens": 34502159.0,
      "step": 13615
    },
    {
      "entropy": 6.248439455032349,
      "epoch": 1.5718407386035778,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004763255432970144,
      "loss": 6.0714,
      "mean_token_accuracy": 0.1683644562959671,
      "num_tokens": 34516247.0,
      "step": 13620
    },
    {
      "entropy": 6.161786127090454,
      "epoch": 1.5724177726485862,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047630711509376235,
      "loss": 5.9466,
      "mean_token_accuracy": 0.17339649200439453,
      "num_tokens": 34528599.0,
      "step": 13625
    },
    {
      "entropy": 6.315026569366455,
      "epoch": 1.572994806693595,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004762886801196455,
      "loss": 6.1762,
      "mean_token_accuracy": 0.15327844619750977,
      "num_tokens": 34541346.0,
      "step": 13630
    },
    {
      "entropy": 6.389235925674439,
      "epoch": 1.5735718407386035,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047627023837528386,
      "loss": 6.1017,
      "mean_token_accuracy": 0.16337597817182542,
      "num_tokens": 34553934.0,
      "step": 13635
    },
    {
      "entropy": 6.329330778121948,
      "epoch": 1.5741488747836123,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047625178986129775,
      "loss": 6.1503,
      "mean_token_accuracy": 0.1708940386772156,
      "num_tokens": 34565739.0,
      "step": 13640
    },
    {
      "entropy": 6.214039516448975,
      "epoch": 1.5747259088286207,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004762333345783078,
      "loss": 6.0717,
      "mean_token_accuracy": 0.17032964378595353,
      "num_tokens": 34579062.0,
      "step": 13645
    },
    {
      "entropy": 6.2890371799469,
      "epoch": 1.5753029428736296,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047621487252693436,
      "loss": 6.109,
      "mean_token_accuracy": 0.17146503627300264,
      "num_tokens": 34592090.0,
      "step": 13650
    },
    {
      "entropy": 6.215422916412353,
      "epoch": 1.5758799769186382,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047619640370779876,
      "loss": 5.9818,
      "mean_token_accuracy": 0.16550679504871368,
      "num_tokens": 34604626.0,
      "step": 13655
    },
    {
      "entropy": 6.333647632598877,
      "epoch": 1.5764570109636469,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047617792812152207,
      "loss": 6.1318,
      "mean_token_accuracy": 0.1639261156320572,
      "num_tokens": 34617174.0,
      "step": 13660
    },
    {
      "entropy": 6.230650806427002,
      "epoch": 1.5770340450086555,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004761594457687256,
      "loss": 5.9971,
      "mean_token_accuracy": 0.17324588149785997,
      "num_tokens": 34629909.0,
      "step": 13665
    },
    {
      "entropy": 6.257308387756348,
      "epoch": 1.5776110790536642,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004761409566500313,
      "loss": 6.0872,
      "mean_token_accuracy": 0.17212423384189607,
      "num_tokens": 34643640.0,
      "step": 13670
    },
    {
      "entropy": 6.281475162506103,
      "epoch": 1.5781881130986728,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047612246076606066,
      "loss": 6.0142,
      "mean_token_accuracy": 0.1610131159424782,
      "num_tokens": 34655250.0,
      "step": 13675
    },
    {
      "entropy": 6.323982858657837,
      "epoch": 1.5787651471436814,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047610395811743594,
      "loss": 6.0816,
      "mean_token_accuracy": 0.16235461086034775,
      "num_tokens": 34669318.0,
      "step": 13680
    },
    {
      "entropy": 6.237914276123047,
      "epoch": 1.5793421811886903,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047608544870477956,
      "loss": 6.0607,
      "mean_token_accuracy": 0.1697299838066101,
      "num_tokens": 34680932.0,
      "step": 13685
    },
    {
      "entropy": 6.323851537704468,
      "epoch": 1.5799192152336987,
      "grad_norm": 1.875,
      "learning_rate": 0.00047606693252871395,
      "loss": 6.1738,
      "mean_token_accuracy": 0.16352701634168626,
      "num_tokens": 34693091.0,
      "step": 13690
    },
    {
      "entropy": 6.353163433074951,
      "epoch": 1.5804962492787076,
      "grad_norm": 1.328125,
      "learning_rate": 0.000476048409589862,
      "loss": 6.1665,
      "mean_token_accuracy": 0.16994237452745437,
      "num_tokens": 34706645.0,
      "step": 13695
    },
    {
      "entropy": 6.312082672119141,
      "epoch": 1.581073283323716,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004760298798888466,
      "loss": 6.1502,
      "mean_token_accuracy": 0.16166957914829255,
      "num_tokens": 34718974.0,
      "step": 13700
    },
    {
      "entropy": 6.258579730987549,
      "epoch": 1.5816503173687249,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004760113434262911,
      "loss": 6.1283,
      "mean_token_accuracy": 0.1639259859919548,
      "num_tokens": 34732424.0,
      "step": 13705
    },
    {
      "entropy": 6.27340145111084,
      "epoch": 1.5822273514137333,
      "grad_norm": 1.125,
      "learning_rate": 0.00047599280020281894,
      "loss": 6.0145,
      "mean_token_accuracy": 0.17224159985780715,
      "num_tokens": 34743865.0,
      "step": 13710
    },
    {
      "entropy": 6.236146402359009,
      "epoch": 1.5828043854587421,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047597425021905364,
      "loss": 5.9892,
      "mean_token_accuracy": 0.1733963370323181,
      "num_tokens": 34756415.0,
      "step": 13715
    },
    {
      "entropy": 6.32954421043396,
      "epoch": 1.5833814195037508,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004759556934756194,
      "loss": 6.1399,
      "mean_token_accuracy": 0.1598810486495495,
      "num_tokens": 34769666.0,
      "step": 13720
    },
    {
      "entropy": 6.132478141784668,
      "epoch": 1.5839584535487594,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047593712997314017,
      "loss": 5.9454,
      "mean_token_accuracy": 0.18104538768529893,
      "num_tokens": 34781523.0,
      "step": 13725
    },
    {
      "entropy": 6.255754470825195,
      "epoch": 1.584535487593768,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047591855971224035,
      "loss": 6.0319,
      "mean_token_accuracy": 0.16793281435966492,
      "num_tokens": 34794597.0,
      "step": 13730
    },
    {
      "entropy": 6.303833389282227,
      "epoch": 1.5851125216387767,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004758999826935446,
      "loss": 6.1652,
      "mean_token_accuracy": 0.1641547754406929,
      "num_tokens": 34807791.0,
      "step": 13735
    },
    {
      "entropy": 6.454448127746582,
      "epoch": 1.5856895556837853,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004758813989176778,
      "loss": 6.1679,
      "mean_token_accuracy": 0.16815381050109862,
      "num_tokens": 34821456.0,
      "step": 13740
    },
    {
      "entropy": 6.289196872711182,
      "epoch": 1.586266589728794,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047586280838526483,
      "loss": 6.0683,
      "mean_token_accuracy": 0.17699868679046632,
      "num_tokens": 34834786.0,
      "step": 13745
    },
    {
      "entropy": 6.2721411228179935,
      "epoch": 1.5868436237738026,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004758442110969312,
      "loss": 6.0719,
      "mean_token_accuracy": 0.16389156579971315,
      "num_tokens": 34846125.0,
      "step": 13750
    },
    {
      "entropy": 6.36556601524353,
      "epoch": 1.5874206578188113,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004758256070533022,
      "loss": 6.1379,
      "mean_token_accuracy": 0.1652582183480263,
      "num_tokens": 34858438.0,
      "step": 13755
    },
    {
      "entropy": 6.269256162643432,
      "epoch": 1.5879976918638201,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004758069962550037,
      "loss": 6.1417,
      "mean_token_accuracy": 0.16086930483579637,
      "num_tokens": 34871340.0,
      "step": 13760
    },
    {
      "entropy": 6.320274209976196,
      "epoch": 1.5885747259088285,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047578837870266156,
      "loss": 6.0939,
      "mean_token_accuracy": 0.16528169810771942,
      "num_tokens": 34882111.0,
      "step": 13765
    },
    {
      "entropy": 6.327654838562012,
      "epoch": 1.5891517599538374,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047576975439690206,
      "loss": 6.1363,
      "mean_token_accuracy": 0.16278501451015473,
      "num_tokens": 34895539.0,
      "step": 13770
    },
    {
      "entropy": 6.314521789550781,
      "epoch": 1.5897287939988458,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047575112333835164,
      "loss": 6.075,
      "mean_token_accuracy": 0.16338768899440764,
      "num_tokens": 34907860.0,
      "step": 13775
    },
    {
      "entropy": 6.347145318984985,
      "epoch": 1.5903058280438547,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047573248552763684,
      "loss": 6.1797,
      "mean_token_accuracy": 0.16526250541210175,
      "num_tokens": 34920595.0,
      "step": 13780
    },
    {
      "entropy": 6.345773601531983,
      "epoch": 1.590882862088863,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047571384096538474,
      "loss": 6.1018,
      "mean_token_accuracy": 0.16914000362157822,
      "num_tokens": 34933458.0,
      "step": 13785
    },
    {
      "entropy": 6.261147594451904,
      "epoch": 1.591459896133872,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004756951896522222,
      "loss": 6.0385,
      "mean_token_accuracy": 0.17352436035871505,
      "num_tokens": 34946872.0,
      "step": 13790
    },
    {
      "entropy": 6.304848003387451,
      "epoch": 1.5920369301788806,
      "grad_norm": 3.375,
      "learning_rate": 0.0004756765315887766,
      "loss": 6.1094,
      "mean_token_accuracy": 0.16546263694763183,
      "num_tokens": 34959384.0,
      "step": 13795
    },
    {
      "entropy": 6.362481641769409,
      "epoch": 1.5926139642238892,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004756578667756756,
      "loss": 6.129,
      "mean_token_accuracy": 0.16471261978149415,
      "num_tokens": 34971571.0,
      "step": 13800
    },
    {
      "entropy": 6.298099565505981,
      "epoch": 1.5931909982688979,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004756391952135469,
      "loss": 6.062,
      "mean_token_accuracy": 0.17933578938245773,
      "num_tokens": 34984897.0,
      "step": 13805
    },
    {
      "entropy": 6.322285318374634,
      "epoch": 1.5937680323139065,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047562051690301855,
      "loss": 6.138,
      "mean_token_accuracy": 0.1619904324412346,
      "num_tokens": 34998185.0,
      "step": 13810
    },
    {
      "entropy": 6.395644950866699,
      "epoch": 1.5943450663589152,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047560183184471873,
      "loss": 6.0972,
      "mean_token_accuracy": 0.1668487012386322,
      "num_tokens": 35010985.0,
      "step": 13815
    },
    {
      "entropy": 6.311239624023438,
      "epoch": 1.5949221004039238,
      "grad_norm": 1.234375,
      "learning_rate": 0.000475583140039276,
      "loss": 6.1207,
      "mean_token_accuracy": 0.16769470274448395,
      "num_tokens": 35023785.0,
      "step": 13820
    },
    {
      "entropy": 6.369430828094482,
      "epoch": 1.5954991344489324,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047556444148731897,
      "loss": 6.1613,
      "mean_token_accuracy": 0.15731503814458847,
      "num_tokens": 35035926.0,
      "step": 13825
    },
    {
      "entropy": 6.3039085388183596,
      "epoch": 1.596076168493941,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004755457361894766,
      "loss": 6.1064,
      "mean_token_accuracy": 0.16659609824419022,
      "num_tokens": 35048290.0,
      "step": 13830
    },
    {
      "entropy": 6.290177297592163,
      "epoch": 1.59665320253895,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047552702414637793,
      "loss": 6.0144,
      "mean_token_accuracy": 0.17391386926174163,
      "num_tokens": 35059883.0,
      "step": 13835
    },
    {
      "entropy": 6.334092044830323,
      "epoch": 1.5972302365839584,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047550830535865245,
      "loss": 6.1131,
      "mean_token_accuracy": 0.16483326703310014,
      "num_tokens": 35071747.0,
      "step": 13840
    },
    {
      "entropy": 6.252234888076782,
      "epoch": 1.5978072706289672,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047548957982692974,
      "loss": 6.132,
      "mean_token_accuracy": 0.16601303368806838,
      "num_tokens": 35083774.0,
      "step": 13845
    },
    {
      "entropy": 6.295589828491211,
      "epoch": 1.5983843046739756,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004754708475518396,
      "loss": 6.0616,
      "mean_token_accuracy": 0.17326804846525193,
      "num_tokens": 35095757.0,
      "step": 13850
    },
    {
      "entropy": 6.31047887802124,
      "epoch": 1.5989613387189845,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047545210853401214,
      "loss": 6.0967,
      "mean_token_accuracy": 0.16631067395210267,
      "num_tokens": 35108788.0,
      "step": 13855
    },
    {
      "entropy": 6.352183723449707,
      "epoch": 1.599538372763993,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047543336277407753,
      "loss": 6.2169,
      "mean_token_accuracy": 0.16050542891025543,
      "num_tokens": 35121641.0,
      "step": 13860
    },
    {
      "entropy": 6.352316522598267,
      "epoch": 1.6001154068090018,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047541461027266624,
      "loss": 6.1502,
      "mean_token_accuracy": 0.16059670001268386,
      "num_tokens": 35133960.0,
      "step": 13865
    },
    {
      "entropy": 6.363188743591309,
      "epoch": 1.6006924408540104,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004753958510304091,
      "loss": 6.1361,
      "mean_token_accuracy": 0.1658671870827675,
      "num_tokens": 35146027.0,
      "step": 13870
    },
    {
      "entropy": 6.350321197509766,
      "epoch": 1.601269474899019,
      "grad_norm": 1.8125,
      "learning_rate": 0.00047537708504793714,
      "loss": 6.1071,
      "mean_token_accuracy": 0.16653528958559036,
      "num_tokens": 35158728.0,
      "step": 13875
    },
    {
      "entropy": 6.2908519268035885,
      "epoch": 1.6018465089440277,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047535831232588146,
      "loss": 6.1269,
      "mean_token_accuracy": 0.16270237565040588,
      "num_tokens": 35171734.0,
      "step": 13880
    },
    {
      "entropy": 6.360332727432251,
      "epoch": 1.6024235429890363,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047533953286487345,
      "loss": 6.1369,
      "mean_token_accuracy": 0.15890816748142242,
      "num_tokens": 35185074.0,
      "step": 13885
    },
    {
      "entropy": 6.3325221061706545,
      "epoch": 1.603000577034045,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004753207466655447,
      "loss": 6.1518,
      "mean_token_accuracy": 0.16376099586486817,
      "num_tokens": 35197972.0,
      "step": 13890
    },
    {
      "entropy": 6.334444665908814,
      "epoch": 1.6035776110790536,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004753019537285273,
      "loss": 6.0818,
      "mean_token_accuracy": 0.1678009122610092,
      "num_tokens": 35210939.0,
      "step": 13895
    },
    {
      "entropy": 6.386923742294312,
      "epoch": 1.6041546451240625,
      "grad_norm": 1.25,
      "learning_rate": 0.00047528315405445296,
      "loss": 6.1734,
      "mean_token_accuracy": 0.16232287734746934,
      "num_tokens": 35223521.0,
      "step": 13900
    },
    {
      "entropy": 6.303661584854126,
      "epoch": 1.604731679169071,
      "grad_norm": 2.203125,
      "learning_rate": 0.00047526434764395435,
      "loss": 6.1578,
      "mean_token_accuracy": 0.16882762908935547,
      "num_tokens": 35235713.0,
      "step": 13905
    },
    {
      "entropy": 6.249205541610718,
      "epoch": 1.6053087132140798,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047524553449766386,
      "loss": 6.1123,
      "mean_token_accuracy": 0.16625256538391114,
      "num_tokens": 35248938.0,
      "step": 13910
    },
    {
      "entropy": 6.343287229537964,
      "epoch": 1.6058857472590882,
      "grad_norm": 1.25,
      "learning_rate": 0.00047522671461621433,
      "loss": 6.1235,
      "mean_token_accuracy": 0.16260135173797607,
      "num_tokens": 35260916.0,
      "step": 13915
    },
    {
      "entropy": 6.35877799987793,
      "epoch": 1.606462781304097,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004752078880002386,
      "loss": 6.0874,
      "mean_token_accuracy": 0.16499348878860473,
      "num_tokens": 35273529.0,
      "step": 13920
    },
    {
      "entropy": 6.319088792800903,
      "epoch": 1.6070398153491054,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047518905465037005,
      "loss": 6.0514,
      "mean_token_accuracy": 0.16407953053712845,
      "num_tokens": 35285500.0,
      "step": 13925
    },
    {
      "entropy": 6.29489312171936,
      "epoch": 1.6076168493941143,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047517021456724214,
      "loss": 6.1082,
      "mean_token_accuracy": 0.1632474958896637,
      "num_tokens": 35300039.0,
      "step": 13930
    },
    {
      "entropy": 6.309400177001953,
      "epoch": 1.608193883439123,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047515136775148843,
      "loss": 6.1973,
      "mean_token_accuracy": 0.15550429224967957,
      "num_tokens": 35313091.0,
      "step": 13935
    },
    {
      "entropy": 6.329781770706177,
      "epoch": 1.6087709174841316,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004751325142037429,
      "loss": 6.1284,
      "mean_token_accuracy": 0.16518935412168503,
      "num_tokens": 35325890.0,
      "step": 13940
    },
    {
      "entropy": 6.278371334075928,
      "epoch": 1.6093479515291402,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047511365392463974,
      "loss": 6.0337,
      "mean_token_accuracy": 0.17123775631189347,
      "num_tokens": 35339395.0,
      "step": 13945
    },
    {
      "entropy": 6.252002334594726,
      "epoch": 1.6099249855741489,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047509478691481317,
      "loss": 6.0848,
      "mean_token_accuracy": 0.16890290081501008,
      "num_tokens": 35351163.0,
      "step": 13950
    },
    {
      "entropy": 6.394436597824097,
      "epoch": 1.6105020196191575,
      "grad_norm": 2.6875,
      "learning_rate": 0.00047507591317489783,
      "loss": 6.1041,
      "mean_token_accuracy": 0.1712552562355995,
      "num_tokens": 35364099.0,
      "step": 13955
    },
    {
      "entropy": 6.3305333137512205,
      "epoch": 1.6110790536641661,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004750570327055286,
      "loss": 6.1306,
      "mean_token_accuracy": 0.1640000745654106,
      "num_tokens": 35376768.0,
      "step": 13960
    },
    {
      "entropy": 6.344961166381836,
      "epoch": 1.6116560877091748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047503814550734034,
      "loss": 6.1351,
      "mean_token_accuracy": 0.1643372043967247,
      "num_tokens": 35388892.0,
      "step": 13965
    },
    {
      "entropy": 6.371452808380127,
      "epoch": 1.6122331217541834,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004750192515809685,
      "loss": 6.1855,
      "mean_token_accuracy": 0.16151557713747025,
      "num_tokens": 35401573.0,
      "step": 13970
    },
    {
      "entropy": 6.332014751434326,
      "epoch": 1.6128101557991923,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047500035092704843,
      "loss": 6.0987,
      "mean_token_accuracy": 0.16190767139196396,
      "num_tokens": 35415073.0,
      "step": 13975
    },
    {
      "entropy": 6.248025035858154,
      "epoch": 1.6133871898442007,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004749814435462159,
      "loss": 6.0687,
      "mean_token_accuracy": 0.1728854924440384,
      "num_tokens": 35428756.0,
      "step": 13980
    },
    {
      "entropy": 6.308794832229614,
      "epoch": 1.6139642238892096,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004749625294391069,
      "loss": 6.062,
      "mean_token_accuracy": 0.16338517218828202,
      "num_tokens": 35440387.0,
      "step": 13985
    },
    {
      "entropy": 6.26749005317688,
      "epoch": 1.614541257934218,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047494360860635755,
      "loss": 5.9829,
      "mean_token_accuracy": 0.18377355188131334,
      "num_tokens": 35454105.0,
      "step": 13990
    },
    {
      "entropy": 6.304738330841064,
      "epoch": 1.6151182919792268,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004749246810486042,
      "loss": 6.1322,
      "mean_token_accuracy": 0.16615962982177734,
      "num_tokens": 35467775.0,
      "step": 13995
    },
    {
      "entropy": 6.271231174468994,
      "epoch": 1.6156953260242353,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0004749057467664836,
      "loss": 6.1092,
      "mean_token_accuracy": 0.16965964138507844,
      "num_tokens": 35479217.0,
      "step": 14000
    },
    {
      "entropy": 6.302905178070068,
      "epoch": 1.6162723600692441,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047488680576063247,
      "loss": 6.076,
      "mean_token_accuracy": 0.1761475294828415,
      "num_tokens": 35490676.0,
      "step": 14005
    },
    {
      "entropy": 6.224405097961426,
      "epoch": 1.6168493941142528,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004748678580316878,
      "loss": 6.0433,
      "mean_token_accuracy": 0.16903021633625032,
      "num_tokens": 35503440.0,
      "step": 14010
    },
    {
      "entropy": 6.247961568832397,
      "epoch": 1.6174264281592614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047484890358028714,
      "loss": 6.0403,
      "mean_token_accuracy": 0.1722070872783661,
      "num_tokens": 35515717.0,
      "step": 14015
    },
    {
      "entropy": 6.377208995819092,
      "epoch": 1.61800346220427,
      "grad_norm": 1.625,
      "learning_rate": 0.0004748299424070678,
      "loss": 6.0816,
      "mean_token_accuracy": 0.16644570976495743,
      "num_tokens": 35528445.0,
      "step": 14020
    },
    {
      "entropy": 6.203998327255249,
      "epoch": 1.6185804962492787,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004748109745126677,
      "loss": 5.976,
      "mean_token_accuracy": 0.17228845208883287,
      "num_tokens": 35541652.0,
      "step": 14025
    },
    {
      "entropy": 6.157395648956299,
      "epoch": 1.6191575302942873,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047479199989772464,
      "loss": 6.0205,
      "mean_token_accuracy": 0.1703417792916298,
      "num_tokens": 35555310.0,
      "step": 14030
    },
    {
      "entropy": 6.334373998641968,
      "epoch": 1.619734564339296,
      "grad_norm": 1.40625,
      "learning_rate": 0.000474773018562877,
      "loss": 6.0952,
      "mean_token_accuracy": 0.16538431942462922,
      "num_tokens": 35567830.0,
      "step": 14035
    },
    {
      "entropy": 6.271333789825439,
      "epoch": 1.6203115983843048,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047475403050876297,
      "loss": 6.0555,
      "mean_token_accuracy": 0.16785277426242828,
      "num_tokens": 35579054.0,
      "step": 14040
    },
    {
      "entropy": 6.363547706604004,
      "epoch": 1.6208886324293132,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004747350357360214,
      "loss": 6.1628,
      "mean_token_accuracy": 0.16527276188135148,
      "num_tokens": 35592463.0,
      "step": 14045
    },
    {
      "entropy": 6.34897894859314,
      "epoch": 1.621465666474322,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004747160342452912,
      "loss": 6.1284,
      "mean_token_accuracy": 0.1657451093196869,
      "num_tokens": 35605127.0,
      "step": 14050
    },
    {
      "entropy": 6.232157468795776,
      "epoch": 1.6220427005193305,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047469702603721134,
      "loss": 6.0409,
      "mean_token_accuracy": 0.17249712347984314,
      "num_tokens": 35617509.0,
      "step": 14055
    },
    {
      "entropy": 6.257739067077637,
      "epoch": 1.6226197345643394,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004746780111124212,
      "loss": 6.0855,
      "mean_token_accuracy": 0.16773395836353303,
      "num_tokens": 35630788.0,
      "step": 14060
    },
    {
      "entropy": 6.354380941390991,
      "epoch": 1.6231967686093478,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047465898947156033,
      "loss": 6.1451,
      "mean_token_accuracy": 0.16356271356344224,
      "num_tokens": 35643613.0,
      "step": 14065
    },
    {
      "entropy": 6.372706079483033,
      "epoch": 1.6237738026543567,
      "grad_norm": 2.53125,
      "learning_rate": 0.00047463996111526854,
      "loss": 6.1067,
      "mean_token_accuracy": 0.1694120690226555,
      "num_tokens": 35656495.0,
      "step": 14070
    },
    {
      "entropy": 6.3641197681427,
      "epoch": 1.6243508366993653,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047462092604418576,
      "loss": 6.191,
      "mean_token_accuracy": 0.1535666763782501,
      "num_tokens": 35669661.0,
      "step": 14075
    },
    {
      "entropy": 6.269794607162476,
      "epoch": 1.624927870744374,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047460188425895236,
      "loss": 6.0978,
      "mean_token_accuracy": 0.16541868150234224,
      "num_tokens": 35682675.0,
      "step": 14080
    },
    {
      "entropy": 6.331015968322754,
      "epoch": 1.6255049047893826,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047458283576020874,
      "loss": 6.1263,
      "mean_token_accuracy": 0.16866612434387207,
      "num_tokens": 35694571.0,
      "step": 14085
    },
    {
      "entropy": 6.305769634246826,
      "epoch": 1.6260819388343912,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047456378054859554,
      "loss": 6.0442,
      "mean_token_accuracy": 0.1698611244559288,
      "num_tokens": 35707196.0,
      "step": 14090
    },
    {
      "entropy": 6.2425220012664795,
      "epoch": 1.6266589728793999,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047454471862475375,
      "loss": 6.0519,
      "mean_token_accuracy": 0.17151181250810624,
      "num_tokens": 35719682.0,
      "step": 14095
    },
    {
      "entropy": 6.317946672439575,
      "epoch": 1.6272360069244085,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047452564998932446,
      "loss": 6.1473,
      "mean_token_accuracy": 0.16216070652008058,
      "num_tokens": 35732595.0,
      "step": 14100
    },
    {
      "entropy": 6.355157709121704,
      "epoch": 1.6278130409694171,
      "grad_norm": 1.40625,
      "learning_rate": 0.000474506574642949,
      "loss": 6.0666,
      "mean_token_accuracy": 0.1663676917552948,
      "num_tokens": 35744611.0,
      "step": 14105
    },
    {
      "entropy": 6.2730059146881105,
      "epoch": 1.6283900750144258,
      "grad_norm": 1.6875,
      "learning_rate": 0.000474487492586269,
      "loss": 6.0772,
      "mean_token_accuracy": 0.16661421656608583,
      "num_tokens": 35757288.0,
      "step": 14110
    },
    {
      "entropy": 6.374433422088623,
      "epoch": 1.6289671090594346,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004744684038199263,
      "loss": 6.1371,
      "mean_token_accuracy": 0.16418022513389588,
      "num_tokens": 35769747.0,
      "step": 14115
    },
    {
      "entropy": 6.3676992893219,
      "epoch": 1.629544143104443,
      "grad_norm": 2.28125,
      "learning_rate": 0.00047444930834456293,
      "loss": 6.1704,
      "mean_token_accuracy": 0.16366525143384933,
      "num_tokens": 35781656.0,
      "step": 14120
    },
    {
      "entropy": 6.329796028137207,
      "epoch": 1.630121177149452,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004744302061608212,
      "loss": 6.153,
      "mean_token_accuracy": 0.16873974055051805,
      "num_tokens": 35794618.0,
      "step": 14125
    },
    {
      "entropy": 6.369326257705689,
      "epoch": 1.6306982111944603,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00047441109726934345,
      "loss": 6.0348,
      "mean_token_accuracy": 0.16814184337854385,
      "num_tokens": 35807504.0,
      "step": 14130
    },
    {
      "entropy": 6.217661762237549,
      "epoch": 1.6312752452394692,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047439198167077256,
      "loss": 6.0602,
      "mean_token_accuracy": 0.17050414085388182,
      "num_tokens": 35819668.0,
      "step": 14135
    },
    {
      "entropy": 6.300717782974243,
      "epoch": 1.6318522792844776,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004743728593657514,
      "loss": 6.099,
      "mean_token_accuracy": 0.17155559360980988,
      "num_tokens": 35832078.0,
      "step": 14140
    },
    {
      "entropy": 6.3105512142181395,
      "epoch": 1.6324293133294865,
      "grad_norm": 1.25,
      "learning_rate": 0.0004743537303549231,
      "loss": 5.9889,
      "mean_token_accuracy": 0.17244088500738144,
      "num_tokens": 35843580.0,
      "step": 14145
    },
    {
      "entropy": 6.236100673675537,
      "epoch": 1.6330063473744951,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004743345946389311,
      "loss": 6.0669,
      "mean_token_accuracy": 0.17535771280527115,
      "num_tokens": 35857146.0,
      "step": 14150
    },
    {
      "entropy": 6.253125715255737,
      "epoch": 1.6335833814195038,
      "grad_norm": 3.171875,
      "learning_rate": 0.00047431545221841907,
      "loss": 6.0763,
      "mean_token_accuracy": 0.16937452852725982,
      "num_tokens": 35869374.0,
      "step": 14155
    },
    {
      "entropy": 6.396531057357788,
      "epoch": 1.6341604154645124,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047429630309403086,
      "loss": 6.1255,
      "mean_token_accuracy": 0.15818234384059907,
      "num_tokens": 35881882.0,
      "step": 14160
    },
    {
      "entropy": 6.185387468338012,
      "epoch": 1.634737449509521,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047427714726641044,
      "loss": 5.9409,
      "mean_token_accuracy": 0.179660464823246,
      "num_tokens": 35895512.0,
      "step": 14165
    },
    {
      "entropy": 6.18449034690857,
      "epoch": 1.6353144835545297,
      "grad_norm": 2.28125,
      "learning_rate": 0.00047425798473620215,
      "loss": 5.9598,
      "mean_token_accuracy": 0.1764688104391098,
      "num_tokens": 35908020.0,
      "step": 14170
    },
    {
      "entropy": 6.300310659408569,
      "epoch": 1.6358915175995383,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004742388155040505,
      "loss": 6.0202,
      "mean_token_accuracy": 0.17187013179063798,
      "num_tokens": 35920261.0,
      "step": 14175
    },
    {
      "entropy": 6.360107517242431,
      "epoch": 1.6364685516445472,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047421963957060026,
      "loss": 6.1752,
      "mean_token_accuracy": 0.16400312632322311,
      "num_tokens": 35932135.0,
      "step": 14180
    },
    {
      "entropy": 6.245635604858398,
      "epoch": 1.6370455856895556,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004742004569364964,
      "loss": 6.0604,
      "mean_token_accuracy": 0.16876980662345886,
      "num_tokens": 35944574.0,
      "step": 14185
    },
    {
      "entropy": 6.269022464752197,
      "epoch": 1.6376226197345645,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047418126760238416,
      "loss": 6.0579,
      "mean_token_accuracy": 0.1743795841932297,
      "num_tokens": 35956877.0,
      "step": 14190
    },
    {
      "entropy": 6.316147232055664,
      "epoch": 1.6381996537795729,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047416207156890887,
      "loss": 6.1537,
      "mean_token_accuracy": 0.16733625382184983,
      "num_tokens": 35968616.0,
      "step": 14195
    },
    {
      "entropy": 6.280966711044312,
      "epoch": 1.6387766878245817,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004741428688367164,
      "loss": 6.1277,
      "mean_token_accuracy": 0.16590285897254944,
      "num_tokens": 35982401.0,
      "step": 14200
    },
    {
      "entropy": 6.303164863586426,
      "epoch": 1.6393537218695902,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047412365940645225,
      "loss": 6.0664,
      "mean_token_accuracy": 0.17370134890079497,
      "num_tokens": 35994907.0,
      "step": 14205
    },
    {
      "entropy": 6.327745819091797,
      "epoch": 1.639930755914599,
      "grad_norm": 2.015625,
      "learning_rate": 0.00047410444327876286,
      "loss": 6.0397,
      "mean_token_accuracy": 0.16742120534181595,
      "num_tokens": 36007109.0,
      "step": 14210
    },
    {
      "entropy": 6.253001260757446,
      "epoch": 1.6405077899596077,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047408522045429426,
      "loss": 6.1394,
      "mean_token_accuracy": 0.16436311453580857,
      "num_tokens": 36018734.0,
      "step": 14215
    },
    {
      "entropy": 6.322027969360351,
      "epoch": 1.6410848240046163,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004740659909336933,
      "loss": 6.182,
      "mean_token_accuracy": 0.16763557493686676,
      "num_tokens": 36031427.0,
      "step": 14220
    },
    {
      "entropy": 6.3208245754241945,
      "epoch": 1.641661858049625,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047404675471760655,
      "loss": 6.0957,
      "mean_token_accuracy": 0.16826531291007996,
      "num_tokens": 36044726.0,
      "step": 14225
    },
    {
      "entropy": 6.2171959400177,
      "epoch": 1.6422388920946336,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0004740275118066811,
      "loss": 6.0971,
      "mean_token_accuracy": 0.16296851485967637,
      "num_tokens": 36056882.0,
      "step": 14230
    },
    {
      "entropy": 6.3228593349456785,
      "epoch": 1.6428159261396422,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00047400826220156416,
      "loss": 6.1013,
      "mean_token_accuracy": 0.162416909635067,
      "num_tokens": 36068548.0,
      "step": 14235
    },
    {
      "entropy": 6.356378126144409,
      "epoch": 1.6433929601846509,
      "grad_norm": 1.8125,
      "learning_rate": 0.00047398900590290313,
      "loss": 6.0699,
      "mean_token_accuracy": 0.16246748268604277,
      "num_tokens": 36080652.0,
      "step": 14240
    },
    {
      "entropy": 6.2382556915283205,
      "epoch": 1.6439699942296595,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047396974291134575,
      "loss": 6.0953,
      "mean_token_accuracy": 0.16950343698263168,
      "num_tokens": 36093372.0,
      "step": 14245
    },
    {
      "entropy": 6.337120532989502,
      "epoch": 1.6445470282746681,
      "grad_norm": 1.375,
      "learning_rate": 0.00047395047322754,
      "loss": 6.1044,
      "mean_token_accuracy": 0.17220360785722733,
      "num_tokens": 36106023.0,
      "step": 14250
    },
    {
      "entropy": 6.349548196792602,
      "epoch": 1.645124062319677,
      "grad_norm": 1.796875,
      "learning_rate": 0.00047393119685213374,
      "loss": 6.0564,
      "mean_token_accuracy": 0.1731447085738182,
      "num_tokens": 36119627.0,
      "step": 14255
    },
    {
      "entropy": 6.130048561096191,
      "epoch": 1.6457010963646854,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004739119137857756,
      "loss": 5.897,
      "mean_token_accuracy": 0.1846640169620514,
      "num_tokens": 36132835.0,
      "step": 14260
    },
    {
      "entropy": 6.287685298919678,
      "epoch": 1.6462781304096943,
      "grad_norm": 2.359375,
      "learning_rate": 0.00047389262402911404,
      "loss": 6.0702,
      "mean_token_accuracy": 0.16985928565263747,
      "num_tokens": 36144234.0,
      "step": 14265
    },
    {
      "entropy": 6.28140811920166,
      "epoch": 1.6468551644547027,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047387332758279784,
      "loss": 6.0292,
      "mean_token_accuracy": 0.16954858899116515,
      "num_tokens": 36158556.0,
      "step": 14270
    },
    {
      "entropy": 6.276857662200928,
      "epoch": 1.6474321984997116,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047385402444747606,
      "loss": 6.0968,
      "mean_token_accuracy": 0.1673444241285324,
      "num_tokens": 36171698.0,
      "step": 14275
    },
    {
      "entropy": 6.280260753631592,
      "epoch": 1.64800923254472,
      "grad_norm": 1.828125,
      "learning_rate": 0.00047383471462379803,
      "loss": 6.1143,
      "mean_token_accuracy": 0.16841488927602768,
      "num_tokens": 36184649.0,
      "step": 14280
    },
    {
      "entropy": 6.3320068359375,
      "epoch": 1.6485862665897288,
      "grad_norm": 1.359375,
      "learning_rate": 0.000473815398112413,
      "loss": 6.0247,
      "mean_token_accuracy": 0.17274200022220612,
      "num_tokens": 36197127.0,
      "step": 14285
    },
    {
      "entropy": 6.297778749465943,
      "epoch": 1.6491633006347375,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004737960749139708,
      "loss": 6.064,
      "mean_token_accuracy": 0.16634163707494737,
      "num_tokens": 36209471.0,
      "step": 14290
    },
    {
      "entropy": 6.3453363418579105,
      "epoch": 1.6497403346797461,
      "grad_norm": 1.25,
      "learning_rate": 0.0004737767450291215,
      "loss": 6.1189,
      "mean_token_accuracy": 0.16584311574697494,
      "num_tokens": 36221798.0,
      "step": 14295
    },
    {
      "entropy": 6.23340950012207,
      "epoch": 1.6503173687247548,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047375740845851506,
      "loss": 6.0453,
      "mean_token_accuracy": 0.16708384156227113,
      "num_tokens": 36235758.0,
      "step": 14300
    },
    {
      "entropy": 6.240923690795898,
      "epoch": 1.6508944027697634,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004737380652028019,
      "loss": 6.0534,
      "mean_token_accuracy": 0.16766662746667862,
      "num_tokens": 36248484.0,
      "step": 14305
    },
    {
      "entropy": 6.218803548812867,
      "epoch": 1.651471436814772,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047371871526263263,
      "loss": 5.9181,
      "mean_token_accuracy": 0.18093321323394776,
      "num_tokens": 36260796.0,
      "step": 14310
    },
    {
      "entropy": 6.229616022109985,
      "epoch": 1.6520484708597807,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004736993586386581,
      "loss": 6.0716,
      "mean_token_accuracy": 0.16974446326494216,
      "num_tokens": 36273617.0,
      "step": 14315
    },
    {
      "entropy": 6.202992248535156,
      "epoch": 1.6526255049047895,
      "grad_norm": 1.921875,
      "learning_rate": 0.00047367999533152934,
      "loss": 6.0228,
      "mean_token_accuracy": 0.17423105984926224,
      "num_tokens": 36286285.0,
      "step": 14320
    },
    {
      "entropy": 6.307953596115112,
      "epoch": 1.653202538949798,
      "grad_norm": 2.328125,
      "learning_rate": 0.00047366062534189756,
      "loss": 6.1595,
      "mean_token_accuracy": 0.16763816475868226,
      "num_tokens": 36298970.0,
      "step": 14325
    },
    {
      "entropy": 6.276805305480957,
      "epoch": 1.6537795729948068,
      "grad_norm": 1.375,
      "learning_rate": 0.00047364124867041446,
      "loss": 6.0462,
      "mean_token_accuracy": 0.176275734603405,
      "num_tokens": 36312684.0,
      "step": 14330
    },
    {
      "entropy": 6.253661394119263,
      "epoch": 1.6543566070398152,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047362186531773156,
      "loss": 6.1084,
      "mean_token_accuracy": 0.16896721422672273,
      "num_tokens": 36326300.0,
      "step": 14335
    },
    {
      "entropy": 6.266036891937256,
      "epoch": 1.654933641084824,
      "grad_norm": 1.625,
      "learning_rate": 0.0004736024752845008,
      "loss": 6.0062,
      "mean_token_accuracy": 0.169168621301651,
      "num_tokens": 36338234.0,
      "step": 14340
    },
    {
      "entropy": 6.282933187484741,
      "epoch": 1.6555106751298325,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004735830785713746,
      "loss": 6.1184,
      "mean_token_accuracy": 0.16418775916099548,
      "num_tokens": 36351032.0,
      "step": 14345
    },
    {
      "entropy": 6.279763555526733,
      "epoch": 1.6560877091748414,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004735636751790051,
      "loss": 5.9989,
      "mean_token_accuracy": 0.1654868558049202,
      "num_tokens": 36363267.0,
      "step": 14350
    },
    {
      "entropy": 6.2751750469207765,
      "epoch": 1.65666474321985,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000473544265108045,
      "loss": 6.0844,
      "mean_token_accuracy": 0.17139649987220765,
      "num_tokens": 36375628.0,
      "step": 14355
    },
    {
      "entropy": 6.262072467803955,
      "epoch": 1.6572417772648587,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047352484835914716,
      "loss": 6.12,
      "mean_token_accuracy": 0.16745211482048034,
      "num_tokens": 36388161.0,
      "step": 14360
    },
    {
      "entropy": 6.334037446975708,
      "epoch": 1.6578188113098673,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004735054249329647,
      "loss": 6.082,
      "mean_token_accuracy": 0.169405734539032,
      "num_tokens": 36400287.0,
      "step": 14365
    },
    {
      "entropy": 6.40230073928833,
      "epoch": 1.658395845354876,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047348599483015087,
      "loss": 6.1608,
      "mean_token_accuracy": 0.16180269569158554,
      "num_tokens": 36413896.0,
      "step": 14370
    },
    {
      "entropy": 6.28045129776001,
      "epoch": 1.6589728793998846,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047346655805135916,
      "loss": 6.0279,
      "mean_token_accuracy": 0.17025929540395737,
      "num_tokens": 36425812.0,
      "step": 14375
    },
    {
      "entropy": 6.212641954421997,
      "epoch": 1.6595499134448932,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004734471145972434,
      "loss": 6.0843,
      "mean_token_accuracy": 0.1649244725704193,
      "num_tokens": 36438977.0,
      "step": 14380
    },
    {
      "entropy": 6.310390663146973,
      "epoch": 1.6601269474899019,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047342766446845753,
      "loss": 6.0707,
      "mean_token_accuracy": 0.1669655993580818,
      "num_tokens": 36453173.0,
      "step": 14385
    },
    {
      "entropy": 6.309250164031982,
      "epoch": 1.6607039815349105,
      "grad_norm": 2.9375,
      "learning_rate": 0.0004734082076656557,
      "loss": 6.0591,
      "mean_token_accuracy": 0.1656690925359726,
      "num_tokens": 36464977.0,
      "step": 14390
    },
    {
      "entropy": 6.293370246887207,
      "epoch": 1.6612810155799194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047338874418949235,
      "loss": 6.1555,
      "mean_token_accuracy": 0.16307076811790466,
      "num_tokens": 36476951.0,
      "step": 14395
    },
    {
      "entropy": 6.3167040824890135,
      "epoch": 1.6618580496249278,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047336927404062213,
      "loss": 6.1511,
      "mean_token_accuracy": 0.16583692729473115,
      "num_tokens": 36490387.0,
      "step": 14400
    },
    {
      "entropy": 6.362817287445068,
      "epoch": 1.6624350836699366,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047334979721969995,
      "loss": 6.0889,
      "mean_token_accuracy": 0.1716933622956276,
      "num_tokens": 36501958.0,
      "step": 14405
    },
    {
      "entropy": 6.351159572601318,
      "epoch": 1.663012117714945,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004733303137273808,
      "loss": 6.182,
      "mean_token_accuracy": 0.16345127671957016,
      "num_tokens": 36515355.0,
      "step": 14410
    },
    {
      "entropy": 6.361408281326294,
      "epoch": 1.663589151759954,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047331082356432015,
      "loss": 6.19,
      "mean_token_accuracy": 0.16078285574913026,
      "num_tokens": 36528288.0,
      "step": 14415
    },
    {
      "entropy": 6.388300895690918,
      "epoch": 1.6641661858049623,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004732913267311734,
      "loss": 6.17,
      "mean_token_accuracy": 0.16457592099905013,
      "num_tokens": 36541577.0,
      "step": 14420
    },
    {
      "entropy": 6.301159620285034,
      "epoch": 1.6647432198499712,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004732718232285964,
      "loss": 6.1006,
      "mean_token_accuracy": 0.1722007229924202,
      "num_tokens": 36554802.0,
      "step": 14425
    },
    {
      "entropy": 6.259671449661255,
      "epoch": 1.6653202538949798,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047325231305724507,
      "loss": 6.0326,
      "mean_token_accuracy": 0.1640797436237335,
      "num_tokens": 36567175.0,
      "step": 14430
    },
    {
      "entropy": 6.251500034332276,
      "epoch": 1.6658972879399885,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004732327962177757,
      "loss": 6.0171,
      "mean_token_accuracy": 0.18299563825130463,
      "num_tokens": 36579391.0,
      "step": 14435
    },
    {
      "entropy": 6.341972541809082,
      "epoch": 1.6664743219849971,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004732132727108447,
      "loss": 6.1526,
      "mean_token_accuracy": 0.16472519487142562,
      "num_tokens": 36592684.0,
      "step": 14440
    },
    {
      "entropy": 6.3064422607421875,
      "epoch": 1.6670513560300058,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047319374253710874,
      "loss": 6.1486,
      "mean_token_accuracy": 0.16581295132637025,
      "num_tokens": 36604996.0,
      "step": 14445
    },
    {
      "entropy": 6.350444650650024,
      "epoch": 1.6676283900750144,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004731742056972247,
      "loss": 6.0836,
      "mean_token_accuracy": 0.16828773766756058,
      "num_tokens": 36618914.0,
      "step": 14450
    },
    {
      "entropy": 6.259874773025513,
      "epoch": 1.668205424120023,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004731546621918497,
      "loss": 6.0123,
      "mean_token_accuracy": 0.1697782725095749,
      "num_tokens": 36630527.0,
      "step": 14455
    },
    {
      "entropy": 6.252018451690674,
      "epoch": 1.668782458165032,
      "grad_norm": 3.328125,
      "learning_rate": 0.000473135112021641,
      "loss": 6.1106,
      "mean_token_accuracy": 0.16840582937002183,
      "num_tokens": 36642605.0,
      "step": 14460
    },
    {
      "entropy": 6.2921864032745365,
      "epoch": 1.6693594922100403,
      "grad_norm": 1.25,
      "learning_rate": 0.00047311555518725617,
      "loss": 6.0538,
      "mean_token_accuracy": 0.16681113839149475,
      "num_tokens": 36655075.0,
      "step": 14465
    },
    {
      "entropy": 6.240460395812988,
      "epoch": 1.6699365262550492,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047309599168935323,
      "loss": 6.0752,
      "mean_token_accuracy": 0.16931247115135192,
      "num_tokens": 36667684.0,
      "step": 14470
    },
    {
      "entropy": 6.286182641983032,
      "epoch": 1.6705135603000576,
      "grad_norm": 4.625,
      "learning_rate": 0.00047307642152858993,
      "loss": 6.0597,
      "mean_token_accuracy": 0.1699221059679985,
      "num_tokens": 36679656.0,
      "step": 14475
    },
    {
      "entropy": 6.269905281066895,
      "epoch": 1.6710905943450665,
      "grad_norm": 2.21875,
      "learning_rate": 0.00047305684470562453,
      "loss": 6.0321,
      "mean_token_accuracy": 0.1783604308962822,
      "num_tokens": 36691691.0,
      "step": 14480
    },
    {
      "entropy": 6.258505725860596,
      "epoch": 1.6716676283900749,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004730372612211156,
      "loss": 6.077,
      "mean_token_accuracy": 0.1736406207084656,
      "num_tokens": 36703615.0,
      "step": 14485
    },
    {
      "entropy": 6.27036304473877,
      "epoch": 1.6722446624350837,
      "grad_norm": 1.953125,
      "learning_rate": 0.00047301767107572174,
      "loss": 6.076,
      "mean_token_accuracy": 0.16444099098443984,
      "num_tokens": 36716299.0,
      "step": 14490
    },
    {
      "entropy": 6.333795166015625,
      "epoch": 1.6728216964800924,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004729980742701018,
      "loss": 6.0409,
      "mean_token_accuracy": 0.16901907622814177,
      "num_tokens": 36727723.0,
      "step": 14495
    },
    {
      "entropy": 6.242728137969971,
      "epoch": 1.673398730525101,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004729784708049151,
      "loss": 5.9147,
      "mean_token_accuracy": 0.17809747010469437,
      "num_tokens": 36740497.0,
      "step": 14500
    },
    {
      "entropy": 6.122395133972168,
      "epoch": 1.6739757645701097,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004729588606808209,
      "loss": 5.9688,
      "mean_token_accuracy": 0.1696045443415642,
      "num_tokens": 36752862.0,
      "step": 14505
    },
    {
      "entropy": 6.329869270324707,
      "epoch": 1.6745527986151183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047293924389847864,
      "loss": 6.0848,
      "mean_token_accuracy": 0.16898933798074722,
      "num_tokens": 36765948.0,
      "step": 14510
    },
    {
      "entropy": 6.2872460842132565,
      "epoch": 1.675129832660127,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004729196204585483,
      "loss": 6.0588,
      "mean_token_accuracy": 0.16923158019781112,
      "num_tokens": 36778768.0,
      "step": 14515
    },
    {
      "entropy": 6.289277219772339,
      "epoch": 1.6757068667051356,
      "grad_norm": 2.171875,
      "learning_rate": 0.00047289999036168983,
      "loss": 6.1205,
      "mean_token_accuracy": 0.1675689622759819,
      "num_tokens": 36790941.0,
      "step": 14520
    },
    {
      "entropy": 6.314367389678955,
      "epoch": 1.6762839007501442,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0004728803536085634,
      "loss": 6.1636,
      "mean_token_accuracy": 0.1690709784626961,
      "num_tokens": 36803984.0,
      "step": 14525
    },
    {
      "entropy": 6.267935705184937,
      "epoch": 1.6768609347951529,
      "grad_norm": 1.90625,
      "learning_rate": 0.00047286071019982974,
      "loss": 6.0147,
      "mean_token_accuracy": 0.172528937458992,
      "num_tokens": 36815079.0,
      "step": 14530
    },
    {
      "entropy": 6.230243587493897,
      "epoch": 1.6774379688401617,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047284106013614926,
      "loss": 6.0844,
      "mean_token_accuracy": 0.17123476564884185,
      "num_tokens": 36828007.0,
      "step": 14535
    },
    {
      "entropy": 6.433234691619873,
      "epoch": 1.6780150028851701,
      "grad_norm": 2.53125,
      "learning_rate": 0.000472821403418183,
      "loss": 6.2349,
      "mean_token_accuracy": 0.15828924477100373,
      "num_tokens": 36841762.0,
      "step": 14540
    },
    {
      "entropy": 6.321075105667115,
      "epoch": 1.678592036930179,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004728017400465921,
      "loss": 6.0574,
      "mean_token_accuracy": 0.16850826889276505,
      "num_tokens": 36854624.0,
      "step": 14545
    },
    {
      "entropy": 6.263060855865478,
      "epoch": 1.6791690709751874,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047278207002203796,
      "loss": 6.0273,
      "mean_token_accuracy": 0.1637829512357712,
      "num_tokens": 36866266.0,
      "step": 14550
    },
    {
      "entropy": 6.309200429916382,
      "epoch": 1.6797461050201963,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047276239334518216,
      "loss": 6.1123,
      "mean_token_accuracy": 0.16346139907836915,
      "num_tokens": 36878813.0,
      "step": 14555
    },
    {
      "entropy": 6.256145763397217,
      "epoch": 1.6803231390652047,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047274271001668646,
      "loss": 6.0709,
      "mean_token_accuracy": 0.16486488208174704,
      "num_tokens": 36891848.0,
      "step": 14560
    },
    {
      "entropy": 6.230090188980102,
      "epoch": 1.6809001731102136,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047272302003721286,
      "loss": 6.0399,
      "mean_token_accuracy": 0.17097180038690568,
      "num_tokens": 36903861.0,
      "step": 14565
    },
    {
      "entropy": 6.310818386077881,
      "epoch": 1.6814772071552222,
      "grad_norm": 1.859375,
      "learning_rate": 0.00047270332340742377,
      "loss": 6.1226,
      "mean_token_accuracy": 0.16238213181495667,
      "num_tokens": 36916849.0,
      "step": 14570
    },
    {
      "entropy": 6.360727548599243,
      "epoch": 1.6820542412002308,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047268362012798157,
      "loss": 6.1313,
      "mean_token_accuracy": 0.1568854719400406,
      "num_tokens": 36929559.0,
      "step": 14575
    },
    {
      "entropy": 6.244599103927612,
      "epoch": 1.6826312752452395,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004726639101995491,
      "loss": 5.9974,
      "mean_token_accuracy": 0.1702466532588005,
      "num_tokens": 36943958.0,
      "step": 14580
    },
    {
      "entropy": 6.22980580329895,
      "epoch": 1.6832083092902481,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047264419362278906,
      "loss": 6.0478,
      "mean_token_accuracy": 0.1754392057657242,
      "num_tokens": 36956542.0,
      "step": 14585
    },
    {
      "entropy": 6.289747095108032,
      "epoch": 1.6837853433352568,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047262447039836484,
      "loss": 6.0759,
      "mean_token_accuracy": 0.1649500235915184,
      "num_tokens": 36968872.0,
      "step": 14590
    },
    {
      "entropy": 6.285870981216431,
      "epoch": 1.6843623773802654,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047260474052693963,
      "loss": 6.0647,
      "mean_token_accuracy": 0.1678180992603302,
      "num_tokens": 36982234.0,
      "step": 14595
    },
    {
      "entropy": 6.246563816070557,
      "epoch": 1.684939411425274,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047258500400917724,
      "loss": 6.0466,
      "mean_token_accuracy": 0.16805330067873,
      "num_tokens": 36994407.0,
      "step": 14600
    },
    {
      "entropy": 6.325880527496338,
      "epoch": 1.6855164454702827,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047256526084574137,
      "loss": 6.0776,
      "mean_token_accuracy": 0.1641208603978157,
      "num_tokens": 37006147.0,
      "step": 14605
    },
    {
      "entropy": 6.266071701049805,
      "epoch": 1.6860934795152915,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047254551103729597,
      "loss": 6.0499,
      "mean_token_accuracy": 0.16772838830947875,
      "num_tokens": 37018206.0,
      "step": 14610
    },
    {
      "entropy": 6.275664186477661,
      "epoch": 1.6866705135603,
      "grad_norm": 2.328125,
      "learning_rate": 0.0004725257545845055,
      "loss": 6.1231,
      "mean_token_accuracy": 0.16369811147451402,
      "num_tokens": 37030632.0,
      "step": 14615
    },
    {
      "entropy": 6.217998313903808,
      "epoch": 1.6872475476053088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047250599148803443,
      "loss": 5.9618,
      "mean_token_accuracy": 0.1760655924677849,
      "num_tokens": 37041700.0,
      "step": 14620
    },
    {
      "entropy": 6.268389558792114,
      "epoch": 1.6878245816503172,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047248622174854746,
      "loss": 6.0853,
      "mean_token_accuracy": 0.1739245593547821,
      "num_tokens": 37054467.0,
      "step": 14625
    },
    {
      "entropy": 6.202273797988892,
      "epoch": 1.688401615695326,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004724664453667094,
      "loss": 5.9802,
      "mean_token_accuracy": 0.17177198827266693,
      "num_tokens": 37066596.0,
      "step": 14630
    },
    {
      "entropy": 6.2765673160552975,
      "epoch": 1.6889786497403345,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004724466623431857,
      "loss": 6.1071,
      "mean_token_accuracy": 0.1682032197713852,
      "num_tokens": 37078812.0,
      "step": 14635
    },
    {
      "entropy": 6.285689496994019,
      "epoch": 1.6895556837853434,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004724268726786415,
      "loss": 6.1031,
      "mean_token_accuracy": 0.16566502600908278,
      "num_tokens": 37090916.0,
      "step": 14640
    },
    {
      "entropy": 6.204647207260132,
      "epoch": 1.690132717830352,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004724070763737424,
      "loss": 5.9859,
      "mean_token_accuracy": 0.1659800134599209,
      "num_tokens": 37104687.0,
      "step": 14645
    },
    {
      "entropy": 6.312428569793701,
      "epoch": 1.6907097518753607,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004723872734291545,
      "loss": 6.1253,
      "mean_token_accuracy": 0.16727470308542253,
      "num_tokens": 37117763.0,
      "step": 14650
    },
    {
      "entropy": 6.279324579238891,
      "epoch": 1.6912867859203693,
      "grad_norm": 2.0,
      "learning_rate": 0.00047236746384554364,
      "loss": 6.0441,
      "mean_token_accuracy": 0.1683202251791954,
      "num_tokens": 37130355.0,
      "step": 14655
    },
    {
      "entropy": 6.299044466018676,
      "epoch": 1.691863819965378,
      "grad_norm": 3.609375,
      "learning_rate": 0.0004723476476235762,
      "loss": 6.0565,
      "mean_token_accuracy": 0.17116303145885467,
      "num_tokens": 37142116.0,
      "step": 14660
    },
    {
      "entropy": 6.232453632354736,
      "epoch": 1.6924408540103866,
      "grad_norm": 2.125,
      "learning_rate": 0.0004723278247639186,
      "loss": 6.023,
      "mean_token_accuracy": 0.1750013381242752,
      "num_tokens": 37153841.0,
      "step": 14665
    },
    {
      "entropy": 6.205951595306397,
      "epoch": 1.6930178880553952,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004723079952672377,
      "loss": 6.0276,
      "mean_token_accuracy": 0.1719754680991173,
      "num_tokens": 37165119.0,
      "step": 14670
    },
    {
      "entropy": 6.308799934387207,
      "epoch": 1.693594922100404,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047228815913420035,
      "loss": 6.1138,
      "mean_token_accuracy": 0.16897120028734208,
      "num_tokens": 37178188.0,
      "step": 14675
    },
    {
      "entropy": 6.232822418212891,
      "epoch": 1.6941719561454125,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004722683163654738,
      "loss": 5.9946,
      "mean_token_accuracy": 0.17685008347034453,
      "num_tokens": 37190211.0,
      "step": 14680
    },
    {
      "entropy": 6.244020748138428,
      "epoch": 1.6947489901904214,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004722484669617254,
      "loss": 5.9841,
      "mean_token_accuracy": 0.17107994109392166,
      "num_tokens": 37202408.0,
      "step": 14685
    },
    {
      "entropy": 6.241586112976075,
      "epoch": 1.6953260242354298,
      "grad_norm": 1.734375,
      "learning_rate": 0.00047222861092362277,
      "loss": 6.097,
      "mean_token_accuracy": 0.16470692306756973,
      "num_tokens": 37214393.0,
      "step": 14690
    },
    {
      "entropy": 6.283365488052368,
      "epoch": 1.6959030582804386,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047220874825183387,
      "loss": 6.012,
      "mean_token_accuracy": 0.16655275076627732,
      "num_tokens": 37227212.0,
      "step": 14695
    },
    {
      "entropy": 6.161387777328491,
      "epoch": 1.696480092325447,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047218887894702656,
      "loss": 6.0363,
      "mean_token_accuracy": 0.1703656107187271,
      "num_tokens": 37239742.0,
      "step": 14700
    },
    {
      "entropy": 6.316285276412964,
      "epoch": 1.697057126370456,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004721690030098693,
      "loss": 6.1202,
      "mean_token_accuracy": 0.16381090879440308,
      "num_tokens": 37254037.0,
      "step": 14705
    },
    {
      "entropy": 6.318480205535889,
      "epoch": 1.6976341604154646,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004721491204410305,
      "loss": 6.0954,
      "mean_token_accuracy": 0.16951050013303756,
      "num_tokens": 37265221.0,
      "step": 14710
    },
    {
      "entropy": 6.3138233661651615,
      "epoch": 1.6982111944604732,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00047212923124117915,
      "loss": 6.0778,
      "mean_token_accuracy": 0.16995845586061478,
      "num_tokens": 37277992.0,
      "step": 14715
    },
    {
      "entropy": 6.260839462280273,
      "epoch": 1.6987882285054818,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004721093354109839,
      "loss": 6.0369,
      "mean_token_accuracy": 0.1694336861371994,
      "num_tokens": 37291608.0,
      "step": 14720
    },
    {
      "entropy": 6.2488401412963865,
      "epoch": 1.6993652625504905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047208943295111406,
      "loss": 6.0675,
      "mean_token_accuracy": 0.16663674563169478,
      "num_tokens": 37303619.0,
      "step": 14725
    },
    {
      "entropy": 6.286404466629028,
      "epoch": 1.6999422965954991,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047206952386223905,
      "loss": 6.1235,
      "mean_token_accuracy": 0.1630891889333725,
      "num_tokens": 37316215.0,
      "step": 14730
    },
    {
      "entropy": 6.264311647415161,
      "epoch": 1.7005193306405078,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004720496081450285,
      "loss": 6.0987,
      "mean_token_accuracy": 0.17511145025491714,
      "num_tokens": 37328643.0,
      "step": 14735
    },
    {
      "entropy": 6.204177665710449,
      "epoch": 1.7010963646855164,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047202968580015224,
      "loss": 5.9841,
      "mean_token_accuracy": 0.17457183748483657,
      "num_tokens": 37340792.0,
      "step": 14740
    },
    {
      "entropy": 6.296930885314941,
      "epoch": 1.701673398730525,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047200975682828045,
      "loss": 6.0496,
      "mean_token_accuracy": 0.16684187352657318,
      "num_tokens": 37353866.0,
      "step": 14745
    },
    {
      "entropy": 6.251866245269776,
      "epoch": 1.702250432775534,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004719898212300832,
      "loss": 6.0468,
      "mean_token_accuracy": 0.16618564426898957,
      "num_tokens": 37365403.0,
      "step": 14750
    },
    {
      "entropy": 6.303374767303467,
      "epoch": 1.7028274668205423,
      "grad_norm": 1.84375,
      "learning_rate": 0.00047196987900623134,
      "loss": 6.0669,
      "mean_token_accuracy": 0.16828102320432664,
      "num_tokens": 37379487.0,
      "step": 14755
    },
    {
      "entropy": 6.296183347702026,
      "epoch": 1.7034045008655512,
      "grad_norm": 2.34375,
      "learning_rate": 0.00047194993015739527,
      "loss": 6.0298,
      "mean_token_accuracy": 0.16699840426445006,
      "num_tokens": 37392497.0,
      "step": 14760
    },
    {
      "entropy": 6.2065503120422365,
      "epoch": 1.7039815349105596,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047192997468424624,
      "loss": 5.9886,
      "mean_token_accuracy": 0.16989068686962128,
      "num_tokens": 37407126.0,
      "step": 14765
    },
    {
      "entropy": 6.148230314254761,
      "epoch": 1.7045585689555685,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004719100125874553,
      "loss": 6.0219,
      "mean_token_accuracy": 0.17883527129888535,
      "num_tokens": 37419768.0,
      "step": 14770
    },
    {
      "entropy": 6.247872018814087,
      "epoch": 1.7051356030005769,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004718900438676939,
      "loss": 6.0585,
      "mean_token_accuracy": 0.1680767372250557,
      "num_tokens": 37432512.0,
      "step": 14775
    },
    {
      "entropy": 6.35809907913208,
      "epoch": 1.7057126370455857,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004718700685256337,
      "loss": 6.1538,
      "mean_token_accuracy": 0.1664692386984825,
      "num_tokens": 37445006.0,
      "step": 14780
    },
    {
      "entropy": 6.297749042510986,
      "epoch": 1.7062896710905944,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004718500865619465,
      "loss": 6.1331,
      "mean_token_accuracy": 0.16584955304861068,
      "num_tokens": 37457386.0,
      "step": 14785
    },
    {
      "entropy": 6.260836982727051,
      "epoch": 1.706866705135603,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004718300979773044,
      "loss": 6.0229,
      "mean_token_accuracy": 0.1740827426314354,
      "num_tokens": 37469773.0,
      "step": 14790
    },
    {
      "entropy": 6.3344052791595455,
      "epoch": 1.7074437391806117,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047181010277237977,
      "loss": 6.0796,
      "mean_token_accuracy": 0.1703534796833992,
      "num_tokens": 37483301.0,
      "step": 14795
    },
    {
      "entropy": 6.317808389663696,
      "epoch": 1.7080207732256203,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004717901009478451,
      "loss": 6.0808,
      "mean_token_accuracy": 0.1657814547419548,
      "num_tokens": 37495110.0,
      "step": 14800
    },
    {
      "entropy": 6.229995775222778,
      "epoch": 1.708597807270629,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047177009250437313,
      "loss": 6.0358,
      "mean_token_accuracy": 0.17732647508382798,
      "num_tokens": 37509281.0,
      "step": 14805
    },
    {
      "entropy": 6.24862265586853,
      "epoch": 1.7091748413156376,
      "grad_norm": 1.25,
      "learning_rate": 0.00047175007744263683,
      "loss": 6.0537,
      "mean_token_accuracy": 0.1722698614001274,
      "num_tokens": 37522690.0,
      "step": 14810
    },
    {
      "entropy": 6.26502947807312,
      "epoch": 1.7097518753606464,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047173005576330935,
      "loss": 6.0587,
      "mean_token_accuracy": 0.1729898691177368,
      "num_tokens": 37534737.0,
      "step": 14815
    },
    {
      "entropy": 6.346290874481201,
      "epoch": 1.7103289094056549,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047171002746706424,
      "loss": 6.119,
      "mean_token_accuracy": 0.16451414823532104,
      "num_tokens": 37547486.0,
      "step": 14820
    },
    {
      "entropy": 6.410305833816528,
      "epoch": 1.7109059434506637,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047168999255457506,
      "loss": 6.1556,
      "mean_token_accuracy": 0.16630878448486328,
      "num_tokens": 37560083.0,
      "step": 14825
    },
    {
      "entropy": 6.255903625488282,
      "epoch": 1.7114829774956721,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047166995102651565,
      "loss": 6.0414,
      "mean_token_accuracy": 0.17067583948373793,
      "num_tokens": 37573260.0,
      "step": 14830
    },
    {
      "entropy": 6.331271076202393,
      "epoch": 1.712060011540681,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004716499028835601,
      "loss": 6.1287,
      "mean_token_accuracy": 0.16774001866579055,
      "num_tokens": 37586403.0,
      "step": 14835
    },
    {
      "entropy": 6.257412528991699,
      "epoch": 1.7126370455856894,
      "grad_norm": 1.84375,
      "learning_rate": 0.0004716298481263828,
      "loss": 6.0818,
      "mean_token_accuracy": 0.17623667418956757,
      "num_tokens": 37600771.0,
      "step": 14840
    },
    {
      "entropy": 6.320488119125367,
      "epoch": 1.7132140796306983,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004716097867556583,
      "loss": 6.0556,
      "mean_token_accuracy": 0.16797322779893875,
      "num_tokens": 37613623.0,
      "step": 14845
    },
    {
      "entropy": 6.303245306015015,
      "epoch": 1.713791113675707,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047158971877206114,
      "loss": 5.9841,
      "mean_token_accuracy": 0.17387645542621613,
      "num_tokens": 37624539.0,
      "step": 14850
    },
    {
      "entropy": 6.279318332672119,
      "epoch": 1.7143681477207156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004715696441762665,
      "loss": 6.0715,
      "mean_token_accuracy": 0.16603559702634813,
      "num_tokens": 37636556.0,
      "step": 14855
    },
    {
      "entropy": 6.298802757263184,
      "epoch": 1.7149451817657242,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047154956296894954,
      "loss": 6.0558,
      "mean_token_accuracy": 0.168888820707798,
      "num_tokens": 37648369.0,
      "step": 14860
    },
    {
      "entropy": 6.315916061401367,
      "epoch": 1.7155222158107328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004715294751507856,
      "loss": 6.1196,
      "mean_token_accuracy": 0.1606738820672035,
      "num_tokens": 37661118.0,
      "step": 14865
    },
    {
      "entropy": 6.247011470794678,
      "epoch": 1.7160992498557415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004715093807224505,
      "loss": 6.0313,
      "mean_token_accuracy": 0.16982904225587844,
      "num_tokens": 37673915.0,
      "step": 14870
    },
    {
      "entropy": 6.208894538879394,
      "epoch": 1.71667628390075,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047148927968462,
      "loss": 6.0498,
      "mean_token_accuracy": 0.1695188343524933,
      "num_tokens": 37687359.0,
      "step": 14875
    },
    {
      "entropy": 6.257250356674194,
      "epoch": 1.7172533179457588,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047146917203797,
      "loss": 6.0668,
      "mean_token_accuracy": 0.16386288702487944,
      "num_tokens": 37700111.0,
      "step": 14880
    },
    {
      "entropy": 6.407634592056274,
      "epoch": 1.7178303519907674,
      "grad_norm": 1.25,
      "learning_rate": 0.0004714490577831771,
      "loss": 6.2126,
      "mean_token_accuracy": 0.1564957782626152,
      "num_tokens": 37712707.0,
      "step": 14885
    },
    {
      "entropy": 6.378075122833252,
      "epoch": 1.7184073860357763,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004714289369209177,
      "loss": 6.125,
      "mean_token_accuracy": 0.1596768781542778,
      "num_tokens": 37724242.0,
      "step": 14890
    },
    {
      "entropy": 6.327669858932495,
      "epoch": 1.7189844200807847,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047140880945186863,
      "loss": 6.0808,
      "mean_token_accuracy": 0.16967980414628983,
      "num_tokens": 37737451.0,
      "step": 14895
    },
    {
      "entropy": 6.327920198440552,
      "epoch": 1.7195614541257935,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047138867537670676,
      "loss": 6.1087,
      "mean_token_accuracy": 0.16197078824043273,
      "num_tokens": 37749968.0,
      "step": 14900
    },
    {
      "entropy": 6.268460416793824,
      "epoch": 1.720138488170802,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047136853469610933,
      "loss": 5.9986,
      "mean_token_accuracy": 0.17554560899734498,
      "num_tokens": 37761961.0,
      "step": 14905
    },
    {
      "entropy": 6.233225727081299,
      "epoch": 1.7207155222158108,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047134838741075383,
      "loss": 6.064,
      "mean_token_accuracy": 0.17252259254455565,
      "num_tokens": 37775089.0,
      "step": 14910
    },
    {
      "entropy": 6.273133707046509,
      "epoch": 1.7212925562608192,
      "grad_norm": 3.875,
      "learning_rate": 0.00047132823352131787,
      "loss": 6.1222,
      "mean_token_accuracy": 0.1664955273270607,
      "num_tokens": 37787072.0,
      "step": 14915
    },
    {
      "entropy": 6.251991224288941,
      "epoch": 1.721869590305828,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004713080730284793,
      "loss": 6.0785,
      "mean_token_accuracy": 0.1657878965139389,
      "num_tokens": 37799162.0,
      "step": 14920
    },
    {
      "entropy": 6.262150382995605,
      "epoch": 1.7224466243508367,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047128790593291617,
      "loss": 6.079,
      "mean_token_accuracy": 0.17039323300123216,
      "num_tokens": 37811560.0,
      "step": 14925
    },
    {
      "entropy": 6.338023710250854,
      "epoch": 1.7230236583958454,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047126773223530677,
      "loss": 6.0091,
      "mean_token_accuracy": 0.1752539336681366,
      "num_tokens": 37823832.0,
      "step": 14930
    },
    {
      "entropy": 6.280074548721314,
      "epoch": 1.723600692440854,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00047124755193632975,
      "loss": 6.099,
      "mean_token_accuracy": 0.1672136142849922,
      "num_tokens": 37836718.0,
      "step": 14935
    },
    {
      "entropy": 6.227765607833862,
      "epoch": 1.7241777264858626,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047122736503666377,
      "loss": 6.0278,
      "mean_token_accuracy": 0.16516398787498474,
      "num_tokens": 37850001.0,
      "step": 14940
    },
    {
      "entropy": 6.167520570755005,
      "epoch": 1.7247547605308713,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004712071715369878,
      "loss": 5.9453,
      "mean_token_accuracy": 0.1785365030169487,
      "num_tokens": 37862425.0,
      "step": 14945
    },
    {
      "entropy": 6.249232196807862,
      "epoch": 1.72533179457588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711869714379812,
      "loss": 5.9925,
      "mean_token_accuracy": 0.17359684109687806,
      "num_tokens": 37874429.0,
      "step": 14950
    },
    {
      "entropy": 6.233495426177979,
      "epoch": 1.7259088286208888,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004711667647403232,
      "loss": 6.0376,
      "mean_token_accuracy": 0.17523322850465775,
      "num_tokens": 37888317.0,
      "step": 14955
    },
    {
      "entropy": 6.216939306259155,
      "epoch": 1.7264858626658972,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047114655144469354,
      "loss": 6.0329,
      "mean_token_accuracy": 0.17252569198608397,
      "num_tokens": 37901101.0,
      "step": 14960
    },
    {
      "entropy": 6.268746280670166,
      "epoch": 1.727062896710906,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047112633155177203,
      "loss": 6.0901,
      "mean_token_accuracy": 0.16735256612300872,
      "num_tokens": 37914013.0,
      "step": 14965
    },
    {
      "entropy": 6.229385900497436,
      "epoch": 1.7276399307559145,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004711061050622388,
      "loss": 5.9856,
      "mean_token_accuracy": 0.17679166793823242,
      "num_tokens": 37926799.0,
      "step": 14970
    },
    {
      "entropy": 6.24853138923645,
      "epoch": 1.7282169648009233,
      "grad_norm": 2.421875,
      "learning_rate": 0.00047108587197677404,
      "loss": 6.0088,
      "mean_token_accuracy": 0.17115989178419114,
      "num_tokens": 37939173.0,
      "step": 14975
    },
    {
      "entropy": 6.348738718032837,
      "epoch": 1.7287939988459318,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047106563229605845,
      "loss": 6.1661,
      "mean_token_accuracy": 0.16205629259347915,
      "num_tokens": 37951030.0,
      "step": 14980
    },
    {
      "entropy": 6.321481847763062,
      "epoch": 1.7293710328909406,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047104538602077276,
      "loss": 6.1005,
      "mean_token_accuracy": 0.16491590738296508,
      "num_tokens": 37964005.0,
      "step": 14985
    },
    {
      "entropy": 6.220985364913941,
      "epoch": 1.7299480669359493,
      "grad_norm": 2.0,
      "learning_rate": 0.0004710251331515978,
      "loss": 6.0606,
      "mean_token_accuracy": 0.170199716091156,
      "num_tokens": 37975851.0,
      "step": 14990
    },
    {
      "entropy": 6.298024463653564,
      "epoch": 1.730525100980958,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047100487368921485,
      "loss": 6.0444,
      "mean_token_accuracy": 0.16759447902441024,
      "num_tokens": 37988008.0,
      "step": 14995
    },
    {
      "entropy": 6.294814109802246,
      "epoch": 1.7311021350259665,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004709846076343055,
      "loss": 6.1433,
      "mean_token_accuracy": 0.16583738923072816,
      "num_tokens": 38000915.0,
      "step": 15000
    },
    {
      "epoch": 1.7311021350259665,
      "eval_entropy": 6.123954276588447,
      "eval_loss": 6.1699981689453125,
      "eval_mean_token_accuracy": 0.16904069022094115,
      "eval_num_tokens": 38000915.0,
      "eval_runtime": 22.6623,
      "eval_samples_per_second": 1241.532,
      "eval_steps_per_second": 155.191,
      "step": 15000
    },
    {
      "entropy": 6.24792218208313,
      "epoch": 1.7316791690709752,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047096433498755103,
      "loss": 6.0619,
      "mean_token_accuracy": 0.17191755622625352,
      "num_tokens": 38013109.0,
      "step": 15005
    },
    {
      "entropy": 6.256828117370605,
      "epoch": 1.7322562031159838,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047094405574963364,
      "loss": 6.0626,
      "mean_token_accuracy": 0.17066695243120195,
      "num_tokens": 38025969.0,
      "step": 15010
    },
    {
      "entropy": 6.28236403465271,
      "epoch": 1.7328332371609925,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047092376992123516,
      "loss": 6.048,
      "mean_token_accuracy": 0.16960776448249817,
      "num_tokens": 38039104.0,
      "step": 15015
    },
    {
      "entropy": 6.308421182632446,
      "epoch": 1.733410271206001,
      "grad_norm": 1.71875,
      "learning_rate": 0.00047090347750303803,
      "loss": 6.1151,
      "mean_token_accuracy": 0.17100338786840438,
      "num_tokens": 38053453.0,
      "step": 15020
    },
    {
      "entropy": 6.361363124847412,
      "epoch": 1.7339873052510097,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004708831784957247,
      "loss": 6.1232,
      "mean_token_accuracy": 0.15991011261940002,
      "num_tokens": 38064932.0,
      "step": 15025
    },
    {
      "entropy": 6.315028858184815,
      "epoch": 1.7345643392960186,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004708628728999781,
      "loss": 6.0741,
      "mean_token_accuracy": 0.1607021138072014,
      "num_tokens": 38078904.0,
      "step": 15030
    },
    {
      "entropy": 6.308137226104736,
      "epoch": 1.735141373341027,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004708425607164809,
      "loss": 6.1239,
      "mean_token_accuracy": 0.16347237974405288,
      "num_tokens": 38092605.0,
      "step": 15035
    },
    {
      "entropy": 6.280299425125122,
      "epoch": 1.7357184073860359,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004708222419459165,
      "loss": 6.0219,
      "mean_token_accuracy": 0.1774211421608925,
      "num_tokens": 38105953.0,
      "step": 15040
    },
    {
      "entropy": 6.285515403747558,
      "epoch": 1.7362954414310443,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004708019165889683,
      "loss": 6.1891,
      "mean_token_accuracy": 0.16232980489730836,
      "num_tokens": 38118485.0,
      "step": 15045
    },
    {
      "entropy": 6.34949746131897,
      "epoch": 1.7368724754760532,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004707815846463199,
      "loss": 6.1597,
      "mean_token_accuracy": 0.16315972208976745,
      "num_tokens": 38131220.0,
      "step": 15050
    },
    {
      "entropy": 6.239033269882202,
      "epoch": 1.7374495095210616,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004707612461186552,
      "loss": 6.031,
      "mean_token_accuracy": 0.17054777294397355,
      "num_tokens": 38143012.0,
      "step": 15055
    },
    {
      "entropy": 6.256315422058106,
      "epoch": 1.7380265435660704,
      "grad_norm": 1.984375,
      "learning_rate": 0.00047074090100665805,
      "loss": 6.0141,
      "mean_token_accuracy": 0.17277972102165223,
      "num_tokens": 38155810.0,
      "step": 15060
    },
    {
      "entropy": 6.285384273529052,
      "epoch": 1.738603577611079,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047072054931101306,
      "loss": 6.0988,
      "mean_token_accuracy": 0.16316191703081132,
      "num_tokens": 38169183.0,
      "step": 15065
    },
    {
      "entropy": 6.235709571838379,
      "epoch": 1.7391806116560877,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004707001910324046,
      "loss": 6.0805,
      "mean_token_accuracy": 0.1654953509569168,
      "num_tokens": 38182098.0,
      "step": 15070
    },
    {
      "entropy": 6.342658376693725,
      "epoch": 1.7397576457010964,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047067982617151737,
      "loss": 6.0952,
      "mean_token_accuracy": 0.16227303296327591,
      "num_tokens": 38194950.0,
      "step": 15075
    },
    {
      "entropy": 6.27519679069519,
      "epoch": 1.740334679746105,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004706594547290365,
      "loss": 6.0493,
      "mean_token_accuracy": 0.17537856101989746,
      "num_tokens": 38207600.0,
      "step": 15080
    },
    {
      "entropy": 6.234506750106812,
      "epoch": 1.7409117137911136,
      "grad_norm": 2.03125,
      "learning_rate": 0.00047063907670564695,
      "loss": 5.9848,
      "mean_token_accuracy": 0.18101305812597274,
      "num_tokens": 38219489.0,
      "step": 15085
    },
    {
      "entropy": 6.29203085899353,
      "epoch": 1.7414887478361223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004706186921020342,
      "loss": 6.1122,
      "mean_token_accuracy": 0.16346038281917571,
      "num_tokens": 38233240.0,
      "step": 15090
    },
    {
      "entropy": 6.314747762680054,
      "epoch": 1.7420657818811311,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047059830091888407,
      "loss": 6.059,
      "mean_token_accuracy": 0.16740975230932237,
      "num_tokens": 38245967.0,
      "step": 15095
    },
    {
      "entropy": 6.268871784210205,
      "epoch": 1.7426428159261396,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004705779031568821,
      "loss": 6.0446,
      "mean_token_accuracy": 0.16653717309236526,
      "num_tokens": 38258586.0,
      "step": 15100
    },
    {
      "entropy": 6.25105881690979,
      "epoch": 1.7432198499711484,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047055749881671463,
      "loss": 6.0664,
      "mean_token_accuracy": 0.17233491986989974,
      "num_tokens": 38270007.0,
      "step": 15105
    },
    {
      "entropy": 6.320896196365356,
      "epoch": 1.7437968840161568,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004705370878990677,
      "loss": 6.0114,
      "mean_token_accuracy": 0.17320594638586045,
      "num_tokens": 38282580.0,
      "step": 15110
    },
    {
      "entropy": 6.226344156265259,
      "epoch": 1.7443739180611657,
      "grad_norm": 2.03125,
      "learning_rate": 0.00047051667040462806,
      "loss": 6.0185,
      "mean_token_accuracy": 0.16522823870182038,
      "num_tokens": 38295577.0,
      "step": 15115
    },
    {
      "entropy": 6.269693994522095,
      "epoch": 1.7449509521061741,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047049624633408224,
      "loss": 6.1309,
      "mean_token_accuracy": 0.16538017094135285,
      "num_tokens": 38307924.0,
      "step": 15120
    },
    {
      "entropy": 6.359875202178955,
      "epoch": 1.745527986151183,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047047581568811724,
      "loss": 6.0743,
      "mean_token_accuracy": 0.16466730535030366,
      "num_tokens": 38321573.0,
      "step": 15125
    },
    {
      "entropy": 6.304333209991455,
      "epoch": 1.7461050201961916,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047045537846742043,
      "loss": 6.0607,
      "mean_token_accuracy": 0.17050682455301286,
      "num_tokens": 38333953.0,
      "step": 15130
    },
    {
      "entropy": 6.261137628555298,
      "epoch": 1.7466820542412003,
      "grad_norm": 1.296875,
      "learning_rate": 0.000470434934672679,
      "loss": 6.0718,
      "mean_token_accuracy": 0.16425079554319383,
      "num_tokens": 38346028.0,
      "step": 15135
    },
    {
      "entropy": 6.317224311828613,
      "epoch": 1.747259088286209,
      "grad_norm": 1.5,
      "learning_rate": 0.00047041448430458054,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17166270464658737,
      "num_tokens": 38358806.0,
      "step": 15140
    },
    {
      "entropy": 6.222446775436401,
      "epoch": 1.7478361223312175,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047039402736381305,
      "loss": 5.9498,
      "mean_token_accuracy": 0.1746882528066635,
      "num_tokens": 38370396.0,
      "step": 15145
    },
    {
      "entropy": 6.24567985534668,
      "epoch": 1.7484131563762262,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004703735638510645,
      "loss": 6.172,
      "mean_token_accuracy": 0.16677465289831161,
      "num_tokens": 38383747.0,
      "step": 15150
    },
    {
      "entropy": 6.333199834823608,
      "epoch": 1.7489901904212348,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004703530937670232,
      "loss": 6.0572,
      "mean_token_accuracy": 0.1690805047750473,
      "num_tokens": 38396611.0,
      "step": 15155
    },
    {
      "entropy": 6.281210470199585,
      "epoch": 1.7495672244662435,
      "grad_norm": 1.375,
      "learning_rate": 0.0004703326171123776,
      "loss": 6.094,
      "mean_token_accuracy": 0.16737785041332245,
      "num_tokens": 38407627.0,
      "step": 15160
    },
    {
      "entropy": 6.160182142257691,
      "epoch": 1.750144258511252,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004703121338878164,
      "loss": 5.9978,
      "mean_token_accuracy": 0.17774373441934585,
      "num_tokens": 38420522.0,
      "step": 15165
    },
    {
      "entropy": 6.282178735733032,
      "epoch": 1.750721292556261,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004702916440940286,
      "loss": 6.1264,
      "mean_token_accuracy": 0.16249785125255584,
      "num_tokens": 38433968.0,
      "step": 15170
    },
    {
      "entropy": 6.312160015106201,
      "epoch": 1.7512983266012694,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004702711477317034,
      "loss": 6.0757,
      "mean_token_accuracy": 0.17204999625682832,
      "num_tokens": 38445819.0,
      "step": 15175
    },
    {
      "entropy": 6.340288782119751,
      "epoch": 1.7518753606462782,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004702506448015301,
      "loss": 6.0909,
      "mean_token_accuracy": 0.16886577010154724,
      "num_tokens": 38458823.0,
      "step": 15180
    },
    {
      "entropy": 6.2418297290802,
      "epoch": 1.7524523946912867,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00047023013530419843,
      "loss": 6.0596,
      "mean_token_accuracy": 0.16895245462656022,
      "num_tokens": 38472553.0,
      "step": 15185
    },
    {
      "entropy": 6.294234561920166,
      "epoch": 1.7530294287362955,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004702096192403981,
      "loss": 6.0384,
      "mean_token_accuracy": 0.1733073890209198,
      "num_tokens": 38484143.0,
      "step": 15190
    },
    {
      "entropy": 6.221842432022095,
      "epoch": 1.753606462781304,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004701890966108192,
      "loss": 6.0565,
      "mean_token_accuracy": 0.16363780051469803,
      "num_tokens": 38496795.0,
      "step": 15195
    },
    {
      "entropy": 6.2286309719085695,
      "epoch": 1.7541834968263128,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000470168567416152,
      "loss": 5.9689,
      "mean_token_accuracy": 0.17548104077577592,
      "num_tokens": 38509548.0,
      "step": 15200
    },
    {
      "entropy": 6.299403381347656,
      "epoch": 1.7547605308713214,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004701480316570869,
      "loss": 6.0866,
      "mean_token_accuracy": 0.16311995685100555,
      "num_tokens": 38524011.0,
      "step": 15205
    },
    {
      "entropy": 6.3120521068572994,
      "epoch": 1.75533756491633,
      "grad_norm": 1.25,
      "learning_rate": 0.0004701274893343147,
      "loss": 6.1045,
      "mean_token_accuracy": 0.1634999006986618,
      "num_tokens": 38536913.0,
      "step": 15210
    },
    {
      "entropy": 6.233125972747803,
      "epoch": 1.7559145989613387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047010694044852643,
      "loss": 6.068,
      "mean_token_accuracy": 0.17134792655706405,
      "num_tokens": 38548759.0,
      "step": 15215
    },
    {
      "entropy": 6.314376878738403,
      "epoch": 1.7564916330063474,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000470086385000413,
      "loss": 6.0631,
      "mean_token_accuracy": 0.17076184898614882,
      "num_tokens": 38562071.0,
      "step": 15220
    },
    {
      "entropy": 6.2503444194793705,
      "epoch": 1.757068667051356,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004700658229906661,
      "loss": 6.0194,
      "mean_token_accuracy": 0.1687441125512123,
      "num_tokens": 38573816.0,
      "step": 15225
    },
    {
      "entropy": 6.396673727035522,
      "epoch": 1.7576457010963646,
      "grad_norm": 2.265625,
      "learning_rate": 0.00047004525441997694,
      "loss": 6.1596,
      "mean_token_accuracy": 0.16280039995908738,
      "num_tokens": 38586625.0,
      "step": 15230
    },
    {
      "entropy": 6.261827564239502,
      "epoch": 1.7582227351413735,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004700246792890376,
      "loss": 6.097,
      "mean_token_accuracy": 0.17354324609041213,
      "num_tokens": 38599391.0,
      "step": 15235
    },
    {
      "entropy": 6.203934097290039,
      "epoch": 1.758799769186382,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004700040975985401,
      "loss": 5.9781,
      "mean_token_accuracy": 0.1821871742606163,
      "num_tokens": 38610822.0,
      "step": 15240
    },
    {
      "entropy": 6.23805685043335,
      "epoch": 1.7593768032313908,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046998350934917654,
      "loss": 6.0029,
      "mean_token_accuracy": 0.1770789697766304,
      "num_tokens": 38623886.0,
      "step": 15245
    },
    {
      "entropy": 6.236564683914184,
      "epoch": 1.7599538372763992,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046996291454163956,
      "loss": 5.9812,
      "mean_token_accuracy": 0.17420328557491302,
      "num_tokens": 38636680.0,
      "step": 15250
    },
    {
      "entropy": 6.2587113857269285,
      "epoch": 1.760530871321408,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004699423131766218,
      "loss": 6.0823,
      "mean_token_accuracy": 0.17095499336719513,
      "num_tokens": 38648841.0,
      "step": 15255
    },
    {
      "entropy": 6.211136674880981,
      "epoch": 1.7611079053664165,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004699217052548161,
      "loss": 6.0056,
      "mean_token_accuracy": 0.17292061299085618,
      "num_tokens": 38660603.0,
      "step": 15260
    },
    {
      "entropy": 6.264367008209229,
      "epoch": 1.7616849394114253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046990109077691577,
      "loss": 6.0593,
      "mean_token_accuracy": 0.17040781825780868,
      "num_tokens": 38674082.0,
      "step": 15265
    },
    {
      "entropy": 6.251727819442749,
      "epoch": 1.762261973456434,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046988046974361406,
      "loss": 5.9876,
      "mean_token_accuracy": 0.17488901168107987,
      "num_tokens": 38686274.0,
      "step": 15270
    },
    {
      "entropy": 6.312477970123291,
      "epoch": 1.7628390075014426,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004698598421556045,
      "loss": 6.0687,
      "mean_token_accuracy": 0.17141490131616594,
      "num_tokens": 38699260.0,
      "step": 15275
    },
    {
      "entropy": 6.240328598022461,
      "epoch": 1.7634160415464513,
      "grad_norm": 2.921875,
      "learning_rate": 0.0004698392080135809,
      "loss": 6.0628,
      "mean_token_accuracy": 0.17000517696142198,
      "num_tokens": 38712054.0,
      "step": 15280
    },
    {
      "entropy": 6.236440801620484,
      "epoch": 1.76399307559146,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004698185673182374,
      "loss": 5.9923,
      "mean_token_accuracy": 0.1765800029039383,
      "num_tokens": 38724634.0,
      "step": 15285
    },
    {
      "entropy": 6.25202693939209,
      "epoch": 1.7645701096364685,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046979792007026817,
      "loss": 5.9628,
      "mean_token_accuracy": 0.17394617944955826,
      "num_tokens": 38737294.0,
      "step": 15290
    },
    {
      "entropy": 6.330475282669068,
      "epoch": 1.7651471436814772,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004697772662703676,
      "loss": 6.1466,
      "mean_token_accuracy": 0.15854198187589646,
      "num_tokens": 38749578.0,
      "step": 15295
    },
    {
      "entropy": 6.262237167358398,
      "epoch": 1.7657241777264858,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046975660591923047,
      "loss": 6.0442,
      "mean_token_accuracy": 0.15982145816087723,
      "num_tokens": 38762282.0,
      "step": 15300
    },
    {
      "entropy": 6.3007566928863525,
      "epoch": 1.7663012117714945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004697359390175516,
      "loss": 6.0285,
      "mean_token_accuracy": 0.17306247353553772,
      "num_tokens": 38773861.0,
      "step": 15305
    },
    {
      "entropy": 6.301147747039795,
      "epoch": 1.7668782458165033,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046971526556602625,
      "loss": 6.1427,
      "mean_token_accuracy": 0.16681707203388213,
      "num_tokens": 38786427.0,
      "step": 15310
    },
    {
      "entropy": 6.30169415473938,
      "epoch": 1.7674552798615117,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004696945855653495,
      "loss": 6.1021,
      "mean_token_accuracy": 0.16261772960424423,
      "num_tokens": 38798496.0,
      "step": 15315
    },
    {
      "entropy": 6.261436891555786,
      "epoch": 1.7680323139065206,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004696738990162172,
      "loss": 6.1365,
      "mean_token_accuracy": 0.16127245575189592,
      "num_tokens": 38812069.0,
      "step": 15320
    },
    {
      "entropy": 6.29582257270813,
      "epoch": 1.768609347951529,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000469653205919325,
      "loss": 6.1237,
      "mean_token_accuracy": 0.1679990828037262,
      "num_tokens": 38824465.0,
      "step": 15325
    },
    {
      "entropy": 6.353593111038208,
      "epoch": 1.7691863819965379,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046963250627536884,
      "loss": 6.1229,
      "mean_token_accuracy": 0.16071866005659102,
      "num_tokens": 38837328.0,
      "step": 15330
    },
    {
      "entropy": 6.258993768692017,
      "epoch": 1.7697634160415463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004696118000850451,
      "loss": 6.0489,
      "mean_token_accuracy": 0.16911157071590424,
      "num_tokens": 38849238.0,
      "step": 15335
    },
    {
      "entropy": 6.222685432434082,
      "epoch": 1.7703404500865552,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046959108734905003,
      "loss": 6.0825,
      "mean_token_accuracy": 0.1641075626015663,
      "num_tokens": 38861479.0,
      "step": 15340
    },
    {
      "entropy": 6.246473884582519,
      "epoch": 1.7709174841315638,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046957036806808045,
      "loss": 6.007,
      "mean_token_accuracy": 0.1772461786866188,
      "num_tokens": 38874342.0,
      "step": 15345
    },
    {
      "entropy": 6.231625413894653,
      "epoch": 1.7714945181765724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004695496422428332,
      "loss": 6.0482,
      "mean_token_accuracy": 0.1703166276216507,
      "num_tokens": 38886595.0,
      "step": 15350
    },
    {
      "entropy": 6.297798919677734,
      "epoch": 1.772071552221581,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004695289098740054,
      "loss": 6.0516,
      "mean_token_accuracy": 0.16702995151281358,
      "num_tokens": 38899338.0,
      "step": 15355
    },
    {
      "entropy": 6.289673328399658,
      "epoch": 1.7726485862665897,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004695081709622943,
      "loss": 6.0265,
      "mean_token_accuracy": 0.17137792557477952,
      "num_tokens": 38911765.0,
      "step": 15360
    },
    {
      "entropy": 6.209077072143555,
      "epoch": 1.7732256203115984,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046948742550839744,
      "loss": 5.9768,
      "mean_token_accuracy": 0.1811099350452423,
      "num_tokens": 38924137.0,
      "step": 15365
    },
    {
      "entropy": 6.293831920623779,
      "epoch": 1.773802654356607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004694666735130127,
      "loss": 6.042,
      "mean_token_accuracy": 0.16767871677875518,
      "num_tokens": 38936401.0,
      "step": 15370
    },
    {
      "entropy": 6.273082637786866,
      "epoch": 1.7743796884016156,
      "grad_norm": 2.03125,
      "learning_rate": 0.000469445914976838,
      "loss": 6.0749,
      "mean_token_accuracy": 0.16592430025339128,
      "num_tokens": 38948224.0,
      "step": 15375
    },
    {
      "entropy": 6.273223733901977,
      "epoch": 1.7749567224466243,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004694251499005715,
      "loss": 6.037,
      "mean_token_accuracy": 0.17387449145317077,
      "num_tokens": 38959924.0,
      "step": 15380
    },
    {
      "entropy": 6.215410614013672,
      "epoch": 1.7755337564916331,
      "grad_norm": 3.65625,
      "learning_rate": 0.0004694043782849116,
      "loss": 6.0723,
      "mean_token_accuracy": 0.1672881633043289,
      "num_tokens": 38973018.0,
      "step": 15385
    },
    {
      "entropy": 6.2906115531921385,
      "epoch": 1.7761107905366416,
      "grad_norm": 1.125,
      "learning_rate": 0.00046938360013055715,
      "loss": 6.1164,
      "mean_token_accuracy": 0.16135561764240264,
      "num_tokens": 38985634.0,
      "step": 15390
    },
    {
      "entropy": 6.3686277866363525,
      "epoch": 1.7766878245816504,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046936281543820673,
      "loss": 6.1421,
      "mean_token_accuracy": 0.1644715130329132,
      "num_tokens": 38999446.0,
      "step": 15395
    },
    {
      "entropy": 6.323219871520996,
      "epoch": 1.7772648586266588,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046934202420855967,
      "loss": 6.1244,
      "mean_token_accuracy": 0.17120789289474486,
      "num_tokens": 39012974.0,
      "step": 15400
    },
    {
      "entropy": 6.157790231704712,
      "epoch": 1.7778418926716677,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046932122644231507,
      "loss": 6.0038,
      "mean_token_accuracy": 0.17291020452976227,
      "num_tokens": 39025316.0,
      "step": 15405
    },
    {
      "entropy": 6.209767818450928,
      "epoch": 1.7784189267166761,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046930042214017256,
      "loss": 5.969,
      "mean_token_accuracy": 0.1822973057627678,
      "num_tokens": 39038174.0,
      "step": 15410
    },
    {
      "entropy": 6.363025379180908,
      "epoch": 1.778995960761685,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004692796113028319,
      "loss": 6.1545,
      "mean_token_accuracy": 0.15527651011943816,
      "num_tokens": 39051943.0,
      "step": 15415
    },
    {
      "entropy": 6.228987741470337,
      "epoch": 1.7795729948066936,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000469258793930993,
      "loss": 6.0561,
      "mean_token_accuracy": 0.16570292711257933,
      "num_tokens": 39064765.0,
      "step": 15420
    },
    {
      "entropy": 6.302031183242798,
      "epoch": 1.7801500288517023,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046923797002535605,
      "loss": 6.1258,
      "mean_token_accuracy": 0.16474866718053818,
      "num_tokens": 39077326.0,
      "step": 15425
    },
    {
      "entropy": 6.303848218917847,
      "epoch": 1.780727062896711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004692171395866214,
      "loss": 6.0094,
      "mean_token_accuracy": 0.17039090692996978,
      "num_tokens": 39090049.0,
      "step": 15430
    },
    {
      "entropy": 6.22796106338501,
      "epoch": 1.7813040969417195,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046919630261548986,
      "loss": 6.0243,
      "mean_token_accuracy": 0.17326881736516953,
      "num_tokens": 39102805.0,
      "step": 15435
    },
    {
      "entropy": 6.148467016220093,
      "epoch": 1.7818811309867282,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046917545911266207,
      "loss": 5.9339,
      "mean_token_accuracy": 0.18568576574325563,
      "num_tokens": 39115966.0,
      "step": 15440
    },
    {
      "entropy": 6.2328081130981445,
      "epoch": 1.7824581650317368,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046915460907883923,
      "loss": 6.0783,
      "mean_token_accuracy": 0.16815552562475206,
      "num_tokens": 39128417.0,
      "step": 15445
    },
    {
      "entropy": 6.260302972793579,
      "epoch": 1.7830351990767457,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046913375251472246,
      "loss": 6.028,
      "mean_token_accuracy": 0.17435874044895172,
      "num_tokens": 39142338.0,
      "step": 15450
    },
    {
      "entropy": 6.25403733253479,
      "epoch": 1.783612233121754,
      "grad_norm": 3.0,
      "learning_rate": 0.00046911288942101345,
      "loss": 6.0697,
      "mean_token_accuracy": 0.17178328335285187,
      "num_tokens": 39155055.0,
      "step": 15455
    },
    {
      "entropy": 6.327509355545044,
      "epoch": 1.784189267166763,
      "grad_norm": 3.546875,
      "learning_rate": 0.0004690920197984138,
      "loss": 6.088,
      "mean_token_accuracy": 0.16623275429010392,
      "num_tokens": 39167170.0,
      "step": 15460
    },
    {
      "entropy": 6.431978321075439,
      "epoch": 1.7847663012117714,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004690711436476254,
      "loss": 6.1754,
      "mean_token_accuracy": 0.1634928062558174,
      "num_tokens": 39181389.0,
      "step": 15465
    },
    {
      "entropy": 6.399842214584351,
      "epoch": 1.7853433352567802,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046905026096935056,
      "loss": 6.1594,
      "mean_token_accuracy": 0.16029439866542816,
      "num_tokens": 39194728.0,
      "step": 15470
    },
    {
      "entropy": 6.165746212005615,
      "epoch": 1.7859203693017887,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004690293717642916,
      "loss": 5.9747,
      "mean_token_accuracy": 0.1781982809305191,
      "num_tokens": 39207945.0,
      "step": 15475
    },
    {
      "entropy": 6.262561368942261,
      "epoch": 1.7864974033467975,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000469008476033151,
      "loss": 6.1209,
      "mean_token_accuracy": 0.1702757239341736,
      "num_tokens": 39221314.0,
      "step": 15480
    },
    {
      "entropy": 6.340551853179932,
      "epoch": 1.7870744373918062,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046898757377663176,
      "loss": 6.0694,
      "mean_token_accuracy": 0.16916078925132752,
      "num_tokens": 39233910.0,
      "step": 15485
    },
    {
      "entropy": 6.28997769355774,
      "epoch": 1.7876514714368148,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046896666499543683,
      "loss": 6.0606,
      "mean_token_accuracy": 0.17222110480070113,
      "num_tokens": 39246027.0,
      "step": 15490
    },
    {
      "entropy": 6.221077728271484,
      "epoch": 1.7882285054818234,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00046894574969026946,
      "loss": 6.0333,
      "mean_token_accuracy": 0.16958654075860977,
      "num_tokens": 39258456.0,
      "step": 15495
    },
    {
      "entropy": 6.365782451629639,
      "epoch": 1.788805539526832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046892482786183313,
      "loss": 6.0906,
      "mean_token_accuracy": 0.16027437299489974,
      "num_tokens": 39270624.0,
      "step": 15500
    },
    {
      "entropy": 6.262175798416138,
      "epoch": 1.7893825735718407,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046890389951083155,
      "loss": 5.9906,
      "mean_token_accuracy": 0.16951557099819184,
      "num_tokens": 39283350.0,
      "step": 15505
    },
    {
      "entropy": 6.29289026260376,
      "epoch": 1.7899596076168494,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046888296463796857,
      "loss": 6.0783,
      "mean_token_accuracy": 0.17297891080379485,
      "num_tokens": 39296321.0,
      "step": 15510
    },
    {
      "entropy": 6.256748819351197,
      "epoch": 1.790536641661858,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004688620232439485,
      "loss": 5.9794,
      "mean_token_accuracy": 0.1770552009344101,
      "num_tokens": 39309223.0,
      "step": 15515
    },
    {
      "entropy": 6.167249011993408,
      "epoch": 1.7911136757068666,
      "grad_norm": 2.640625,
      "learning_rate": 0.00046884107532947547,
      "loss": 5.9671,
      "mean_token_accuracy": 0.1709598571062088,
      "num_tokens": 39322348.0,
      "step": 15520
    },
    {
      "entropy": 6.285998582839966,
      "epoch": 1.7916907097518755,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046882012089525415,
      "loss": 6.0561,
      "mean_token_accuracy": 0.1638656571507454,
      "num_tokens": 39335375.0,
      "step": 15525
    },
    {
      "entropy": 6.330535793304444,
      "epoch": 1.792267743796884,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004687991599419895,
      "loss": 6.0534,
      "mean_token_accuracy": 0.16950902789831163,
      "num_tokens": 39347865.0,
      "step": 15530
    },
    {
      "entropy": 6.270843029022217,
      "epoch": 1.7928447778418928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046877819247038624,
      "loss": 6.0894,
      "mean_token_accuracy": 0.1685513213276863,
      "num_tokens": 39361402.0,
      "step": 15535
    },
    {
      "entropy": 6.257634687423706,
      "epoch": 1.7934218118869012,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004687572184811497,
      "loss": 5.946,
      "mean_token_accuracy": 0.17115099728107452,
      "num_tokens": 39372839.0,
      "step": 15540
    },
    {
      "entropy": 6.258444881439209,
      "epoch": 1.79399884593191,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046873623797498545,
      "loss": 5.9962,
      "mean_token_accuracy": 0.16875378042459488,
      "num_tokens": 39385189.0,
      "step": 15545
    },
    {
      "entropy": 6.356871175765991,
      "epoch": 1.7945758799769185,
      "grad_norm": 1.875,
      "learning_rate": 0.000468715250952599,
      "loss": 6.1782,
      "mean_token_accuracy": 0.16155981868505478,
      "num_tokens": 39397871.0,
      "step": 15550
    },
    {
      "entropy": 6.331059503555298,
      "epoch": 1.7951529140219273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046869425741469635,
      "loss": 6.0892,
      "mean_token_accuracy": 0.16324354857206344,
      "num_tokens": 39411222.0,
      "step": 15555
    },
    {
      "entropy": 6.283884048461914,
      "epoch": 1.795729948066936,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004686732573619835,
      "loss": 6.0929,
      "mean_token_accuracy": 0.1666042074561119,
      "num_tokens": 39423614.0,
      "step": 15560
    },
    {
      "entropy": 6.3039412021636965,
      "epoch": 1.7963069821119446,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004686522507951669,
      "loss": 6.0173,
      "mean_token_accuracy": 0.16653052419424058,
      "num_tokens": 39436299.0,
      "step": 15565
    },
    {
      "entropy": 6.258326625823974,
      "epoch": 1.7968840161569533,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004686312377149531,
      "loss": 6.1395,
      "mean_token_accuracy": 0.15843801349401473,
      "num_tokens": 39448994.0,
      "step": 15570
    },
    {
      "entropy": 6.3444324970245365,
      "epoch": 1.797461050201962,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046861021812204874,
      "loss": 6.1192,
      "mean_token_accuracy": 0.1605915457010269,
      "num_tokens": 39462387.0,
      "step": 15575
    },
    {
      "entropy": 6.342525243759155,
      "epoch": 1.7980380842469705,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046858919201716085,
      "loss": 6.1111,
      "mean_token_accuracy": 0.1640731140971184,
      "num_tokens": 39475082.0,
      "step": 15580
    },
    {
      "entropy": 6.29853310585022,
      "epoch": 1.7986151182919792,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004685681594009966,
      "loss": 6.0898,
      "mean_token_accuracy": 0.16256304532289506,
      "num_tokens": 39487272.0,
      "step": 15585
    },
    {
      "entropy": 6.239833211898803,
      "epoch": 1.799192152336988,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046854712027426357,
      "loss": 6.0776,
      "mean_token_accuracy": 0.16334139704704284,
      "num_tokens": 39502499.0,
      "step": 15590
    },
    {
      "entropy": 6.328673505783081,
      "epoch": 1.7997691863819965,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046852607463766923,
      "loss": 6.0839,
      "mean_token_accuracy": 0.16911648958921432,
      "num_tokens": 39514652.0,
      "step": 15595
    },
    {
      "entropy": 6.401641321182251,
      "epoch": 1.8003462204270053,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004685050224919215,
      "loss": 6.1618,
      "mean_token_accuracy": 0.1580326199531555,
      "num_tokens": 39527859.0,
      "step": 15600
    },
    {
      "entropy": 6.261534261703491,
      "epoch": 1.8009232544720137,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046848396383772844,
      "loss": 6.0345,
      "mean_token_accuracy": 0.1679571434855461,
      "num_tokens": 39541099.0,
      "step": 15605
    },
    {
      "entropy": 6.204374647140503,
      "epoch": 1.8015002885170226,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004684628986757984,
      "loss": 6.0081,
      "mean_token_accuracy": 0.1721584215760231,
      "num_tokens": 39553233.0,
      "step": 15610
    },
    {
      "entropy": 6.285806322097779,
      "epoch": 1.802077322562031,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004684418270068398,
      "loss": 6.0228,
      "mean_token_accuracy": 0.16974806785583496,
      "num_tokens": 39565379.0,
      "step": 15615
    },
    {
      "entropy": 6.238617944717407,
      "epoch": 1.8026543566070399,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004684207488315615,
      "loss": 5.975,
      "mean_token_accuracy": 0.17460445761680604,
      "num_tokens": 39577509.0,
      "step": 15620
    },
    {
      "entropy": 6.322760725021363,
      "epoch": 1.8032313906520485,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004683996641506724,
      "loss": 6.0739,
      "mean_token_accuracy": 0.1633619874715805,
      "num_tokens": 39589960.0,
      "step": 15625
    },
    {
      "entropy": 6.227279090881348,
      "epoch": 1.8038084246970572,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046837857296488163,
      "loss": 5.9781,
      "mean_token_accuracy": 0.17601959705352782,
      "num_tokens": 39602231.0,
      "step": 15630
    },
    {
      "entropy": 6.260754919052124,
      "epoch": 1.8043854587420658,
      "grad_norm": 1.625,
      "learning_rate": 0.00046835747527489857,
      "loss": 6.0532,
      "mean_token_accuracy": 0.1809969499707222,
      "num_tokens": 39615756.0,
      "step": 15635
    },
    {
      "entropy": 6.310202407836914,
      "epoch": 1.8049624927870744,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000468336371081433,
      "loss": 6.0821,
      "mean_token_accuracy": 0.1634535536170006,
      "num_tokens": 39627810.0,
      "step": 15640
    },
    {
      "entropy": 6.2715802669525145,
      "epoch": 1.805539526832083,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046831526038519444,
      "loss": 6.0666,
      "mean_token_accuracy": 0.16980360597372054,
      "num_tokens": 39640930.0,
      "step": 15645
    },
    {
      "entropy": 6.269176197052002,
      "epoch": 1.8061165608770917,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004682941431868933,
      "loss": 6.0093,
      "mean_token_accuracy": 0.1734122857451439,
      "num_tokens": 39652146.0,
      "step": 15650
    },
    {
      "entropy": 6.161038208007812,
      "epoch": 1.8066935949221004,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046827301948723963,
      "loss": 5.9362,
      "mean_token_accuracy": 0.17960268557071685,
      "num_tokens": 39665156.0,
      "step": 15655
    },
    {
      "entropy": 6.255029582977295,
      "epoch": 1.807270628967109,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046825188928694393,
      "loss": 6.0758,
      "mean_token_accuracy": 0.16673412621021272,
      "num_tokens": 39677346.0,
      "step": 15660
    },
    {
      "entropy": 6.312059164047241,
      "epoch": 1.8078476630121179,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004682307525867169,
      "loss": 6.0247,
      "mean_token_accuracy": 0.1778455853462219,
      "num_tokens": 39690613.0,
      "step": 15665
    },
    {
      "entropy": 6.335729742050171,
      "epoch": 1.8084246970571263,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004682096093872695,
      "loss": 6.0163,
      "mean_token_accuracy": 0.17328280210494995,
      "num_tokens": 39703111.0,
      "step": 15670
    },
    {
      "entropy": 6.2205384254455565,
      "epoch": 1.8090017311021351,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046818845968931284,
      "loss": 6.0623,
      "mean_token_accuracy": 0.1691240191459656,
      "num_tokens": 39715441.0,
      "step": 15675
    },
    {
      "entropy": 6.262801313400269,
      "epoch": 1.8095787651471436,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046816730349355843,
      "loss": 6.0161,
      "mean_token_accuracy": 0.17307625114917755,
      "num_tokens": 39728437.0,
      "step": 15680
    },
    {
      "entropy": 6.241965198516846,
      "epoch": 1.8101557991921524,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046814614080071756,
      "loss": 6.0466,
      "mean_token_accuracy": 0.17203230857849122,
      "num_tokens": 39740324.0,
      "step": 15685
    },
    {
      "entropy": 6.294243240356446,
      "epoch": 1.8107328332371608,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046812497161150224,
      "loss": 6.0171,
      "mean_token_accuracy": 0.16689405590295792,
      "num_tokens": 39752158.0,
      "step": 15690
    },
    {
      "entropy": 6.28272008895874,
      "epoch": 1.8113098672821697,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046810379592662445,
      "loss": 6.0772,
      "mean_token_accuracy": 0.17248584032058717,
      "num_tokens": 39766296.0,
      "step": 15695
    },
    {
      "entropy": 6.234416913986206,
      "epoch": 1.8118869013271783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046808261374679637,
      "loss": 6.0515,
      "mean_token_accuracy": 0.17033150643110276,
      "num_tokens": 39779690.0,
      "step": 15700
    },
    {
      "entropy": 6.25835657119751,
      "epoch": 1.812463935372187,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004680614250727305,
      "loss": 6.0099,
      "mean_token_accuracy": 0.1731962352991104,
      "num_tokens": 39790617.0,
      "step": 15705
    },
    {
      "entropy": 6.252574872970581,
      "epoch": 1.8130409694171956,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004680402299051395,
      "loss": 6.036,
      "mean_token_accuracy": 0.1660449355840683,
      "num_tokens": 39802994.0,
      "step": 15710
    },
    {
      "entropy": 6.25236554145813,
      "epoch": 1.8136180034622043,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004680190282447363,
      "loss": 6.0521,
      "mean_token_accuracy": 0.169236396253109,
      "num_tokens": 39815618.0,
      "step": 15715
    },
    {
      "entropy": 6.24625039100647,
      "epoch": 1.814195037507213,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004679978200922339,
      "loss": 6.0194,
      "mean_token_accuracy": 0.1698412522673607,
      "num_tokens": 39827283.0,
      "step": 15720
    },
    {
      "entropy": 6.2522207736969,
      "epoch": 1.8147720715522215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004679766054483457,
      "loss": 5.9816,
      "mean_token_accuracy": 0.17221676111221312,
      "num_tokens": 39840116.0,
      "step": 15725
    },
    {
      "entropy": 6.15627088546753,
      "epoch": 1.8153491055972304,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004679553843137852,
      "loss": 5.979,
      "mean_token_accuracy": 0.18153376132249832,
      "num_tokens": 39851770.0,
      "step": 15730
    },
    {
      "entropy": 6.156879806518555,
      "epoch": 1.8159261396422388,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046793415668926625,
      "loss": 5.9414,
      "mean_token_accuracy": 0.17151160389184952,
      "num_tokens": 39863803.0,
      "step": 15735
    },
    {
      "entropy": 6.248953819274902,
      "epoch": 1.8165031736872477,
      "grad_norm": 1.25,
      "learning_rate": 0.0004679129225755028,
      "loss": 5.9871,
      "mean_token_accuracy": 0.17617322951555253,
      "num_tokens": 39875713.0,
      "step": 15740
    },
    {
      "entropy": 6.18450894355774,
      "epoch": 1.817080207732256,
      "grad_norm": 1.234375,
      "learning_rate": 0.000467891681973209,
      "loss": 5.9588,
      "mean_token_accuracy": 0.17611952871084213,
      "num_tokens": 39888670.0,
      "step": 15745
    },
    {
      "entropy": 6.142603588104248,
      "epoch": 1.817657241777265,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00046787043488309926,
      "loss": 5.9825,
      "mean_token_accuracy": 0.18451358526945114,
      "num_tokens": 39900845.0,
      "step": 15750
    },
    {
      "entropy": 6.346268844604492,
      "epoch": 1.8182342758222734,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004678491813058882,
      "loss": 6.0927,
      "mean_token_accuracy": 0.1648651421070099,
      "num_tokens": 39912268.0,
      "step": 15755
    },
    {
      "entropy": 6.277725410461426,
      "epoch": 1.8188113098672822,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004678279212422908,
      "loss": 6.0369,
      "mean_token_accuracy": 0.1688872143626213,
      "num_tokens": 39924213.0,
      "step": 15760
    },
    {
      "entropy": 6.276968383789063,
      "epoch": 1.8193883439122909,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004678066546930221,
      "loss": 6.0825,
      "mean_token_accuracy": 0.1626095324754715,
      "num_tokens": 39937069.0,
      "step": 15765
    },
    {
      "entropy": 6.237314939498901,
      "epoch": 1.8199653779572995,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046778538165879725,
      "loss": 6.0041,
      "mean_token_accuracy": 0.16766275614500045,
      "num_tokens": 39948988.0,
      "step": 15770
    },
    {
      "entropy": 6.2724080085754395,
      "epoch": 1.8205424120023082,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046776410214033185,
      "loss": 6.0754,
      "mean_token_accuracy": 0.16913768351078035,
      "num_tokens": 39962177.0,
      "step": 15775
    },
    {
      "entropy": 6.317848825454712,
      "epoch": 1.8211194460473168,
      "grad_norm": 2.34375,
      "learning_rate": 0.0004677428161383417,
      "loss": 6.0752,
      "mean_token_accuracy": 0.16904523521661757,
      "num_tokens": 39975329.0,
      "step": 15780
    },
    {
      "entropy": 6.264898729324341,
      "epoch": 1.8216964800923254,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004677215236535426,
      "loss": 6.0407,
      "mean_token_accuracy": 0.16779443323612214,
      "num_tokens": 39987923.0,
      "step": 15785
    },
    {
      "entropy": 6.34999647140503,
      "epoch": 1.822273514137334,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004677002246866508,
      "loss": 6.125,
      "mean_token_accuracy": 0.16695744693279266,
      "num_tokens": 40001596.0,
      "step": 15790
    },
    {
      "entropy": 6.289482831954956,
      "epoch": 1.8228505481823427,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046767891923838264,
      "loss": 6.0273,
      "mean_token_accuracy": 0.1737503945827484,
      "num_tokens": 40014207.0,
      "step": 15795
    },
    {
      "entropy": 6.369691467285156,
      "epoch": 1.8234275822273514,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004676576073094548,
      "loss": 6.0887,
      "mean_token_accuracy": 0.16101481169462203,
      "num_tokens": 40027931.0,
      "step": 15800
    },
    {
      "entropy": 6.339088201522827,
      "epoch": 1.8240046162723602,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046763628890058396,
      "loss": 6.0971,
      "mean_token_accuracy": 0.16571219712495805,
      "num_tokens": 40040463.0,
      "step": 15805
    },
    {
      "entropy": 6.221599912643432,
      "epoch": 1.8245816503173686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046761496401248734,
      "loss": 5.9669,
      "mean_token_accuracy": 0.17340231388807298,
      "num_tokens": 40053085.0,
      "step": 15810
    },
    {
      "entropy": 6.316984224319458,
      "epoch": 1.8251586843623775,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000467593632645882,
      "loss": 6.1351,
      "mean_token_accuracy": 0.1690345823764801,
      "num_tokens": 40065771.0,
      "step": 15815
    },
    {
      "entropy": 6.296414136886597,
      "epoch": 1.825735718407386,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004675722948014855,
      "loss": 6.0543,
      "mean_token_accuracy": 0.17363547384738923,
      "num_tokens": 40078939.0,
      "step": 15820
    },
    {
      "entropy": 6.23414626121521,
      "epoch": 1.8263127524523948,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046755095048001556,
      "loss": 5.9846,
      "mean_token_accuracy": 0.17467279732227325,
      "num_tokens": 40091565.0,
      "step": 15825
    },
    {
      "entropy": 6.167823934555054,
      "epoch": 1.8268897864974032,
      "grad_norm": 1.96875,
      "learning_rate": 0.0004675295996821899,
      "loss": 6.0007,
      "mean_token_accuracy": 0.16983738988637925,
      "num_tokens": 40103619.0,
      "step": 15830
    },
    {
      "entropy": 6.311656141281128,
      "epoch": 1.827466820542412,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000467508242408727,
      "loss": 6.0797,
      "mean_token_accuracy": 0.1673286721110344,
      "num_tokens": 40116370.0,
      "step": 15835
    },
    {
      "entropy": 6.304900693893432,
      "epoch": 1.8280438545874207,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004674868786603448,
      "loss": 5.9791,
      "mean_token_accuracy": 0.1691321074962616,
      "num_tokens": 40129015.0,
      "step": 15840
    },
    {
      "entropy": 6.241245222091675,
      "epoch": 1.8286208886324293,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004674655084377622,
      "loss": 5.9891,
      "mean_token_accuracy": 0.1670050650835037,
      "num_tokens": 40141437.0,
      "step": 15845
    },
    {
      "entropy": 6.212176847457886,
      "epoch": 1.829197922677438,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004674441317416978,
      "loss": 6.0434,
      "mean_token_accuracy": 0.16276399046182632,
      "num_tokens": 40154516.0,
      "step": 15850
    },
    {
      "entropy": 6.238998651504517,
      "epoch": 1.8297749567224466,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046742274857287057,
      "loss": 5.9534,
      "mean_token_accuracy": 0.1708405777812004,
      "num_tokens": 40166598.0,
      "step": 15855
    },
    {
      "entropy": 6.257824850082398,
      "epoch": 1.8303519907674553,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004674013589319998,
      "loss": 6.0465,
      "mean_token_accuracy": 0.1640569418668747,
      "num_tokens": 40178905.0,
      "step": 15860
    },
    {
      "entropy": 6.317538118362426,
      "epoch": 1.830929024812464,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004673799628198049,
      "loss": 6.1051,
      "mean_token_accuracy": 0.16486846506595612,
      "num_tokens": 40191678.0,
      "step": 15865
    },
    {
      "entropy": 6.285384607315064,
      "epoch": 1.8315060588574728,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046735856023700546,
      "loss": 6.0262,
      "mean_token_accuracy": 0.16484537124633789,
      "num_tokens": 40204680.0,
      "step": 15870
    },
    {
      "entropy": 6.233611536026001,
      "epoch": 1.8320830929024812,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004673371511843215,
      "loss": 6.0817,
      "mean_token_accuracy": 0.17095134109258653,
      "num_tokens": 40217026.0,
      "step": 15875
    },
    {
      "entropy": 6.258399248123169,
      "epoch": 1.83266012694749,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004673157356624729,
      "loss": 5.9863,
      "mean_token_accuracy": 0.1706467315554619,
      "num_tokens": 40228950.0,
      "step": 15880
    },
    {
      "entropy": 6.287282991409302,
      "epoch": 1.8332371609924984,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004672943136721801,
      "loss": 6.0836,
      "mean_token_accuracy": 0.16891248077154158,
      "num_tokens": 40241294.0,
      "step": 15885
    },
    {
      "entropy": 6.341698074340821,
      "epoch": 1.8338141950375073,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004672728852141636,
      "loss": 6.0943,
      "mean_token_accuracy": 0.16762983351945876,
      "num_tokens": 40253331.0,
      "step": 15890
    },
    {
      "entropy": 6.258095550537109,
      "epoch": 1.8343912290825157,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046725145028914404,
      "loss": 6.0801,
      "mean_token_accuracy": 0.16581851840019227,
      "num_tokens": 40266059.0,
      "step": 15895
    },
    {
      "entropy": 6.256051778793335,
      "epoch": 1.8349682631275246,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004672300088978425,
      "loss": 6.0242,
      "mean_token_accuracy": 0.1764928877353668,
      "num_tokens": 40278732.0,
      "step": 15900
    },
    {
      "entropy": 6.2313501834869385,
      "epoch": 1.8355452971725332,
      "grad_norm": 1.625,
      "learning_rate": 0.0004672085610409801,
      "loss": 5.946,
      "mean_token_accuracy": 0.18284683674573898,
      "num_tokens": 40290428.0,
      "step": 15905
    },
    {
      "entropy": 6.177774572372437,
      "epoch": 1.8361223312175419,
      "grad_norm": 1.0,
      "learning_rate": 0.00046718710671927815,
      "loss": 6.0255,
      "mean_token_accuracy": 0.17687230110168456,
      "num_tokens": 40302153.0,
      "step": 15910
    },
    {
      "entropy": 6.310884857177735,
      "epoch": 1.8366993652625505,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046716564593345843,
      "loss": 6.0497,
      "mean_token_accuracy": 0.1659906193614006,
      "num_tokens": 40315493.0,
      "step": 15915
    },
    {
      "entropy": 6.3170911312103275,
      "epoch": 1.8372763993075591,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046714417868424265,
      "loss": 6.071,
      "mean_token_accuracy": 0.1596185088157654,
      "num_tokens": 40328157.0,
      "step": 15920
    },
    {
      "entropy": 6.326702356338501,
      "epoch": 1.8378534333525678,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046712270497235284,
      "loss": 6.0964,
      "mean_token_accuracy": 0.1661912366747856,
      "num_tokens": 40341684.0,
      "step": 15925
    },
    {
      "entropy": 6.280466222763062,
      "epoch": 1.8384304673975764,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004671012247985112,
      "loss": 6.1368,
      "mean_token_accuracy": 0.16310458034276962,
      "num_tokens": 40354002.0,
      "step": 15930
    },
    {
      "entropy": 6.223762655258179,
      "epoch": 1.839007501442585,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046707973816344044,
      "loss": 6.0223,
      "mean_token_accuracy": 0.17710338532924652,
      "num_tokens": 40366211.0,
      "step": 15935
    },
    {
      "entropy": 6.310409069061279,
      "epoch": 1.8395845354875937,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046705824506786304,
      "loss": 6.098,
      "mean_token_accuracy": 0.1691246896982193,
      "num_tokens": 40377957.0,
      "step": 15940
    },
    {
      "entropy": 6.291127347946167,
      "epoch": 1.8401615695326026,
      "grad_norm": 3.203125,
      "learning_rate": 0.00046703674551250193,
      "loss": 6.0688,
      "mean_token_accuracy": 0.16395938247442246,
      "num_tokens": 40391018.0,
      "step": 15945
    },
    {
      "entropy": 6.228165149688721,
      "epoch": 1.840738603577611,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004670152394980803,
      "loss": 6.0586,
      "mean_token_accuracy": 0.168018639087677,
      "num_tokens": 40405134.0,
      "step": 15950
    },
    {
      "entropy": 6.349109077453614,
      "epoch": 1.8413156376226198,
      "grad_norm": 4.15625,
      "learning_rate": 0.0004669937270253214,
      "loss": 6.1221,
      "mean_token_accuracy": 0.1632961302995682,
      "num_tokens": 40418131.0,
      "step": 15955
    },
    {
      "entropy": 6.31708927154541,
      "epoch": 1.8418926716676283,
      "grad_norm": 1.0625,
      "learning_rate": 0.000466972208094949,
      "loss": 6.0529,
      "mean_token_accuracy": 0.16544150561094284,
      "num_tokens": 40430614.0,
      "step": 15960
    },
    {
      "entropy": 6.308804941177368,
      "epoch": 1.8424697057126371,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046695068270768665,
      "loss": 6.0683,
      "mean_token_accuracy": 0.16248993575572968,
      "num_tokens": 40443200.0,
      "step": 15965
    },
    {
      "entropy": 6.2004454135894775,
      "epoch": 1.8430467397576455,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046692915086425846,
      "loss": 6.0024,
      "mean_token_accuracy": 0.1657170385122299,
      "num_tokens": 40455662.0,
      "step": 15970
    },
    {
      "entropy": 6.319042015075683,
      "epoch": 1.8436237738026544,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046690761256538857,
      "loss": 6.0141,
      "mean_token_accuracy": 0.1710167035460472,
      "num_tokens": 40468247.0,
      "step": 15975
    },
    {
      "entropy": 6.324203443527222,
      "epoch": 1.844200807847663,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004668860678118015,
      "loss": 6.1585,
      "mean_token_accuracy": 0.16913821399211884,
      "num_tokens": 40480607.0,
      "step": 15980
    },
    {
      "entropy": 6.323934125900268,
      "epoch": 1.8447778418926717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046686451660422185,
      "loss": 6.0394,
      "mean_token_accuracy": 0.16791746020317078,
      "num_tokens": 40492140.0,
      "step": 15985
    },
    {
      "entropy": 6.285348176956177,
      "epoch": 1.8453548759376803,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046684295894337455,
      "loss": 6.0897,
      "mean_token_accuracy": 0.16872338950634003,
      "num_tokens": 40504513.0,
      "step": 15990
    },
    {
      "entropy": 6.237533950805664,
      "epoch": 1.845931909982689,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004668213948299845,
      "loss": 6.0184,
      "mean_token_accuracy": 0.16824758648872376,
      "num_tokens": 40517258.0,
      "step": 15995
    },
    {
      "entropy": 6.234326934814453,
      "epoch": 1.8465089440276976,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004667998242647772,
      "loss": 6.0702,
      "mean_token_accuracy": 0.16491250544786454,
      "num_tokens": 40529092.0,
      "step": 16000
    },
    {
      "entropy": 6.266136550903321,
      "epoch": 1.8470859780727062,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000466778247248478,
      "loss": 6.0419,
      "mean_token_accuracy": 0.1642804428935051,
      "num_tokens": 40543586.0,
      "step": 16005
    },
    {
      "entropy": 6.313886404037476,
      "epoch": 1.847663012117715,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046675666378181275,
      "loss": 6.0448,
      "mean_token_accuracy": 0.17515656352043152,
      "num_tokens": 40555721.0,
      "step": 16010
    },
    {
      "entropy": 6.2824138641357425,
      "epoch": 1.8482400461627235,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004667350738655074,
      "loss": 6.1204,
      "mean_token_accuracy": 0.1662580594420433,
      "num_tokens": 40567547.0,
      "step": 16015
    },
    {
      "entropy": 6.233053827285767,
      "epoch": 1.8488170802077324,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046671347750028806,
      "loss": 5.9538,
      "mean_token_accuracy": 0.17195529937744142,
      "num_tokens": 40579498.0,
      "step": 16020
    },
    {
      "entropy": 6.263859844207763,
      "epoch": 1.8493941142527408,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004666918746868811,
      "loss": 6.1004,
      "mean_token_accuracy": 0.1605726420879364,
      "num_tokens": 40593853.0,
      "step": 16025
    },
    {
      "entropy": 6.250426197052002,
      "epoch": 1.8499711482977497,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004666702654260133,
      "loss": 5.992,
      "mean_token_accuracy": 0.17157431542873383,
      "num_tokens": 40605737.0,
      "step": 16030
    },
    {
      "entropy": 6.225319910049438,
      "epoch": 1.850548182342758,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046664864971841113,
      "loss": 5.9642,
      "mean_token_accuracy": 0.172415129840374,
      "num_tokens": 40617839.0,
      "step": 16035
    },
    {
      "entropy": 6.265378141403199,
      "epoch": 1.851125216387767,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046662702756480195,
      "loss": 6.0229,
      "mean_token_accuracy": 0.17276886850595474,
      "num_tokens": 40630247.0,
      "step": 16040
    },
    {
      "entropy": 6.222239351272583,
      "epoch": 1.8517022504327756,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046660539896591286,
      "loss": 5.9997,
      "mean_token_accuracy": 0.17419626861810683,
      "num_tokens": 40643869.0,
      "step": 16045
    },
    {
      "entropy": 6.3070024967193605,
      "epoch": 1.8522792844777842,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004665837639224713,
      "loss": 6.012,
      "mean_token_accuracy": 0.16875723153352737,
      "num_tokens": 40655389.0,
      "step": 16050
    },
    {
      "entropy": 6.279503965377808,
      "epoch": 1.8528563185227929,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046656212243520505,
      "loss": 6.0528,
      "mean_token_accuracy": 0.165364670753479,
      "num_tokens": 40667663.0,
      "step": 16055
    },
    {
      "entropy": 6.244712591171265,
      "epoch": 1.8534333525678015,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046654047450484193,
      "loss": 5.9987,
      "mean_token_accuracy": 0.17057995349168778,
      "num_tokens": 40680623.0,
      "step": 16060
    },
    {
      "entropy": 6.220044422149658,
      "epoch": 1.8540103866128101,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004665188201321102,
      "loss": 5.9599,
      "mean_token_accuracy": 0.1741309642791748,
      "num_tokens": 40692679.0,
      "step": 16065
    },
    {
      "entropy": 6.258392667770385,
      "epoch": 1.8545874206578188,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046649715931773795,
      "loss": 6.1387,
      "mean_token_accuracy": 0.1673455372452736,
      "num_tokens": 40705187.0,
      "step": 16070
    },
    {
      "entropy": 6.291736078262329,
      "epoch": 1.8551644547028274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000466475492062454,
      "loss": 6.087,
      "mean_token_accuracy": 0.17339060008525847,
      "num_tokens": 40717854.0,
      "step": 16075
    },
    {
      "entropy": 6.2766063690185545,
      "epoch": 1.855741488747836,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046645381836698684,
      "loss": 6.1085,
      "mean_token_accuracy": 0.16265238225460052,
      "num_tokens": 40731039.0,
      "step": 16080
    },
    {
      "entropy": 6.2548644065856935,
      "epoch": 1.856318522792845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004664321382320657,
      "loss": 6.0184,
      "mean_token_accuracy": 0.17314105480909348,
      "num_tokens": 40745065.0,
      "step": 16085
    },
    {
      "entropy": 6.258414602279663,
      "epoch": 1.8568955568378533,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046641045165841965,
      "loss": 6.1141,
      "mean_token_accuracy": 0.16226121485233308,
      "num_tokens": 40758107.0,
      "step": 16090
    },
    {
      "entropy": 6.270643854141236,
      "epoch": 1.8574725908828622,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046638875864677814,
      "loss": 6.1105,
      "mean_token_accuracy": 0.16872740387916565,
      "num_tokens": 40771082.0,
      "step": 16095
    },
    {
      "entropy": 6.305703115463257,
      "epoch": 1.8580496249278706,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004663670591978708,
      "loss": 6.0308,
      "mean_token_accuracy": 0.16707999408245086,
      "num_tokens": 40784165.0,
      "step": 16100
    },
    {
      "entropy": 6.237900972366333,
      "epoch": 1.8586266589728795,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004663453533124275,
      "loss": 5.9312,
      "mean_token_accuracy": 0.1778552234172821,
      "num_tokens": 40796279.0,
      "step": 16105
    },
    {
      "entropy": 6.1860754013061525,
      "epoch": 1.859203693017888,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004663236409911783,
      "loss": 5.9726,
      "mean_token_accuracy": 0.1731626346707344,
      "num_tokens": 40809074.0,
      "step": 16110
    },
    {
      "entropy": 6.17712368965149,
      "epoch": 1.8597807270628968,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046630192223485345,
      "loss": 5.9326,
      "mean_token_accuracy": 0.17486831545829773,
      "num_tokens": 40821086.0,
      "step": 16115
    },
    {
      "entropy": 6.314848279953003,
      "epoch": 1.8603577611079054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046628019704418345,
      "loss": 6.0452,
      "mean_token_accuracy": 0.16555711328983308,
      "num_tokens": 40832697.0,
      "step": 16120
    },
    {
      "entropy": 6.169562435150146,
      "epoch": 1.860934795152914,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000466258465419899,
      "loss": 6.0551,
      "mean_token_accuracy": 0.16342287957668306,
      "num_tokens": 40845268.0,
      "step": 16125
    },
    {
      "entropy": 6.295875883102417,
      "epoch": 1.8615118291979227,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004662367273627312,
      "loss": 6.0789,
      "mean_token_accuracy": 0.15954152941703797,
      "num_tokens": 40858314.0,
      "step": 16130
    },
    {
      "entropy": 6.303014135360717,
      "epoch": 1.8620888632429313,
      "grad_norm": 1.109375,
      "learning_rate": 0.000466214982873411,
      "loss": 6.0057,
      "mean_token_accuracy": 0.17139186710119247,
      "num_tokens": 40869982.0,
      "step": 16135
    },
    {
      "entropy": 6.248759126663208,
      "epoch": 1.86266589728794,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046619323195266975,
      "loss": 5.9716,
      "mean_token_accuracy": 0.1751272961497307,
      "num_tokens": 40881831.0,
      "step": 16140
    },
    {
      "entropy": 6.161066722869873,
      "epoch": 1.8632429313329486,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004661714746012392,
      "loss": 6.008,
      "mean_token_accuracy": 0.17016322016716004,
      "num_tokens": 40894279.0,
      "step": 16145
    },
    {
      "entropy": 6.233482503890992,
      "epoch": 1.8638199653779572,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000466149710819851,
      "loss": 6.1207,
      "mean_token_accuracy": 0.16817092448472976,
      "num_tokens": 40906960.0,
      "step": 16150
    },
    {
      "entropy": 6.340131759643555,
      "epoch": 1.8643969994229659,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004661279406092373,
      "loss": 6.0666,
      "mean_token_accuracy": 0.16523224860429764,
      "num_tokens": 40920688.0,
      "step": 16155
    },
    {
      "entropy": 6.30043215751648,
      "epoch": 1.8649740334679747,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046610616397013027,
      "loss": 6.0865,
      "mean_token_accuracy": 0.1653803125023842,
      "num_tokens": 40933017.0,
      "step": 16160
    },
    {
      "entropy": 6.219679450988769,
      "epoch": 1.8655510675129832,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004660843809032623,
      "loss": 5.9787,
      "mean_token_accuracy": 0.17148686349391937,
      "num_tokens": 40947179.0,
      "step": 16165
    },
    {
      "entropy": 6.283222913742065,
      "epoch": 1.866128101557992,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004660625914093661,
      "loss": 6.0981,
      "mean_token_accuracy": 0.1636438712477684,
      "num_tokens": 40958806.0,
      "step": 16170
    },
    {
      "entropy": 6.313401699066162,
      "epoch": 1.8667051356030004,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004660407954891745,
      "loss": 6.0434,
      "mean_token_accuracy": 0.17289685904979707,
      "num_tokens": 40970667.0,
      "step": 16175
    },
    {
      "entropy": 6.22429256439209,
      "epoch": 1.8672821696480093,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004660189931434207,
      "loss": 6.0033,
      "mean_token_accuracy": 0.16919858902692794,
      "num_tokens": 40982663.0,
      "step": 16180
    },
    {
      "entropy": 6.23825216293335,
      "epoch": 1.8678592036930177,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004659971843728379,
      "loss": 6.0046,
      "mean_token_accuracy": 0.16918305009603501,
      "num_tokens": 40995168.0,
      "step": 16185
    },
    {
      "entropy": 6.307004165649414,
      "epoch": 1.8684362377380266,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004659753691781597,
      "loss": 6.0791,
      "mean_token_accuracy": 0.1678227812051773,
      "num_tokens": 41007159.0,
      "step": 16190
    },
    {
      "entropy": 6.231713056564331,
      "epoch": 1.8690132717830352,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004659535475601198,
      "loss": 5.9706,
      "mean_token_accuracy": 0.16688550263643265,
      "num_tokens": 41019860.0,
      "step": 16195
    },
    {
      "entropy": 6.203270149230957,
      "epoch": 1.8695903058280439,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046593171951945226,
      "loss": 6.0379,
      "mean_token_accuracy": 0.1657431960105896,
      "num_tokens": 41031635.0,
      "step": 16200
    },
    {
      "entropy": 6.259737777709961,
      "epoch": 1.8701673398730525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046590988505689114,
      "loss": 6.0626,
      "mean_token_accuracy": 0.16913105845451354,
      "num_tokens": 41044678.0,
      "step": 16205
    },
    {
      "entropy": 6.280846738815308,
      "epoch": 1.8707443739180611,
      "grad_norm": 1.390625,
      "learning_rate": 0.00046588804417317084,
      "loss": 6.1166,
      "mean_token_accuracy": 0.16795769333839417,
      "num_tokens": 41056799.0,
      "step": 16210
    },
    {
      "entropy": 6.287213611602783,
      "epoch": 1.8713214079630698,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046586619686902597,
      "loss": 6.1045,
      "mean_token_accuracy": 0.16530012637376784,
      "num_tokens": 41069700.0,
      "step": 16215
    },
    {
      "entropy": 6.308014869689941,
      "epoch": 1.8718984420080784,
      "grad_norm": 1.25,
      "learning_rate": 0.00046584434314519144,
      "loss": 6.0134,
      "mean_token_accuracy": 0.17250948995351792,
      "num_tokens": 41081427.0,
      "step": 16220
    },
    {
      "entropy": 6.290389108657837,
      "epoch": 1.8724754760530873,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004658224830024022,
      "loss": 6.0395,
      "mean_token_accuracy": 0.17142314016819,
      "num_tokens": 41094770.0,
      "step": 16225
    },
    {
      "entropy": 6.228277778625488,
      "epoch": 1.8730525100980957,
      "grad_norm": 1.25,
      "learning_rate": 0.0004658006164413935,
      "loss": 5.9786,
      "mean_token_accuracy": 0.1729537844657898,
      "num_tokens": 41106802.0,
      "step": 16230
    },
    {
      "entropy": 6.216348171234131,
      "epoch": 1.8736295441431046,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004657787434629009,
      "loss": 6.0816,
      "mean_token_accuracy": 0.17646737545728683,
      "num_tokens": 41120125.0,
      "step": 16235
    },
    {
      "entropy": 6.105568265914917,
      "epoch": 1.874206578188113,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046575686406765993,
      "loss": 6.052,
      "mean_token_accuracy": 0.17620332092046737,
      "num_tokens": 41134267.0,
      "step": 16240
    },
    {
      "entropy": 6.314919900894165,
      "epoch": 1.8747836122331218,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046573497825640664,
      "loss": 6.0825,
      "mean_token_accuracy": 0.1711890734732151,
      "num_tokens": 41147266.0,
      "step": 16245
    },
    {
      "entropy": 6.271852016448975,
      "epoch": 1.8753606462781303,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004657130860298771,
      "loss": 6.0167,
      "mean_token_accuracy": 0.17353598028421402,
      "num_tokens": 41159666.0,
      "step": 16250
    },
    {
      "entropy": 6.266560459136963,
      "epoch": 1.8759376803231391,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004656911873888077,
      "loss": 6.0694,
      "mean_token_accuracy": 0.16777328699827193,
      "num_tokens": 41171947.0,
      "step": 16255
    },
    {
      "entropy": 6.214862108230591,
      "epoch": 1.8765147143681478,
      "grad_norm": 1.625,
      "learning_rate": 0.0004656692823339349,
      "loss": 6.0512,
      "mean_token_accuracy": 0.1780415490269661,
      "num_tokens": 41186376.0,
      "step": 16260
    },
    {
      "entropy": 6.257151699066162,
      "epoch": 1.8770917484131564,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004656473708659955,
      "loss": 6.073,
      "mean_token_accuracy": 0.1731489807367325,
      "num_tokens": 41199730.0,
      "step": 16265
    },
    {
      "entropy": 6.250900554656982,
      "epoch": 1.877668782458165,
      "grad_norm": 1.375,
      "learning_rate": 0.00046562545298572646,
      "loss": 6.0228,
      "mean_token_accuracy": 0.17911915779113768,
      "num_tokens": 41212502.0,
      "step": 16270
    },
    {
      "entropy": 6.185820722579956,
      "epoch": 1.8782458165031737,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000465603528693865,
      "loss": 5.9887,
      "mean_token_accuracy": 0.17167483866214753,
      "num_tokens": 41226006.0,
      "step": 16275
    },
    {
      "entropy": 6.316561460494995,
      "epoch": 1.8788228505481823,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046558159799114853,
      "loss": 6.044,
      "mean_token_accuracy": 0.16871319115161895,
      "num_tokens": 41239172.0,
      "step": 16280
    },
    {
      "entropy": 6.207419347763062,
      "epoch": 1.879399884593191,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004655596608783147,
      "loss": 5.9569,
      "mean_token_accuracy": 0.17857006043195725,
      "num_tokens": 41251307.0,
      "step": 16285
    },
    {
      "entropy": 6.265816640853882,
      "epoch": 1.8799769186381996,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046553771735610136,
      "loss": 6.0126,
      "mean_token_accuracy": 0.16932079792022706,
      "num_tokens": 41264634.0,
      "step": 16290
    },
    {
      "entropy": 6.2975006103515625,
      "epoch": 1.8805539526832082,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004655157674252465,
      "loss": 6.0935,
      "mean_token_accuracy": 0.15814897865056993,
      "num_tokens": 41278205.0,
      "step": 16295
    },
    {
      "entropy": 6.317185497283935,
      "epoch": 1.881130986728217,
      "grad_norm": 1.125,
      "learning_rate": 0.00046549381108648843,
      "loss": 6.0527,
      "mean_token_accuracy": 0.1710874080657959,
      "num_tokens": 41290447.0,
      "step": 16300
    },
    {
      "entropy": 6.305042934417725,
      "epoch": 1.8817080207732255,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004654718483405656,
      "loss": 5.9676,
      "mean_token_accuracy": 0.17561424076557158,
      "num_tokens": 41302948.0,
      "step": 16305
    },
    {
      "entropy": 6.232265901565552,
      "epoch": 1.8822850548182344,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00046544987918821685,
      "loss": 6.1008,
      "mean_token_accuracy": 0.16179861277341842,
      "num_tokens": 41315237.0,
      "step": 16310
    },
    {
      "entropy": 6.131192016601562,
      "epoch": 1.8828620888632428,
      "grad_norm": 2.0625,
      "learning_rate": 0.000465427903630181,
      "loss": 5.9238,
      "mean_token_accuracy": 0.18151000291109085,
      "num_tokens": 41327426.0,
      "step": 16315
    },
    {
      "entropy": 6.361543273925781,
      "epoch": 1.8834391229082517,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004654059216671972,
      "loss": 6.1255,
      "mean_token_accuracy": 0.1627553418278694,
      "num_tokens": 41340517.0,
      "step": 16320
    },
    {
      "entropy": 6.213067150115966,
      "epoch": 1.88401615695326,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004653839333000048,
      "loss": 6.0259,
      "mean_token_accuracy": 0.1693815842270851,
      "num_tokens": 41353520.0,
      "step": 16325
    },
    {
      "entropy": 6.259006404876709,
      "epoch": 1.884593190998269,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046536193852934334,
      "loss": 6.0648,
      "mean_token_accuracy": 0.1702921360731125,
      "num_tokens": 41366326.0,
      "step": 16330
    },
    {
      "entropy": 6.230393600463867,
      "epoch": 1.8851702250432776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046533993735595273,
      "loss": 5.974,
      "mean_token_accuracy": 0.17455315291881562,
      "num_tokens": 41377405.0,
      "step": 16335
    },
    {
      "entropy": 6.313044166564941,
      "epoch": 1.8857472590882862,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046531792978057277,
      "loss": 6.1152,
      "mean_token_accuracy": 0.1636503666639328,
      "num_tokens": 41391038.0,
      "step": 16340
    },
    {
      "entropy": 6.267929553985596,
      "epoch": 1.8863242931332949,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004652959158039438,
      "loss": 6.0527,
      "mean_token_accuracy": 0.1683339387178421,
      "num_tokens": 41404069.0,
      "step": 16345
    },
    {
      "entropy": 6.2727337837219235,
      "epoch": 1.8869013271783035,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004652738954268062,
      "loss": 6.0436,
      "mean_token_accuracy": 0.1729697972536087,
      "num_tokens": 41416698.0,
      "step": 16350
    },
    {
      "entropy": 6.213017225265503,
      "epoch": 1.8874783612233121,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00046525186864990073,
      "loss": 6.0148,
      "mean_token_accuracy": 0.17483233362436296,
      "num_tokens": 41429953.0,
      "step": 16355
    },
    {
      "entropy": 6.182444763183594,
      "epoch": 1.8880553952683208,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046522983547396806,
      "loss": 6.0384,
      "mean_token_accuracy": 0.16858932971954346,
      "num_tokens": 41442824.0,
      "step": 16360
    },
    {
      "entropy": 6.258214044570923,
      "epoch": 1.8886324293133296,
      "grad_norm": 1.859375,
      "learning_rate": 0.0004652077958997494,
      "loss": 6.0562,
      "mean_token_accuracy": 0.17258467972278596,
      "num_tokens": 41455062.0,
      "step": 16365
    },
    {
      "entropy": 6.214363622665405,
      "epoch": 1.889209463358338,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046518574992798604,
      "loss": 5.9594,
      "mean_token_accuracy": 0.18245875984430313,
      "num_tokens": 41468147.0,
      "step": 16370
    },
    {
      "entropy": 6.234375715255737,
      "epoch": 1.889786497403347,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046516369755941945,
      "loss": 5.9635,
      "mean_token_accuracy": 0.1727774828672409,
      "num_tokens": 41480664.0,
      "step": 16375
    },
    {
      "entropy": 6.266945314407349,
      "epoch": 1.8903635314483553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004651416387947914,
      "loss": 6.0572,
      "mean_token_accuracy": 0.16870816498994828,
      "num_tokens": 41494578.0,
      "step": 16380
    },
    {
      "entropy": 6.309160470962524,
      "epoch": 1.8909405654933642,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004651195736348437,
      "loss": 6.0935,
      "mean_token_accuracy": 0.1649898260831833,
      "num_tokens": 41506528.0,
      "step": 16385
    },
    {
      "entropy": 6.325799036026001,
      "epoch": 1.8915175995383726,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004650975020803186,
      "loss": 6.1549,
      "mean_token_accuracy": 0.16303233057260513,
      "num_tokens": 41520861.0,
      "step": 16390
    },
    {
      "entropy": 6.268583345413208,
      "epoch": 1.8920946335833815,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004650754241319584,
      "loss": 6.0208,
      "mean_token_accuracy": 0.17048420161008834,
      "num_tokens": 41533811.0,
      "step": 16395
    },
    {
      "entropy": 6.324508285522461,
      "epoch": 1.8926716676283901,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046505333979050573,
      "loss": 6.0909,
      "mean_token_accuracy": 0.16676845252513886,
      "num_tokens": 41546291.0,
      "step": 16400
    },
    {
      "entropy": 6.226980972290039,
      "epoch": 1.8932487016733988,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004650312490567035,
      "loss": 5.9405,
      "mean_token_accuracy": 0.18085718005895615,
      "num_tokens": 41558157.0,
      "step": 16405
    },
    {
      "entropy": 6.266756439208985,
      "epoch": 1.8938257357184074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004650091519312945,
      "loss": 6.0759,
      "mean_token_accuracy": 0.17133705019950868,
      "num_tokens": 41571346.0,
      "step": 16410
    },
    {
      "entropy": 6.298261976242065,
      "epoch": 1.894402769763416,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046498704841502203,
      "loss": 6.1362,
      "mean_token_accuracy": 0.16385228782892228,
      "num_tokens": 41583973.0,
      "step": 16415
    },
    {
      "entropy": 6.292732286453247,
      "epoch": 1.8949798038084247,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004649649385086296,
      "loss": 6.0488,
      "mean_token_accuracy": 0.16712913364171983,
      "num_tokens": 41596210.0,
      "step": 16420
    },
    {
      "entropy": 6.289768648147583,
      "epoch": 1.8955568378534333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004649428222128608,
      "loss": 6.1271,
      "mean_token_accuracy": 0.16882113367319107,
      "num_tokens": 41608297.0,
      "step": 16425
    },
    {
      "entropy": 6.313119888305664,
      "epoch": 1.896133871898442,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046492069952845953,
      "loss": 6.0681,
      "mean_token_accuracy": 0.1669365108013153,
      "num_tokens": 41621319.0,
      "step": 16430
    },
    {
      "entropy": 6.320606660842896,
      "epoch": 1.8967109059434506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046489857045617,
      "loss": 5.9629,
      "mean_token_accuracy": 0.17546855211257933,
      "num_tokens": 41633282.0,
      "step": 16435
    },
    {
      "entropy": 6.16384072303772,
      "epoch": 1.8972879399884595,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046487643499673625,
      "loss": 6.0354,
      "mean_token_accuracy": 0.16361497044563295,
      "num_tokens": 41646573.0,
      "step": 16440
    },
    {
      "entropy": 6.252078866958618,
      "epoch": 1.8978649740334679,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648542931509029,
      "loss": 5.999,
      "mean_token_accuracy": 0.17221661657094955,
      "num_tokens": 41659596.0,
      "step": 16445
    },
    {
      "entropy": 6.289477062225342,
      "epoch": 1.8984420080784767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004648321449194148,
      "loss": 6.0247,
      "mean_token_accuracy": 0.16427045613527297,
      "num_tokens": 41673612.0,
      "step": 16450
    },
    {
      "entropy": 6.313947153091431,
      "epoch": 1.8990190421234852,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046480999030301673,
      "loss": 6.1183,
      "mean_token_accuracy": 0.1691570982336998,
      "num_tokens": 41686583.0,
      "step": 16455
    },
    {
      "entropy": 6.2747705459594725,
      "epoch": 1.899596076168494,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046478782930245395,
      "loss": 6.0277,
      "mean_token_accuracy": 0.16731169670820237,
      "num_tokens": 41698542.0,
      "step": 16460
    },
    {
      "entropy": 6.274620056152344,
      "epoch": 1.9001731102135024,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046476566191847176,
      "loss": 6.0712,
      "mean_token_accuracy": 0.16262018978595733,
      "num_tokens": 41712461.0,
      "step": 16465
    },
    {
      "entropy": 6.319371175765991,
      "epoch": 1.9007501442585113,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004647434881518159,
      "loss": 6.0762,
      "mean_token_accuracy": 0.16481758207082747,
      "num_tokens": 41725450.0,
      "step": 16470
    },
    {
      "entropy": 6.324736642837524,
      "epoch": 1.90132717830352,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046472130800323194,
      "loss": 6.0425,
      "mean_token_accuracy": 0.17140338122844695,
      "num_tokens": 41737255.0,
      "step": 16475
    },
    {
      "entropy": 6.234397840499878,
      "epoch": 1.9019042123485286,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004646991214734661,
      "loss": 6.0915,
      "mean_token_accuracy": 0.1634446457028389,
      "num_tokens": 41749459.0,
      "step": 16480
    },
    {
      "entropy": 6.248945045471191,
      "epoch": 1.9024812463935372,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004646769285632645,
      "loss": 6.0365,
      "mean_token_accuracy": 0.17080003917217254,
      "num_tokens": 41761603.0,
      "step": 16485
    },
    {
      "entropy": 6.2919713973999025,
      "epoch": 1.9030582804385459,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004646547292733737,
      "loss": 6.1242,
      "mean_token_accuracy": 0.16482744812965394,
      "num_tokens": 41774865.0,
      "step": 16490
    },
    {
      "entropy": 6.240364503860474,
      "epoch": 1.9036353144835545,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004646325236045402,
      "loss": 6.0715,
      "mean_token_accuracy": 0.17352937161922455,
      "num_tokens": 41787558.0,
      "step": 16495
    },
    {
      "entropy": 6.2833081722259525,
      "epoch": 1.9042123485285631,
      "grad_norm": 1.234375,
      "learning_rate": 0.000464610311557511,
      "loss": 5.9958,
      "mean_token_accuracy": 0.1723578542470932,
      "num_tokens": 41800754.0,
      "step": 16500
    },
    {
      "entropy": 6.26614818572998,
      "epoch": 1.904789382573572,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004645880931330331,
      "loss": 5.9604,
      "mean_token_accuracy": 0.1790569856762886,
      "num_tokens": 41813133.0,
      "step": 16505
    },
    {
      "entropy": 6.22602105140686,
      "epoch": 1.9053664166185804,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004645658683318539,
      "loss": 6.0363,
      "mean_token_accuracy": 0.16803948283195497,
      "num_tokens": 41825288.0,
      "step": 16510
    },
    {
      "entropy": 6.312449026107788,
      "epoch": 1.9059434506635893,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004645436371547209,
      "loss": 6.1145,
      "mean_token_accuracy": 0.1666354864835739,
      "num_tokens": 41837331.0,
      "step": 16515
    },
    {
      "entropy": 6.313289451599121,
      "epoch": 1.9065204847085977,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046452139960238186,
      "loss": 6.0976,
      "mean_token_accuracy": 0.170522104203701,
      "num_tokens": 41850536.0,
      "step": 16520
    },
    {
      "entropy": 6.264101219177246,
      "epoch": 1.9070975187536066,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046449915567558467,
      "loss": 5.9937,
      "mean_token_accuracy": 0.17788014858961104,
      "num_tokens": 41862199.0,
      "step": 16525
    },
    {
      "entropy": 6.239650011062622,
      "epoch": 1.907674552798615,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004644769053750775,
      "loss": 6.0172,
      "mean_token_accuracy": 0.17359801530838012,
      "num_tokens": 41872945.0,
      "step": 16530
    },
    {
      "entropy": 6.2783403396606445,
      "epoch": 1.9082515868436238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004644546487016087,
      "loss": 5.9911,
      "mean_token_accuracy": 0.1715606838464737,
      "num_tokens": 41884223.0,
      "step": 16535
    },
    {
      "entropy": 6.246064186096191,
      "epoch": 1.9088286208886325,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046443238565592687,
      "loss": 6.0527,
      "mean_token_accuracy": 0.16985336393117906,
      "num_tokens": 41896997.0,
      "step": 16540
    },
    {
      "entropy": 6.231339931488037,
      "epoch": 1.9094056549336411,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00046441011623878087,
      "loss": 6.0796,
      "mean_token_accuracy": 0.16984277367591857,
      "num_tokens": 41910210.0,
      "step": 16545
    },
    {
      "entropy": 6.087150573730469,
      "epoch": 1.9099826889786498,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004643878404509197,
      "loss": 5.8959,
      "mean_token_accuracy": 0.18043217658996583,
      "num_tokens": 41922882.0,
      "step": 16550
    },
    {
      "entropy": 6.229480791091919,
      "epoch": 1.9105597230236584,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046436555829309264,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1844248279929161,
      "num_tokens": 41934655.0,
      "step": 16555
    },
    {
      "entropy": 6.253088045120239,
      "epoch": 1.911136757068667,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000464343269766049,
      "loss": 6.04,
      "mean_token_accuracy": 0.16934573352336885,
      "num_tokens": 41947946.0,
      "step": 16560
    },
    {
      "entropy": 6.3426737785339355,
      "epoch": 1.9117137911136757,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046432097487053857,
      "loss": 6.1505,
      "mean_token_accuracy": 0.16838859617710114,
      "num_tokens": 41962316.0,
      "step": 16565
    },
    {
      "entropy": 6.282298374176025,
      "epoch": 1.9122908251586843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046429867360731124,
      "loss": 6.0264,
      "mean_token_accuracy": 0.17507038414478301,
      "num_tokens": 41976076.0,
      "step": 16570
    },
    {
      "entropy": 6.328230476379394,
      "epoch": 1.912867859203693,
      "grad_norm": 2.0625,
      "learning_rate": 0.00046427636597711695,
      "loss": 6.0817,
      "mean_token_accuracy": 0.16239470541477202,
      "num_tokens": 41988762.0,
      "step": 16575
    },
    {
      "entropy": 6.192126226425171,
      "epoch": 1.9134448932487018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046425405198070624,
      "loss": 5.9454,
      "mean_token_accuracy": 0.17111702710390092,
      "num_tokens": 42000525.0,
      "step": 16580
    },
    {
      "entropy": 6.195065450668335,
      "epoch": 1.9140219272937102,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046423173161882944,
      "loss": 5.9222,
      "mean_token_accuracy": 0.17092446684837342,
      "num_tokens": 42013716.0,
      "step": 16585
    },
    {
      "entropy": 6.277429103851318,
      "epoch": 1.914598961338719,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046420940489223735,
      "loss": 6.0313,
      "mean_token_accuracy": 0.17426891773939132,
      "num_tokens": 42026689.0,
      "step": 16590
    },
    {
      "entropy": 6.289929008483886,
      "epoch": 1.9151759953837275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004641870718016809,
      "loss": 6.0947,
      "mean_token_accuracy": 0.16388041228055955,
      "num_tokens": 42039907.0,
      "step": 16595
    },
    {
      "entropy": 6.276375436782837,
      "epoch": 1.9157530294287364,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004641647323479113,
      "loss": 6.0451,
      "mean_token_accuracy": 0.173135207593441,
      "num_tokens": 42052781.0,
      "step": 16600
    },
    {
      "entropy": 6.243412160873413,
      "epoch": 1.9163300634737448,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004641423865316798,
      "loss": 6.0177,
      "mean_token_accuracy": 0.17488401681184768,
      "num_tokens": 42066695.0,
      "step": 16605
    },
    {
      "entropy": 6.1937644481658936,
      "epoch": 1.9169070975187537,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004641200343537381,
      "loss": 5.941,
      "mean_token_accuracy": 0.17501333504915237,
      "num_tokens": 42079315.0,
      "step": 16610
    },
    {
      "entropy": 6.265763568878174,
      "epoch": 1.9174841315637623,
      "grad_norm": 1.109375,
      "learning_rate": 0.000464097675814838,
      "loss": 6.058,
      "mean_token_accuracy": 0.16800168603658677,
      "num_tokens": 42091915.0,
      "step": 16615
    },
    {
      "entropy": 6.3285560131073,
      "epoch": 1.918061165608771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004640753109157316,
      "loss": 6.1058,
      "mean_token_accuracy": 0.16329863667488098,
      "num_tokens": 42105316.0,
      "step": 16620
    },
    {
      "entropy": 6.2566338062286375,
      "epoch": 1.9186381996537796,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046405293965717093,
      "loss": 5.9762,
      "mean_token_accuracy": 0.1756441429257393,
      "num_tokens": 42118078.0,
      "step": 16625
    },
    {
      "entropy": 6.134314250946045,
      "epoch": 1.9192152336987882,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004640305620399086,
      "loss": 5.961,
      "mean_token_accuracy": 0.18055089563131332,
      "num_tokens": 42130845.0,
      "step": 16630
    },
    {
      "entropy": 6.211712741851807,
      "epoch": 1.9197922677437969,
      "grad_norm": 1.125,
      "learning_rate": 0.0004640081780646971,
      "loss": 6.0069,
      "mean_token_accuracy": 0.17803016901016236,
      "num_tokens": 42142487.0,
      "step": 16635
    },
    {
      "entropy": 6.216660976409912,
      "epoch": 1.9203693017888055,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046398578773228954,
      "loss": 6.0323,
      "mean_token_accuracy": 0.17111165076494217,
      "num_tokens": 42154747.0,
      "step": 16640
    },
    {
      "entropy": 6.373775386810303,
      "epoch": 1.9209463358338144,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046396339104343886,
      "loss": 6.0687,
      "mean_token_accuracy": 0.1637737326323986,
      "num_tokens": 42167697.0,
      "step": 16645
    },
    {
      "entropy": 6.2525794506073,
      "epoch": 1.9215233698788228,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004639409879988984,
      "loss": 6.0448,
      "mean_token_accuracy": 0.17294813096523284,
      "num_tokens": 42179874.0,
      "step": 16650
    },
    {
      "entropy": 6.323226881027222,
      "epoch": 1.9221004039238316,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004639185785994216,
      "loss": 6.1124,
      "mean_token_accuracy": 0.1681394025683403,
      "num_tokens": 42192747.0,
      "step": 16655
    },
    {
      "entropy": 6.206807041168213,
      "epoch": 1.92267743796884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046389616284576226,
      "loss": 6.0361,
      "mean_token_accuracy": 0.16698787957429886,
      "num_tokens": 42205534.0,
      "step": 16660
    },
    {
      "entropy": 6.251869058609008,
      "epoch": 1.923254472013849,
      "grad_norm": 2.609375,
      "learning_rate": 0.00046387374073867435,
      "loss": 6.044,
      "mean_token_accuracy": 0.17046794891357422,
      "num_tokens": 42217825.0,
      "step": 16665
    },
    {
      "entropy": 6.248140335083008,
      "epoch": 1.9238315060588573,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046385131227891197,
      "loss": 6.0233,
      "mean_token_accuracy": 0.17224513739347458,
      "num_tokens": 42230162.0,
      "step": 16670
    },
    {
      "entropy": 6.189965867996216,
      "epoch": 1.9244085401038662,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004638288774672295,
      "loss": 6.0039,
      "mean_token_accuracy": 0.17534219622612,
      "num_tokens": 42242458.0,
      "step": 16675
    },
    {
      "entropy": 6.134315395355225,
      "epoch": 1.9249855741488748,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004638064363043816,
      "loss": 5.9364,
      "mean_token_accuracy": 0.17521344423294066,
      "num_tokens": 42254389.0,
      "step": 16680
    },
    {
      "entropy": 6.185854196548462,
      "epoch": 1.9255626081938835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046378398879112293,
      "loss": 6.0245,
      "mean_token_accuracy": 0.17561282962560654,
      "num_tokens": 42266421.0,
      "step": 16685
    },
    {
      "entropy": 6.145219898223877,
      "epoch": 1.9261396422388921,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004637615349282086,
      "loss": 5.9901,
      "mean_token_accuracy": 0.17165957987308503,
      "num_tokens": 42279189.0,
      "step": 16690
    },
    {
      "entropy": 6.226276779174805,
      "epoch": 1.9267166762839008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004637390747163938,
      "loss": 6.0187,
      "mean_token_accuracy": 0.17052838951349258,
      "num_tokens": 42291970.0,
      "step": 16695
    },
    {
      "entropy": 6.288958930969239,
      "epoch": 1.9272937103289094,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004637166081564339,
      "loss": 6.1009,
      "mean_token_accuracy": 0.16882906556129457,
      "num_tokens": 42305185.0,
      "step": 16700
    },
    {
      "entropy": 6.345633840560913,
      "epoch": 1.927870744373918,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046369413524908473,
      "loss": 6.0885,
      "mean_token_accuracy": 0.16870655864477158,
      "num_tokens": 42317823.0,
      "step": 16705
    },
    {
      "entropy": 6.186788845062256,
      "epoch": 1.9284477784189267,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046367165599510207,
      "loss": 6.0494,
      "mean_token_accuracy": 0.17140022963285445,
      "num_tokens": 42330540.0,
      "step": 16710
    },
    {
      "entropy": 6.201239490509034,
      "epoch": 1.9290248124639353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004636491703952419,
      "loss": 5.8622,
      "mean_token_accuracy": 0.18321040272712708,
      "num_tokens": 42342347.0,
      "step": 16715
    },
    {
      "entropy": 6.2329206466674805,
      "epoch": 1.9296018465089442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046362667845026057,
      "loss": 6.0089,
      "mean_token_accuracy": 0.17573418319225312,
      "num_tokens": 42355004.0,
      "step": 16720
    },
    {
      "entropy": 6.245389366149903,
      "epoch": 1.9301788805539526,
      "grad_norm": 1.25,
      "learning_rate": 0.00046360418016091467,
      "loss": 5.9942,
      "mean_token_accuracy": 0.17386723011732103,
      "num_tokens": 42366700.0,
      "step": 16725
    },
    {
      "entropy": 6.2483916759490965,
      "epoch": 1.9307559145989615,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046358167552796085,
      "loss": 6.0717,
      "mean_token_accuracy": 0.16806585043668748,
      "num_tokens": 42379723.0,
      "step": 16730
    },
    {
      "entropy": 6.20191330909729,
      "epoch": 1.9313329486439699,
      "grad_norm": 1.125,
      "learning_rate": 0.00046355916455215597,
      "loss": 5.9851,
      "mean_token_accuracy": 0.18643948137760163,
      "num_tokens": 42393505.0,
      "step": 16735
    },
    {
      "entropy": 6.259112405776977,
      "epoch": 1.9319099826889787,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004635366472342573,
      "loss": 6.003,
      "mean_token_accuracy": 0.164619180560112,
      "num_tokens": 42406334.0,
      "step": 16740
    },
    {
      "entropy": 6.253616046905518,
      "epoch": 1.9324870167339872,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004635141235750222,
      "loss": 6.022,
      "mean_token_accuracy": 0.1686533898115158,
      "num_tokens": 42418585.0,
      "step": 16745
    },
    {
      "entropy": 6.243710851669311,
      "epoch": 1.933064050778996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046349159357520815,
      "loss": 5.98,
      "mean_token_accuracy": 0.18272491544485092,
      "num_tokens": 42430741.0,
      "step": 16750
    },
    {
      "entropy": 6.187639904022217,
      "epoch": 1.9336410848240047,
      "grad_norm": 1.203125,
      "learning_rate": 0.000463469057235573,
      "loss": 5.9783,
      "mean_token_accuracy": 0.17807655930519103,
      "num_tokens": 42443241.0,
      "step": 16755
    },
    {
      "entropy": 6.195738410949707,
      "epoch": 1.9342181188690133,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046344651455687476,
      "loss": 6.0625,
      "mean_token_accuracy": 0.17015268355607988,
      "num_tokens": 42456476.0,
      "step": 16760
    },
    {
      "entropy": 6.336266040802002,
      "epoch": 1.934795152914022,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046342396553987155,
      "loss": 6.0827,
      "mean_token_accuracy": 0.16396595537662506,
      "num_tokens": 42468875.0,
      "step": 16765
    },
    {
      "entropy": 6.288597249984742,
      "epoch": 1.9353721869590306,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046340141018532186,
      "loss": 6.0448,
      "mean_token_accuracy": 0.16845245212316512,
      "num_tokens": 42483300.0,
      "step": 16770
    },
    {
      "entropy": 6.210137605667114,
      "epoch": 1.9359492210040392,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004633788484939843,
      "loss": 6.0626,
      "mean_token_accuracy": 0.17025937288999557,
      "num_tokens": 42497748.0,
      "step": 16775
    },
    {
      "entropy": 6.2557402610778805,
      "epoch": 1.9365262550490479,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046335628046661776,
      "loss": 6.0758,
      "mean_token_accuracy": 0.1684375673532486,
      "num_tokens": 42510464.0,
      "step": 16780
    },
    {
      "entropy": 6.335916376113891,
      "epoch": 1.9371032890940567,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046333370610398135,
      "loss": 6.094,
      "mean_token_accuracy": 0.16555423587560653,
      "num_tokens": 42522298.0,
      "step": 16785
    },
    {
      "entropy": 6.318788480758667,
      "epoch": 1.9376803231390651,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004633111254068342,
      "loss": 6.0237,
      "mean_token_accuracy": 0.16976431906223297,
      "num_tokens": 42535094.0,
      "step": 16790
    },
    {
      "entropy": 6.217743253707885,
      "epoch": 1.938257357184074,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004632885383759359,
      "loss": 5.9903,
      "mean_token_accuracy": 0.17622564733028412,
      "num_tokens": 42547272.0,
      "step": 16795
    },
    {
      "entropy": 6.130745124816895,
      "epoch": 1.9388343912290824,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046326594501204624,
      "loss": 5.8878,
      "mean_token_accuracy": 0.1842793568968773,
      "num_tokens": 42559568.0,
      "step": 16800
    },
    {
      "entropy": 6.2765467166900635,
      "epoch": 1.9394114252740913,
      "grad_norm": 1.28125,
      "learning_rate": 0.000463243345315925,
      "loss": 6.0805,
      "mean_token_accuracy": 0.16390738338232042,
      "num_tokens": 42573177.0,
      "step": 16805
    },
    {
      "entropy": 6.263757228851318,
      "epoch": 1.9399884593190997,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046322073928833224,
      "loss": 5.9897,
      "mean_token_accuracy": 0.17142129838466644,
      "num_tokens": 42585436.0,
      "step": 16810
    },
    {
      "entropy": 6.274092197418213,
      "epoch": 1.9405654933641086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004631981269300285,
      "loss": 6.066,
      "mean_token_accuracy": 0.16510260999202728,
      "num_tokens": 42599826.0,
      "step": 16815
    },
    {
      "entropy": 6.223218488693237,
      "epoch": 1.9411425274091172,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004631755082417742,
      "loss": 6.0508,
      "mean_token_accuracy": 0.1671837419271469,
      "num_tokens": 42613901.0,
      "step": 16820
    },
    {
      "entropy": 6.325699520111084,
      "epoch": 1.9417195614541258,
      "grad_norm": 1.875,
      "learning_rate": 0.0004631528832243302,
      "loss": 6.0685,
      "mean_token_accuracy": 0.1655432403087616,
      "num_tokens": 42626592.0,
      "step": 16825
    },
    {
      "entropy": 6.229167652130127,
      "epoch": 1.9422965954991345,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046313025187845735,
      "loss": 5.9619,
      "mean_token_accuracy": 0.17458304166793823,
      "num_tokens": 42639754.0,
      "step": 16830
    },
    {
      "entropy": 6.262476778030395,
      "epoch": 1.942873629544143,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046310761420491705,
      "loss": 5.9884,
      "mean_token_accuracy": 0.16979699730873107,
      "num_tokens": 42651321.0,
      "step": 16835
    },
    {
      "entropy": 6.246843433380127,
      "epoch": 1.9434506635891518,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004630849702044705,
      "loss": 6.0391,
      "mean_token_accuracy": 0.17260385304689407,
      "num_tokens": 42664188.0,
      "step": 16840
    },
    {
      "entropy": 6.198481512069702,
      "epoch": 1.9440276976341604,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046306231987787937,
      "loss": 6.0095,
      "mean_token_accuracy": 0.1676759049296379,
      "num_tokens": 42677882.0,
      "step": 16845
    },
    {
      "entropy": 6.212922239303589,
      "epoch": 1.944604731679169,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046303966322590556,
      "loss": 5.9278,
      "mean_token_accuracy": 0.17638970911502838,
      "num_tokens": 42690595.0,
      "step": 16850
    },
    {
      "entropy": 6.2258812427520756,
      "epoch": 1.9451817657241777,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004630170002493111,
      "loss": 5.9277,
      "mean_token_accuracy": 0.17367104142904283,
      "num_tokens": 42704784.0,
      "step": 16855
    },
    {
      "entropy": 6.168443441390991,
      "epoch": 1.9457587997691865,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046299433094885826,
      "loss": 5.9728,
      "mean_token_accuracy": 0.17192557603120803,
      "num_tokens": 42717012.0,
      "step": 16860
    },
    {
      "entropy": 6.279626703262329,
      "epoch": 1.946335833814195,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046297165532530944,
      "loss": 6.0293,
      "mean_token_accuracy": 0.16916215866804124,
      "num_tokens": 42728879.0,
      "step": 16865
    },
    {
      "entropy": 6.295999956130982,
      "epoch": 1.9469128678592038,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004629489733794274,
      "loss": 6.076,
      "mean_token_accuracy": 0.16610499918460847,
      "num_tokens": 42742688.0,
      "step": 16870
    },
    {
      "entropy": 6.300163602828979,
      "epoch": 1.9474899019042122,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000462926285111975,
      "loss": 6.0136,
      "mean_token_accuracy": 0.17265246957540512,
      "num_tokens": 42754955.0,
      "step": 16875
    },
    {
      "entropy": 6.188889646530152,
      "epoch": 1.948066935949221,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046290359052371535,
      "loss": 6.0522,
      "mean_token_accuracy": 0.16892243772745133,
      "num_tokens": 42767760.0,
      "step": 16880
    },
    {
      "entropy": 6.246478271484375,
      "epoch": 1.9486439699942295,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004628808896154117,
      "loss": 6.0019,
      "mean_token_accuracy": 0.16956818401813506,
      "num_tokens": 42779683.0,
      "step": 16885
    },
    {
      "entropy": 6.2639851570129395,
      "epoch": 1.9492210040392384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004628581823878277,
      "loss": 6.0136,
      "mean_token_accuracy": 0.16905308216810228,
      "num_tokens": 42791289.0,
      "step": 16890
    },
    {
      "entropy": 6.233004856109619,
      "epoch": 1.949798038084247,
      "grad_norm": 1.40625,
      "learning_rate": 0.000462835468841727,
      "loss": 6.0669,
      "mean_token_accuracy": 0.17280749827623368,
      "num_tokens": 42803318.0,
      "step": 16895
    },
    {
      "entropy": 6.268746852874756,
      "epoch": 1.9503750721292556,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004628127489778737,
      "loss": 6.0779,
      "mean_token_accuracy": 0.1690628483891487,
      "num_tokens": 42815935.0,
      "step": 16900
    },
    {
      "entropy": 6.278210783004761,
      "epoch": 1.9509521061742643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004627900227970318,
      "loss": 6.0285,
      "mean_token_accuracy": 0.1742333874106407,
      "num_tokens": 42827759.0,
      "step": 16905
    },
    {
      "entropy": 6.227267503738403,
      "epoch": 1.951529140219273,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046276729029996576,
      "loss": 6.003,
      "mean_token_accuracy": 0.17346829771995545,
      "num_tokens": 42840479.0,
      "step": 16910
    },
    {
      "entropy": 6.2318909645080565,
      "epoch": 1.9521061742642816,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046274455148744025,
      "loss": 6.0952,
      "mean_token_accuracy": 0.1659419283270836,
      "num_tokens": 42853972.0,
      "step": 16915
    },
    {
      "entropy": 6.221674251556396,
      "epoch": 1.9526832083092902,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046272180636022,
      "loss": 6.0415,
      "mean_token_accuracy": 0.16805387735366822,
      "num_tokens": 42866192.0,
      "step": 16920
    },
    {
      "entropy": 6.2549063205719,
      "epoch": 1.953260242354299,
      "grad_norm": 1.421875,
      "learning_rate": 0.00046269905491907,
      "loss": 5.9968,
      "mean_token_accuracy": 0.17463415563106538,
      "num_tokens": 42878661.0,
      "step": 16925
    },
    {
      "entropy": 6.230676937103271,
      "epoch": 1.9538372763993075,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004626762971647555,
      "loss": 6.0225,
      "mean_token_accuracy": 0.17351190000772476,
      "num_tokens": 42891498.0,
      "step": 16930
    },
    {
      "entropy": 6.232025623321533,
      "epoch": 1.9544143104443163,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046265353309804205,
      "loss": 5.979,
      "mean_token_accuracy": 0.1768836870789528,
      "num_tokens": 42904657.0,
      "step": 16935
    },
    {
      "entropy": 6.22298035621643,
      "epoch": 1.9549913444893248,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004626307627196952,
      "loss": 5.9937,
      "mean_token_accuracy": 0.17455218583345414,
      "num_tokens": 42916824.0,
      "step": 16940
    },
    {
      "entropy": 6.279146289825439,
      "epoch": 1.9555683785343336,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004626079860304808,
      "loss": 6.053,
      "mean_token_accuracy": 0.17725331634283065,
      "num_tokens": 42929366.0,
      "step": 16945
    },
    {
      "entropy": 6.147270822525025,
      "epoch": 1.956145412579342,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046258520303116496,
      "loss": 5.9108,
      "mean_token_accuracy": 0.18031210005283355,
      "num_tokens": 42941955.0,
      "step": 16950
    },
    {
      "entropy": 6.272320032119751,
      "epoch": 1.956722446624351,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004625624137225141,
      "loss": 6.0865,
      "mean_token_accuracy": 0.166519957780838,
      "num_tokens": 42954302.0,
      "step": 16955
    },
    {
      "entropy": 6.343348264694214,
      "epoch": 1.9572994806693593,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004625396181052945,
      "loss": 6.0364,
      "mean_token_accuracy": 0.16592449694871902,
      "num_tokens": 42967225.0,
      "step": 16960
    },
    {
      "entropy": 6.157590579986572,
      "epoch": 1.9578765147143682,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046251681618027316,
      "loss": 6.0209,
      "mean_token_accuracy": 0.1752626806497574,
      "num_tokens": 42981388.0,
      "step": 16965
    },
    {
      "entropy": 6.245779943466187,
      "epoch": 1.9584535487593768,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004624940079482167,
      "loss": 6.0451,
      "mean_token_accuracy": 0.17435730695724488,
      "num_tokens": 42993110.0,
      "step": 16970
    },
    {
      "entropy": 6.225678825378418,
      "epoch": 1.9590305828043855,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046247119340989254,
      "loss": 6.0588,
      "mean_token_accuracy": 0.16311826258897782,
      "num_tokens": 43007309.0,
      "step": 16975
    },
    {
      "entropy": 6.3057647228240965,
      "epoch": 1.959607616849394,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004624483725660678,
      "loss": 6.0006,
      "mean_token_accuracy": 0.17937014997005463,
      "num_tokens": 43020813.0,
      "step": 16980
    },
    {
      "entropy": 6.267711973190307,
      "epoch": 1.9601846508944027,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004624255454175102,
      "loss": 6.0235,
      "mean_token_accuracy": 0.1743055358529091,
      "num_tokens": 43033357.0,
      "step": 16985
    },
    {
      "entropy": 6.227343893051147,
      "epoch": 1.9607616849394114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004624027119649875,
      "loss": 6.1083,
      "mean_token_accuracy": 0.16687651872634887,
      "num_tokens": 43046685.0,
      "step": 16990
    },
    {
      "entropy": 6.339546060562133,
      "epoch": 1.96133871898442,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046237987220926766,
      "loss": 6.0799,
      "mean_token_accuracy": 0.1672934964299202,
      "num_tokens": 43059670.0,
      "step": 16995
    },
    {
      "entropy": 6.22210693359375,
      "epoch": 1.9619157530294289,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046235702615111886,
      "loss": 5.9468,
      "mean_token_accuracy": 0.17477399855852127,
      "num_tokens": 43073515.0,
      "step": 17000
    },
    {
      "entropy": 6.229201745986939,
      "epoch": 1.9624927870744373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004623341737913096,
      "loss": 6.0422,
      "mean_token_accuracy": 0.17302633225917816,
      "num_tokens": 43086046.0,
      "step": 17005
    },
    {
      "entropy": 6.224934053421021,
      "epoch": 1.9630698211194462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004623113151306085,
      "loss": 5.9269,
      "mean_token_accuracy": 0.17834639847278594,
      "num_tokens": 43097703.0,
      "step": 17010
    },
    {
      "entropy": 6.266271162033081,
      "epoch": 1.9636468551644546,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046228845016978425,
      "loss": 6.0212,
      "mean_token_accuracy": 0.16586049050092697,
      "num_tokens": 43110378.0,
      "step": 17015
    },
    {
      "entropy": 6.296831846237183,
      "epoch": 1.9642238892094634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004622655789096061,
      "loss": 6.042,
      "mean_token_accuracy": 0.1670925036072731,
      "num_tokens": 43122978.0,
      "step": 17020
    },
    {
      "entropy": 6.284532690048218,
      "epoch": 1.9648009232544719,
      "grad_norm": 2.0,
      "learning_rate": 0.00046224270135084315,
      "loss": 6.0846,
      "mean_token_accuracy": 0.1697956383228302,
      "num_tokens": 43135371.0,
      "step": 17025
    },
    {
      "entropy": 6.197793197631836,
      "epoch": 1.9653779572994807,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046221981749426503,
      "loss": 5.9921,
      "mean_token_accuracy": 0.1699984058737755,
      "num_tokens": 43147251.0,
      "step": 17030
    },
    {
      "entropy": 6.269783020019531,
      "epoch": 1.9659549913444894,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046219692734064124,
      "loss": 5.9799,
      "mean_token_accuracy": 0.16827934384346008,
      "num_tokens": 43160276.0,
      "step": 17035
    },
    {
      "entropy": 6.286127424240112,
      "epoch": 1.966532025389498,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004621740308907419,
      "loss": 6.038,
      "mean_token_accuracy": 0.17274793982505798,
      "num_tokens": 43172443.0,
      "step": 17040
    },
    {
      "entropy": 6.257483434677124,
      "epoch": 1.9671090594345066,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004621511281453369,
      "loss": 6.0584,
      "mean_token_accuracy": 0.16886429041624068,
      "num_tokens": 43186350.0,
      "step": 17045
    },
    {
      "entropy": 6.218119812011719,
      "epoch": 1.9676860934795153,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004621282191051967,
      "loss": 5.9859,
      "mean_token_accuracy": 0.17662974447011948,
      "num_tokens": 43198936.0,
      "step": 17050
    },
    {
      "entropy": 6.324996995925903,
      "epoch": 1.968263127524524,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004621053037710918,
      "loss": 6.1302,
      "mean_token_accuracy": 0.1648729309439659,
      "num_tokens": 43211793.0,
      "step": 17055
    },
    {
      "entropy": 6.251626491546631,
      "epoch": 1.9688401615695326,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462082382143793,
      "loss": 5.9818,
      "mean_token_accuracy": 0.17532301992177962,
      "num_tokens": 43223817.0,
      "step": 17060
    },
    {
      "entropy": 6.19627890586853,
      "epoch": 1.9694171956145412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046205945422407113,
      "loss": 6.0233,
      "mean_token_accuracy": 0.17594825774431228,
      "num_tokens": 43235769.0,
      "step": 17065
    },
    {
      "entropy": 6.22838978767395,
      "epoch": 1.9699942296595498,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046203652001269754,
      "loss": 6.037,
      "mean_token_accuracy": 0.17000845670700074,
      "num_tokens": 43249309.0,
      "step": 17070
    },
    {
      "entropy": 6.279005813598633,
      "epoch": 1.9705712637045587,
      "grad_norm": 1.9296875,
      "learning_rate": 0.00046201357951044337,
      "loss": 6.0119,
      "mean_token_accuracy": 0.1684460759162903,
      "num_tokens": 43262814.0,
      "step": 17075
    },
    {
      "entropy": 6.315310621261597,
      "epoch": 1.9711482977495671,
      "grad_norm": 1.9375,
      "learning_rate": 0.00046199063271808047,
      "loss": 6.1099,
      "mean_token_accuracy": 0.16198794841766356,
      "num_tokens": 43276898.0,
      "step": 17080
    },
    {
      "entropy": 6.245305490493775,
      "epoch": 1.971725331794576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004619676796363804,
      "loss": 5.9942,
      "mean_token_accuracy": 0.17359111309051514,
      "num_tokens": 43288902.0,
      "step": 17085
    },
    {
      "entropy": 6.257856369018555,
      "epoch": 1.9723023658395844,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046194472026611546,
      "loss": 6.0347,
      "mean_token_accuracy": 0.17116419225931168,
      "num_tokens": 43301644.0,
      "step": 17090
    },
    {
      "entropy": 6.231047248840332,
      "epoch": 1.9728793998845933,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004619217546080576,
      "loss": 6.024,
      "mean_token_accuracy": 0.1758951172232628,
      "num_tokens": 43315550.0,
      "step": 17095
    },
    {
      "entropy": 6.350954198837281,
      "epoch": 1.9734564339296017,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004618987826629794,
      "loss": 6.1074,
      "mean_token_accuracy": 0.16367050707340242,
      "num_tokens": 43328095.0,
      "step": 17100
    },
    {
      "entropy": 6.213918876647949,
      "epoch": 1.9740334679746105,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046187580443165344,
      "loss": 6.0191,
      "mean_token_accuracy": 0.17428866624832154,
      "num_tokens": 43340461.0,
      "step": 17105
    },
    {
      "entropy": 6.293837833404541,
      "epoch": 1.9746105020196192,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004618528199148527,
      "loss": 5.9914,
      "mean_token_accuracy": 0.17793631553649902,
      "num_tokens": 43352322.0,
      "step": 17110
    },
    {
      "entropy": 6.2946436405181885,
      "epoch": 1.9751875360646278,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046182982911335016,
      "loss": 6.0919,
      "mean_token_accuracy": 0.17028756439685822,
      "num_tokens": 43366212.0,
      "step": 17115
    },
    {
      "entropy": 6.314712762832642,
      "epoch": 1.9757645701096365,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046180683202791905,
      "loss": 6.1199,
      "mean_token_accuracy": 0.16482252925634383,
      "num_tokens": 43379123.0,
      "step": 17120
    },
    {
      "entropy": 6.320395517349243,
      "epoch": 1.976341604154645,
      "grad_norm": 1.09375,
      "learning_rate": 0.000461783828659333,
      "loss": 6.1507,
      "mean_token_accuracy": 0.15963018387556077,
      "num_tokens": 43393188.0,
      "step": 17125
    },
    {
      "entropy": 6.330562162399292,
      "epoch": 1.9769186381996537,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046176081900836565,
      "loss": 6.0731,
      "mean_token_accuracy": 0.16618767976760865,
      "num_tokens": 43407448.0,
      "step": 17130
    },
    {
      "entropy": 6.234742307662964,
      "epoch": 1.9774956722446624,
      "grad_norm": 2.0625,
      "learning_rate": 0.00046173780307579086,
      "loss": 5.9955,
      "mean_token_accuracy": 0.16900865137577056,
      "num_tokens": 43420202.0,
      "step": 17135
    },
    {
      "entropy": 6.276762580871582,
      "epoch": 1.9780727062896712,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004617147808623829,
      "loss": 6.0736,
      "mean_token_accuracy": 0.16688749194145203,
      "num_tokens": 43433218.0,
      "step": 17140
    },
    {
      "entropy": 6.230731058120727,
      "epoch": 1.9786497403346797,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046169175236891605,
      "loss": 6.0409,
      "mean_token_accuracy": 0.17231394499540328,
      "num_tokens": 43445721.0,
      "step": 17145
    },
    {
      "entropy": 6.259533452987671,
      "epoch": 1.9792267743796885,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004616687175961648,
      "loss": 6.0316,
      "mean_token_accuracy": 0.1742004543542862,
      "num_tokens": 43457470.0,
      "step": 17150
    },
    {
      "entropy": 6.2746974468231205,
      "epoch": 1.979803808424697,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004616456765449039,
      "loss": 6.0402,
      "mean_token_accuracy": 0.17770297527313234,
      "num_tokens": 43470693.0,
      "step": 17155
    },
    {
      "entropy": 6.29163088798523,
      "epoch": 1.9803808424697058,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046162262921590845,
      "loss": 6.0524,
      "mean_token_accuracy": 0.1658223256468773,
      "num_tokens": 43482890.0,
      "step": 17160
    },
    {
      "entropy": 6.346973514556884,
      "epoch": 1.9809578765147142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004615995756099535,
      "loss": 6.074,
      "mean_token_accuracy": 0.1686537370085716,
      "num_tokens": 43495444.0,
      "step": 17165
    },
    {
      "entropy": 6.248704862594605,
      "epoch": 1.981534910559723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004615765157278146,
      "loss": 6.0425,
      "mean_token_accuracy": 0.16893480867147445,
      "num_tokens": 43507415.0,
      "step": 17170
    },
    {
      "entropy": 6.229138278961182,
      "epoch": 1.9821119446047317,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046155344957026726,
      "loss": 6.0765,
      "mean_token_accuracy": 0.16631700247526168,
      "num_tokens": 43520832.0,
      "step": 17175
    },
    {
      "entropy": 6.310163736343384,
      "epoch": 1.9826889786497404,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004615303771380873,
      "loss": 5.9988,
      "mean_token_accuracy": 0.17154231816530227,
      "num_tokens": 43534459.0,
      "step": 17180
    },
    {
      "entropy": 6.201480484008789,
      "epoch": 1.983266012694749,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046150729843205077,
      "loss": 5.8558,
      "mean_token_accuracy": 0.18204520791769027,
      "num_tokens": 43546466.0,
      "step": 17185
    },
    {
      "entropy": 6.220339441299439,
      "epoch": 1.9838430467397576,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046148421345293384,
      "loss": 5.9301,
      "mean_token_accuracy": 0.17631548196077346,
      "num_tokens": 43558285.0,
      "step": 17190
    },
    {
      "entropy": 6.1403234004974365,
      "epoch": 1.9844200807847663,
      "grad_norm": 1.25,
      "learning_rate": 0.0004614611222015131,
      "loss": 5.9156,
      "mean_token_accuracy": 0.18074734061956405,
      "num_tokens": 43572117.0,
      "step": 17195
    },
    {
      "entropy": 6.262120151519776,
      "epoch": 1.984997114829775,
      "grad_norm": 2.140625,
      "learning_rate": 0.00046143802467856507,
      "loss": 6.187,
      "mean_token_accuracy": 0.16410153955221177,
      "num_tokens": 43586951.0,
      "step": 17200
    },
    {
      "entropy": 6.277947902679443,
      "epoch": 1.9855741488747836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046141492088486673,
      "loss": 5.9645,
      "mean_token_accuracy": 0.17765239626169205,
      "num_tokens": 43600542.0,
      "step": 17205
    },
    {
      "entropy": 6.247993421554566,
      "epoch": 1.9861511829197922,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004613918108211951,
      "loss": 6.0523,
      "mean_token_accuracy": 0.1717355266213417,
      "num_tokens": 43612928.0,
      "step": 17210
    },
    {
      "entropy": 6.333569145202636,
      "epoch": 1.986728216964801,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004613686944883275,
      "loss": 6.0304,
      "mean_token_accuracy": 0.17236097604036332,
      "num_tokens": 43625362.0,
      "step": 17215
    },
    {
      "entropy": 6.213045835494995,
      "epoch": 1.9873052510098095,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00046134557188704146,
      "loss": 6.0204,
      "mean_token_accuracy": 0.17817834466695787,
      "num_tokens": 43638165.0,
      "step": 17220
    },
    {
      "entropy": 6.135746431350708,
      "epoch": 1.9878822850548183,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046132244301811466,
      "loss": 5.9393,
      "mean_token_accuracy": 0.17782478779554367,
      "num_tokens": 43650688.0,
      "step": 17225
    },
    {
      "entropy": 6.326607179641724,
      "epoch": 1.9884593190998268,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046129930788232497,
      "loss": 6.0594,
      "mean_token_accuracy": 0.16792702078819274,
      "num_tokens": 43663168.0,
      "step": 17230
    },
    {
      "entropy": 6.254304552078247,
      "epoch": 1.9890363531448356,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046127616648045073,
      "loss": 6.029,
      "mean_token_accuracy": 0.16387904584407806,
      "num_tokens": 43676391.0,
      "step": 17235
    },
    {
      "entropy": 6.3034281730651855,
      "epoch": 1.989613387189844,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046125301881327007,
      "loss": 6.0229,
      "mean_token_accuracy": 0.16842008978128434,
      "num_tokens": 43689631.0,
      "step": 17240
    },
    {
      "entropy": 6.273438882827759,
      "epoch": 1.990190421234853,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046122986488156167,
      "loss": 6.0479,
      "mean_token_accuracy": 0.17078548073768615,
      "num_tokens": 43703176.0,
      "step": 17245
    },
    {
      "entropy": 6.339113378524781,
      "epoch": 1.9907674552798615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046120670468610426,
      "loss": 6.1197,
      "mean_token_accuracy": 0.16192102879285813,
      "num_tokens": 43717439.0,
      "step": 17250
    },
    {
      "entropy": 6.272079277038574,
      "epoch": 1.9913444893248702,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046118353822767683,
      "loss": 5.9796,
      "mean_token_accuracy": 0.1686514735221863,
      "num_tokens": 43730500.0,
      "step": 17255
    },
    {
      "entropy": 6.155890417098999,
      "epoch": 1.9919215233698788,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004611603655070586,
      "loss": 5.9199,
      "mean_token_accuracy": 0.17832320630550386,
      "num_tokens": 43744807.0,
      "step": 17260
    },
    {
      "entropy": 6.255143928527832,
      "epoch": 1.9924985574148875,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046113718652502896,
      "loss": 6.0038,
      "mean_token_accuracy": 0.17497580498456955,
      "num_tokens": 43756748.0,
      "step": 17265
    },
    {
      "entropy": 6.257786607742309,
      "epoch": 1.993075591459896,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004611140012823675,
      "loss": 6.0644,
      "mean_token_accuracy": 0.16814451068639755,
      "num_tokens": 43769309.0,
      "step": 17270
    },
    {
      "entropy": 6.215093517303467,
      "epoch": 1.9936526255049047,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046109080977985395,
      "loss": 6.0178,
      "mean_token_accuracy": 0.1774963155388832,
      "num_tokens": 43781057.0,
      "step": 17275
    },
    {
      "entropy": 6.306448173522949,
      "epoch": 1.9942296595499136,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046106761201826854,
      "loss": 6.0576,
      "mean_token_accuracy": 0.16839146763086318,
      "num_tokens": 43794504.0,
      "step": 17280
    },
    {
      "entropy": 6.134201717376709,
      "epoch": 1.994806693594922,
      "grad_norm": 1.375,
      "learning_rate": 0.00046104440799839145,
      "loss": 5.8808,
      "mean_token_accuracy": 0.18590227514505386,
      "num_tokens": 43810214.0,
      "step": 17285
    },
    {
      "entropy": 6.216445350646973,
      "epoch": 1.9953837276399309,
      "grad_norm": 1.59375,
      "learning_rate": 0.000461021197721003,
      "loss": 6.0415,
      "mean_token_accuracy": 0.16848865598440171,
      "num_tokens": 43822197.0,
      "step": 17290
    },
    {
      "entropy": 6.314849901199341,
      "epoch": 1.9959607616849393,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046099798118688407,
      "loss": 6.021,
      "mean_token_accuracy": 0.1691252052783966,
      "num_tokens": 43835585.0,
      "step": 17295
    },
    {
      "entropy": 6.326150131225586,
      "epoch": 1.9965377957299482,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004609747583968154,
      "loss": 5.9943,
      "mean_token_accuracy": 0.1651899218559265,
      "num_tokens": 43848542.0,
      "step": 17300
    },
    {
      "entropy": 6.180682373046875,
      "epoch": 1.9971148297749566,
      "grad_norm": 2.6875,
      "learning_rate": 0.0004609515293515781,
      "loss": 6.0175,
      "mean_token_accuracy": 0.17395762950181962,
      "num_tokens": 43860717.0,
      "step": 17305
    },
    {
      "entropy": 6.2602049827575685,
      "epoch": 1.9976918638199654,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004609282940519535,
      "loss": 6.0779,
      "mean_token_accuracy": 0.1715341955423355,
      "num_tokens": 43873193.0,
      "step": 17310
    },
    {
      "entropy": 6.3034483909606935,
      "epoch": 1.998268897864974,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004609050524987231,
      "loss": 6.0417,
      "mean_token_accuracy": 0.16786195784807206,
      "num_tokens": 43886241.0,
      "step": 17315
    },
    {
      "entropy": 6.274380493164062,
      "epoch": 1.9988459319099827,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004608818046926686,
      "loss": 6.018,
      "mean_token_accuracy": 0.17538921535015106,
      "num_tokens": 43899429.0,
      "step": 17320
    },
    {
      "entropy": 6.220386648178101,
      "epoch": 1.9994229659549914,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004608585506345719,
      "loss": 6.0432,
      "mean_token_accuracy": 0.16782113611698152,
      "num_tokens": 43913073.0,
      "step": 17325
    },
    {
      "entropy": 6.264380693435669,
      "epoch": 2.0,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004608352903252152,
      "loss": 6.0018,
      "mean_token_accuracy": 0.17643982023000718,
      "num_tokens": 43926234.0,
      "step": 17330
    },
    {
      "entropy": 6.297075891494751,
      "epoch": 2.000577034045009,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046081202376538084,
      "loss": 5.9901,
      "mean_token_accuracy": 0.1705596461892128,
      "num_tokens": 43937787.0,
      "step": 17335
    },
    {
      "entropy": 6.214798879623413,
      "epoch": 2.0011540680900173,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004607887509558513,
      "loss": 6.0,
      "mean_token_accuracy": 0.17187341153621674,
      "num_tokens": 43951429.0,
      "step": 17340
    },
    {
      "entropy": 6.2717287063598635,
      "epoch": 2.001731102135026,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004607654718974094,
      "loss": 6.0265,
      "mean_token_accuracy": 0.17009468674659728,
      "num_tokens": 43964257.0,
      "step": 17345
    },
    {
      "entropy": 6.300295209884643,
      "epoch": 2.0023081361800346,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004607421865908382,
      "loss": 5.929,
      "mean_token_accuracy": 0.1742121919989586,
      "num_tokens": 43975995.0,
      "step": 17350
    },
    {
      "entropy": 6.187320327758789,
      "epoch": 2.0028851702250434,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004607188950369207,
      "loss": 5.8799,
      "mean_token_accuracy": 0.17936831712722778,
      "num_tokens": 43989191.0,
      "step": 17355
    },
    {
      "entropy": 6.265660429000855,
      "epoch": 2.003462204270052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004606955972364405,
      "loss": 5.9906,
      "mean_token_accuracy": 0.17080468833446502,
      "num_tokens": 44000583.0,
      "step": 17360
    },
    {
      "entropy": 6.1407544136047365,
      "epoch": 2.0040392383150607,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004606722931901812,
      "loss": 5.8517,
      "mean_token_accuracy": 0.1786182105541229,
      "num_tokens": 44013894.0,
      "step": 17365
    },
    {
      "entropy": 6.202847480773926,
      "epoch": 2.004616272360069,
      "grad_norm": 1.25,
      "learning_rate": 0.0004606489828989264,
      "loss": 5.9156,
      "mean_token_accuracy": 0.17088600397109985,
      "num_tokens": 44026701.0,
      "step": 17370
    },
    {
      "entropy": 6.175824975967407,
      "epoch": 2.005193306405078,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004606256663634604,
      "loss": 5.9326,
      "mean_token_accuracy": 0.1767387419939041,
      "num_tokens": 44039775.0,
      "step": 17375
    },
    {
      "entropy": 6.25781946182251,
      "epoch": 2.0057703404500864,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004606023435845672,
      "loss": 5.974,
      "mean_token_accuracy": 0.17226887345314026,
      "num_tokens": 44052352.0,
      "step": 17380
    },
    {
      "entropy": 6.250827646255493,
      "epoch": 2.0063473744950953,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004605790145630314,
      "loss": 5.9554,
      "mean_token_accuracy": 0.17064084112644196,
      "num_tokens": 44066147.0,
      "step": 17385
    },
    {
      "entropy": 6.238795328140259,
      "epoch": 2.0069244085401037,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004605556792996377,
      "loss": 5.959,
      "mean_token_accuracy": 0.17261969149112702,
      "num_tokens": 44078692.0,
      "step": 17390
    },
    {
      "entropy": 6.296038961410522,
      "epoch": 2.0075014425851125,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004605323377951709,
      "loss": 6.006,
      "mean_token_accuracy": 0.16696172207593918,
      "num_tokens": 44091511.0,
      "step": 17395
    },
    {
      "entropy": 6.22690978050232,
      "epoch": 2.008078476630121,
      "grad_norm": 1.4375,
      "learning_rate": 0.000460508990050416,
      "loss": 5.9448,
      "mean_token_accuracy": 0.170767118036747,
      "num_tokens": 44103423.0,
      "step": 17400
    },
    {
      "entropy": 6.131740474700928,
      "epoch": 2.00865551067513,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004604856360661584,
      "loss": 5.8242,
      "mean_token_accuracy": 0.18300881683826448,
      "num_tokens": 44116024.0,
      "step": 17405
    },
    {
      "entropy": 6.136595010757446,
      "epoch": 2.0092325447201387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004604622758431835,
      "loss": 5.8754,
      "mean_token_accuracy": 0.17836426347494125,
      "num_tokens": 44128595.0,
      "step": 17410
    },
    {
      "entropy": 6.280027580261231,
      "epoch": 2.009809578765147,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046043890938227724,
      "loss": 5.9718,
      "mean_token_accuracy": 0.1769594207406044,
      "num_tokens": 44140104.0,
      "step": 17415
    },
    {
      "entropy": 6.2423299789428714,
      "epoch": 2.010386612810156,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046041553668422526,
      "loss": 5.9318,
      "mean_token_accuracy": 0.1780355080962181,
      "num_tokens": 44152681.0,
      "step": 17420
    },
    {
      "entropy": 6.12680401802063,
      "epoch": 2.0109636468551644,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046039215774981376,
      "loss": 5.965,
      "mean_token_accuracy": 0.17880391776561738,
      "num_tokens": 44165728.0,
      "step": 17425
    },
    {
      "entropy": 6.262257671356201,
      "epoch": 2.0115406809001732,
      "grad_norm": 1.25,
      "learning_rate": 0.00046036877257982917,
      "loss": 5.9904,
      "mean_token_accuracy": 0.16699692755937576,
      "num_tokens": 44178182.0,
      "step": 17430
    },
    {
      "entropy": 6.2879571437835695,
      "epoch": 2.0121177149451817,
      "grad_norm": 1.234375,
      "learning_rate": 0.000460345381175058,
      "loss": 5.9369,
      "mean_token_accuracy": 0.17570238560438156,
      "num_tokens": 44190954.0,
      "step": 17435
    },
    {
      "entropy": 6.242337083816528,
      "epoch": 2.0126947489901905,
      "grad_norm": 1.03125,
      "learning_rate": 0.000460321983536287,
      "loss": 5.9239,
      "mean_token_accuracy": 0.17687204331159592,
      "num_tokens": 44204123.0,
      "step": 17440
    },
    {
      "entropy": 6.183831357955933,
      "epoch": 2.013271783035199,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046029857966430315,
      "loss": 5.9268,
      "mean_token_accuracy": 0.17197465896606445,
      "num_tokens": 44218656.0,
      "step": 17445
    },
    {
      "entropy": 6.203491353988648,
      "epoch": 2.013848817080208,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046027516955989364,
      "loss": 5.9424,
      "mean_token_accuracy": 0.1783543273806572,
      "num_tokens": 44231191.0,
      "step": 17450
    },
    {
      "entropy": 6.273696565628052,
      "epoch": 2.014425851125216,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046025175322384577,
      "loss": 5.9481,
      "mean_token_accuracy": 0.17753272205591203,
      "num_tokens": 44244564.0,
      "step": 17455
    },
    {
      "entropy": 6.272516918182373,
      "epoch": 2.015002885170225,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046022833065694727,
      "loss": 5.984,
      "mean_token_accuracy": 0.17100428342819213,
      "num_tokens": 44257545.0,
      "step": 17460
    },
    {
      "entropy": 6.192347669601441,
      "epoch": 2.0155799192152335,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046020490185998575,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17826618552207946,
      "num_tokens": 44270416.0,
      "step": 17465
    },
    {
      "entropy": 6.214372444152832,
      "epoch": 2.0161569532602424,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004601814668337495,
      "loss": 5.9211,
      "mean_token_accuracy": 0.17736995220184326,
      "num_tokens": 44283366.0,
      "step": 17470
    },
    {
      "entropy": 6.20364727973938,
      "epoch": 2.016733987305251,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046015802557902654,
      "loss": 5.979,
      "mean_token_accuracy": 0.17061620205640793,
      "num_tokens": 44296079.0,
      "step": 17475
    },
    {
      "entropy": 6.216525363922119,
      "epoch": 2.0173110213502596,
      "grad_norm": 2.453125,
      "learning_rate": 0.00046013457809660537,
      "loss": 5.9299,
      "mean_token_accuracy": 0.173180291056633,
      "num_tokens": 44308409.0,
      "step": 17480
    },
    {
      "entropy": 6.206072568893433,
      "epoch": 2.0178880553952685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046011112438727454,
      "loss": 6.0231,
      "mean_token_accuracy": 0.17291929721832275,
      "num_tokens": 44321826.0,
      "step": 17485
    },
    {
      "entropy": 6.285996961593628,
      "epoch": 2.018465089440277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004600876644518231,
      "loss": 5.9954,
      "mean_token_accuracy": 0.1680159881711006,
      "num_tokens": 44335630.0,
      "step": 17490
    },
    {
      "entropy": 6.087948226928711,
      "epoch": 2.0190421234852858,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046006419829104,
      "loss": 5.7567,
      "mean_token_accuracy": 0.18672087639570237,
      "num_tokens": 44349059.0,
      "step": 17495
    },
    {
      "entropy": 6.192037963867188,
      "epoch": 2.019619157530294,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004600407259057145,
      "loss": 5.8764,
      "mean_token_accuracy": 0.17351363152265548,
      "num_tokens": 44361908.0,
      "step": 17500
    },
    {
      "entropy": 6.220724296569824,
      "epoch": 2.020196191575303,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004600172472966361,
      "loss": 5.9722,
      "mean_token_accuracy": 0.16882861256599427,
      "num_tokens": 44375487.0,
      "step": 17505
    },
    {
      "entropy": 6.192804622650146,
      "epoch": 2.0207732256203115,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045999376246459446,
      "loss": 5.9108,
      "mean_token_accuracy": 0.17260125130414963,
      "num_tokens": 44387881.0,
      "step": 17510
    },
    {
      "entropy": 6.209100961685181,
      "epoch": 2.0213502596653203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004599702714103795,
      "loss": 5.8953,
      "mean_token_accuracy": 0.17245278358459473,
      "num_tokens": 44400609.0,
      "step": 17515
    },
    {
      "entropy": 6.170619869232178,
      "epoch": 2.0219272937103288,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0004599467741347814,
      "loss": 5.8932,
      "mean_token_accuracy": 0.1790915235877037,
      "num_tokens": 44413341.0,
      "step": 17520
    },
    {
      "entropy": 6.221302604675293,
      "epoch": 2.0225043277553376,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004599232706385904,
      "loss": 6.0261,
      "mean_token_accuracy": 0.17114851176738738,
      "num_tokens": 44425790.0,
      "step": 17525
    },
    {
      "entropy": 6.184481763839722,
      "epoch": 2.023081361800346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000459899760922597,
      "loss": 5.8961,
      "mean_token_accuracy": 0.18134428858757018,
      "num_tokens": 44438204.0,
      "step": 17530
    },
    {
      "entropy": 6.192145204544067,
      "epoch": 2.023658395845355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004598762449875921,
      "loss": 5.8576,
      "mean_token_accuracy": 0.17917026430368424,
      "num_tokens": 44450717.0,
      "step": 17535
    },
    {
      "entropy": 6.228597211837768,
      "epoch": 2.0242354298903633,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004598527228343665,
      "loss": 5.8816,
      "mean_token_accuracy": 0.1814187839627266,
      "num_tokens": 44463391.0,
      "step": 17540
    },
    {
      "entropy": 6.161728239059448,
      "epoch": 2.024812463935372,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004598291944637112,
      "loss": 5.8686,
      "mean_token_accuracy": 0.17788388729095458,
      "num_tokens": 44477022.0,
      "step": 17545
    },
    {
      "entropy": 6.250133514404297,
      "epoch": 2.025389497980381,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00045980565987641797,
      "loss": 5.9895,
      "mean_token_accuracy": 0.17265195846557618,
      "num_tokens": 44490616.0,
      "step": 17550
    },
    {
      "entropy": 6.236105871200562,
      "epoch": 2.0259665320253895,
      "grad_norm": 1.125,
      "learning_rate": 0.00045978211907327804,
      "loss": 5.9548,
      "mean_token_accuracy": 0.17624759078025817,
      "num_tokens": 44503307.0,
      "step": 17555
    },
    {
      "entropy": 6.265711402893066,
      "epoch": 2.0265435660703983,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004597585720550834,
      "loss": 5.9679,
      "mean_token_accuracy": 0.1739831820130348,
      "num_tokens": 44516078.0,
      "step": 17560
    },
    {
      "entropy": 6.231649446487427,
      "epoch": 2.0271206001154067,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000459735018822626,
      "loss": 5.9245,
      "mean_token_accuracy": 0.17394810765981675,
      "num_tokens": 44528485.0,
      "step": 17565
    },
    {
      "entropy": 6.240048360824585,
      "epoch": 2.0276976341604156,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00045971145937669794,
      "loss": 5.8738,
      "mean_token_accuracy": 0.17736487239599227,
      "num_tokens": 44539946.0,
      "step": 17570
    },
    {
      "entropy": 6.132218885421753,
      "epoch": 2.028274668205424,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004596878937180917,
      "loss": 5.958,
      "mean_token_accuracy": 0.1710592895746231,
      "num_tokens": 44552828.0,
      "step": 17575
    },
    {
      "entropy": 6.226531219482422,
      "epoch": 2.028851702250433,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004596643218475999,
      "loss": 5.9678,
      "mean_token_accuracy": 0.1757684901356697,
      "num_tokens": 44564930.0,
      "step": 17580
    },
    {
      "entropy": 6.2708100318908695,
      "epoch": 2.0294287362954413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045964074376601534,
      "loss": 5.9457,
      "mean_token_accuracy": 0.17014985233545304,
      "num_tokens": 44578226.0,
      "step": 17585
    },
    {
      "entropy": 6.195923471450806,
      "epoch": 2.03000577034045,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045961715947413106,
      "loss": 5.9532,
      "mean_token_accuracy": 0.17502938061952591,
      "num_tokens": 44590704.0,
      "step": 17590
    },
    {
      "entropy": 6.199918508529663,
      "epoch": 2.0305828043854586,
      "grad_norm": 1.0,
      "learning_rate": 0.0004595935689727404,
      "loss": 5.9319,
      "mean_token_accuracy": 0.18110113143920897,
      "num_tokens": 44602758.0,
      "step": 17595
    },
    {
      "entropy": 6.124384832382202,
      "epoch": 2.0311598384304674,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004595699722626367,
      "loss": 5.8945,
      "mean_token_accuracy": 0.1795112207531929,
      "num_tokens": 44614733.0,
      "step": 17600
    },
    {
      "entropy": 6.308047533035278,
      "epoch": 2.031736872475476,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045954636934461367,
      "loss": 6.0143,
      "mean_token_accuracy": 0.17096339911222458,
      "num_tokens": 44628339.0,
      "step": 17605
    },
    {
      "entropy": 6.294468355178833,
      "epoch": 2.0323139065204847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004595227602194652,
      "loss": 5.9252,
      "mean_token_accuracy": 0.17322905659675597,
      "num_tokens": 44641382.0,
      "step": 17610
    },
    {
      "entropy": 6.156301641464234,
      "epoch": 2.0328909405654936,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004594991448879853,
      "loss": 5.9102,
      "mean_token_accuracy": 0.1699449121952057,
      "num_tokens": 44654618.0,
      "step": 17615
    },
    {
      "entropy": 6.299897193908691,
      "epoch": 2.033467974610502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004594755233509683,
      "loss": 6.0546,
      "mean_token_accuracy": 0.16198293268680572,
      "num_tokens": 44667961.0,
      "step": 17620
    },
    {
      "entropy": 6.191911935806274,
      "epoch": 2.034045008655511,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045945189560920875,
      "loss": 5.9306,
      "mean_token_accuracy": 0.17262049913406372,
      "num_tokens": 44682185.0,
      "step": 17625
    },
    {
      "entropy": 6.27217288017273,
      "epoch": 2.0346220427005193,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004594282616635013,
      "loss": 5.9646,
      "mean_token_accuracy": 0.17757879346609115,
      "num_tokens": 44695513.0,
      "step": 17630
    },
    {
      "entropy": 6.232735204696655,
      "epoch": 2.035199076745528,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004594046215146409,
      "loss": 5.9928,
      "mean_token_accuracy": 0.17601801156997682,
      "num_tokens": 44707051.0,
      "step": 17635
    },
    {
      "entropy": 6.237101078033447,
      "epoch": 2.0357761107905366,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00045938097516342257,
      "loss": 5.982,
      "mean_token_accuracy": 0.17229001522064208,
      "num_tokens": 44719915.0,
      "step": 17640
    },
    {
      "entropy": 6.118812084197998,
      "epoch": 2.0363531448355454,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045935732261064184,
      "loss": 5.837,
      "mean_token_accuracy": 0.1862248256802559,
      "num_tokens": 44733115.0,
      "step": 17645
    },
    {
      "entropy": 6.205551958084106,
      "epoch": 2.036930178880554,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045933366385709405,
      "loss": 5.9436,
      "mean_token_accuracy": 0.1748773232102394,
      "num_tokens": 44744480.0,
      "step": 17650
    },
    {
      "entropy": 6.203357934951782,
      "epoch": 2.0375072129255627,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004593099989035751,
      "loss": 5.9754,
      "mean_token_accuracy": 0.17659931629896164,
      "num_tokens": 44758200.0,
      "step": 17655
    },
    {
      "entropy": 6.316721200942993,
      "epoch": 2.038084246970571,
      "grad_norm": 1.25,
      "learning_rate": 0.0004592863277508809,
      "loss": 6.029,
      "mean_token_accuracy": 0.1700327605009079,
      "num_tokens": 44771407.0,
      "step": 17660
    },
    {
      "entropy": 6.287095546722412,
      "epoch": 2.03866128101558,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004592626503998076,
      "loss": 5.9155,
      "mean_token_accuracy": 0.17439333498477935,
      "num_tokens": 44783765.0,
      "step": 17665
    },
    {
      "entropy": 6.228414916992188,
      "epoch": 2.0392383150605884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004592389668511515,
      "loss": 5.9797,
      "mean_token_accuracy": 0.1745125100016594,
      "num_tokens": 44796550.0,
      "step": 17670
    },
    {
      "entropy": 6.192117929458618,
      "epoch": 2.0398153491055973,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004592152771057093,
      "loss": 5.939,
      "mean_token_accuracy": 0.17687230557203293,
      "num_tokens": 44808802.0,
      "step": 17675
    },
    {
      "entropy": 6.250724363327026,
      "epoch": 2.0403923831506057,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045919158116427785,
      "loss": 5.8742,
      "mean_token_accuracy": 0.17866400480270386,
      "num_tokens": 44821023.0,
      "step": 17680
    },
    {
      "entropy": 6.220262145996093,
      "epoch": 2.0409694171956145,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045916787902765396,
      "loss": 5.9044,
      "mean_token_accuracy": 0.175578336417675,
      "num_tokens": 44833822.0,
      "step": 17685
    },
    {
      "entropy": 6.197280168533325,
      "epoch": 2.0415464512406234,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004591441706966349,
      "loss": 5.9274,
      "mean_token_accuracy": 0.17216354012489318,
      "num_tokens": 44846440.0,
      "step": 17690
    },
    {
      "entropy": 6.200249481201172,
      "epoch": 2.042123485285632,
      "grad_norm": 7.25,
      "learning_rate": 0.0004591204561720183,
      "loss": 5.9403,
      "mean_token_accuracy": 0.1695069521665573,
      "num_tokens": 44858257.0,
      "step": 17695
    },
    {
      "entropy": 6.175353002548218,
      "epoch": 2.0427005193306407,
      "grad_norm": 1.125,
      "learning_rate": 0.0004590967354546015,
      "loss": 5.9805,
      "mean_token_accuracy": 0.1676792249083519,
      "num_tokens": 44871339.0,
      "step": 17700
    },
    {
      "entropy": 6.270719003677368,
      "epoch": 2.043277553375649,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004590730085451824,
      "loss": 5.9391,
      "mean_token_accuracy": 0.1787752777338028,
      "num_tokens": 44884190.0,
      "step": 17705
    },
    {
      "entropy": 6.15546760559082,
      "epoch": 2.043854587420658,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045904927544455914,
      "loss": 5.858,
      "mean_token_accuracy": 0.17509367167949677,
      "num_tokens": 44897530.0,
      "step": 17710
    },
    {
      "entropy": 6.192881536483765,
      "epoch": 2.0444316214656664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045902553615353005,
      "loss": 5.8848,
      "mean_token_accuracy": 0.1834646299481392,
      "num_tokens": 44910428.0,
      "step": 17715
    },
    {
      "entropy": 6.223622989654541,
      "epoch": 2.0450086555106752,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004590017906728933,
      "loss": 5.9294,
      "mean_token_accuracy": 0.17263387590646745,
      "num_tokens": 44922904.0,
      "step": 17720
    },
    {
      "entropy": 6.153698301315307,
      "epoch": 2.0455856895556837,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045897803900344774,
      "loss": 5.8738,
      "mean_token_accuracy": 0.17572209388017654,
      "num_tokens": 44935541.0,
      "step": 17725
    },
    {
      "entropy": 6.176003122329712,
      "epoch": 2.0461627236006925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004589542811459923,
      "loss": 5.9264,
      "mean_token_accuracy": 0.1788767859339714,
      "num_tokens": 44948483.0,
      "step": 17730
    },
    {
      "entropy": 6.233722305297851,
      "epoch": 2.046739757645701,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004589305171013259,
      "loss": 5.9362,
      "mean_token_accuracy": 0.1753383070230484,
      "num_tokens": 44961797.0,
      "step": 17735
    },
    {
      "entropy": 6.2383228778839115,
      "epoch": 2.04731679169071,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000458906746870248,
      "loss": 5.9733,
      "mean_token_accuracy": 0.16818111687898635,
      "num_tokens": 44975426.0,
      "step": 17740
    },
    {
      "entropy": 6.220548105239868,
      "epoch": 2.047893825735718,
      "grad_norm": 1.625,
      "learning_rate": 0.000458882970453558,
      "loss": 5.9421,
      "mean_token_accuracy": 0.17429644018411636,
      "num_tokens": 44988272.0,
      "step": 17745
    },
    {
      "entropy": 6.1548073291778564,
      "epoch": 2.048470859780727,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004588591878520556,
      "loss": 5.8904,
      "mean_token_accuracy": 0.17755962610244752,
      "num_tokens": 45000609.0,
      "step": 17750
    },
    {
      "entropy": 6.313485479354858,
      "epoch": 2.049047893825736,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004588353990665407,
      "loss": 6.0308,
      "mean_token_accuracy": 0.16356184035539628,
      "num_tokens": 45012651.0,
      "step": 17755
    },
    {
      "entropy": 6.2108770370483395,
      "epoch": 2.0496249278707444,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004588116040978136,
      "loss": 5.9973,
      "mean_token_accuracy": 0.17209396362304688,
      "num_tokens": 45025237.0,
      "step": 17760
    },
    {
      "entropy": 6.18538384437561,
      "epoch": 2.050201961915753,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045878780294667437,
      "loss": 6.0153,
      "mean_token_accuracy": 0.16943909376859664,
      "num_tokens": 45038329.0,
      "step": 17765
    },
    {
      "entropy": 6.2096922397613525,
      "epoch": 2.0507789959607616,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004587639956139237,
      "loss": 5.8662,
      "mean_token_accuracy": 0.17784364074468612,
      "num_tokens": 45050251.0,
      "step": 17770
    },
    {
      "entropy": 6.265796804428101,
      "epoch": 2.0513560300057705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004587401821003624,
      "loss": 5.9827,
      "mean_token_accuracy": 0.1691094607114792,
      "num_tokens": 45062701.0,
      "step": 17775
    },
    {
      "entropy": 6.1350870609283445,
      "epoch": 2.051933064050779,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045871636240679133,
      "loss": 5.878,
      "mean_token_accuracy": 0.18649837225675583,
      "num_tokens": 45077059.0,
      "step": 17780
    },
    {
      "entropy": 6.225623273849488,
      "epoch": 2.0525100980957878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004586925365340117,
      "loss": 5.8858,
      "mean_token_accuracy": 0.17775747776031495,
      "num_tokens": 45089571.0,
      "step": 17785
    },
    {
      "entropy": 6.204924058914185,
      "epoch": 2.053087132140796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004586687044828247,
      "loss": 5.8941,
      "mean_token_accuracy": 0.174189892411232,
      "num_tokens": 45101085.0,
      "step": 17790
    },
    {
      "entropy": 6.162897729873658,
      "epoch": 2.053664166185805,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004586448662540322,
      "loss": 5.9485,
      "mean_token_accuracy": 0.1684534505009651,
      "num_tokens": 45113844.0,
      "step": 17795
    },
    {
      "entropy": 6.239241313934326,
      "epoch": 2.0542412002308135,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004586210218484358,
      "loss": 5.9866,
      "mean_token_accuracy": 0.1690887078642845,
      "num_tokens": 45126297.0,
      "step": 17800
    },
    {
      "entropy": 6.265953493118286,
      "epoch": 2.0548182342758223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045859717126683745,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18210013806819916,
      "num_tokens": 45139337.0,
      "step": 17805
    },
    {
      "entropy": 6.206180477142334,
      "epoch": 2.0553952683208307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045857331451003953,
      "loss": 5.9986,
      "mean_token_accuracy": 0.17050107568502426,
      "num_tokens": 45151698.0,
      "step": 17810
    },
    {
      "entropy": 6.199662876129151,
      "epoch": 2.0559723023658396,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045854945157884435,
      "loss": 5.8944,
      "mean_token_accuracy": 0.17789710760116578,
      "num_tokens": 45164615.0,
      "step": 17815
    },
    {
      "entropy": 6.235444402694702,
      "epoch": 2.056549336410848,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045852558247405455,
      "loss": 5.9488,
      "mean_token_accuracy": 0.17498012036085128,
      "num_tokens": 45176630.0,
      "step": 17820
    },
    {
      "entropy": 6.209130382537841,
      "epoch": 2.057126370455857,
      "grad_norm": 2.375,
      "learning_rate": 0.00045850170719647287,
      "loss": 5.8856,
      "mean_token_accuracy": 0.18172849118709564,
      "num_tokens": 45188879.0,
      "step": 17825
    },
    {
      "entropy": 6.245654630661011,
      "epoch": 2.0577034045008658,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045847782574690254,
      "loss": 5.9596,
      "mean_token_accuracy": 0.16994517296552658,
      "num_tokens": 45201694.0,
      "step": 17830
    },
    {
      "entropy": 6.198729944229126,
      "epoch": 2.058280438545874,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045845393812614664,
      "loss": 5.9081,
      "mean_token_accuracy": 0.17356704622507096,
      "num_tokens": 45215873.0,
      "step": 17835
    },
    {
      "entropy": 6.292256259918213,
      "epoch": 2.058857472590883,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004584300443350086,
      "loss": 6.0325,
      "mean_token_accuracy": 0.16657613068819047,
      "num_tokens": 45229660.0,
      "step": 17840
    },
    {
      "entropy": 6.233248519897461,
      "epoch": 2.0594345066358914,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004584061443742922,
      "loss": 5.8899,
      "mean_token_accuracy": 0.17116474509239196,
      "num_tokens": 45243257.0,
      "step": 17845
    },
    {
      "entropy": 6.1754065990448,
      "epoch": 2.0600115406809003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045838223824480124,
      "loss": 5.8782,
      "mean_token_accuracy": 0.18036632239818573,
      "num_tokens": 45256212.0,
      "step": 17850
    },
    {
      "entropy": 6.207752275466919,
      "epoch": 2.0605885747259087,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004583583259473397,
      "loss": 5.9363,
      "mean_token_accuracy": 0.1834060311317444,
      "num_tokens": 45268292.0,
      "step": 17855
    },
    {
      "entropy": 6.1441535472869875,
      "epoch": 2.0611656087709176,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045833440748271203,
      "loss": 5.7908,
      "mean_token_accuracy": 0.20127416551113128,
      "num_tokens": 45284112.0,
      "step": 17860
    },
    {
      "entropy": 6.193862199783325,
      "epoch": 2.061742642815926,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045831048285172266,
      "loss": 5.8931,
      "mean_token_accuracy": 0.1810559496283531,
      "num_tokens": 45295839.0,
      "step": 17865
    },
    {
      "entropy": 6.141061782836914,
      "epoch": 2.062319676860935,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004582865520551762,
      "loss": 5.9293,
      "mean_token_accuracy": 0.1708131492137909,
      "num_tokens": 45309142.0,
      "step": 17870
    },
    {
      "entropy": 6.269522953033447,
      "epoch": 2.0628967109059433,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045826261509387755,
      "loss": 5.9701,
      "mean_token_accuracy": 0.17792370170354843,
      "num_tokens": 45322096.0,
      "step": 17875
    },
    {
      "entropy": 6.263595771789551,
      "epoch": 2.063473744950952,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045823867196863197,
      "loss": 5.992,
      "mean_token_accuracy": 0.17633334845304488,
      "num_tokens": 45333685.0,
      "step": 17880
    },
    {
      "entropy": 6.1949646949768065,
      "epoch": 2.0640507789959606,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004582147226802446,
      "loss": 5.9256,
      "mean_token_accuracy": 0.17821483314037323,
      "num_tokens": 45345417.0,
      "step": 17885
    },
    {
      "entropy": 6.200151824951172,
      "epoch": 2.0646278130409694,
      "grad_norm": 1.125,
      "learning_rate": 0.0004581907672295211,
      "loss": 5.934,
      "mean_token_accuracy": 0.17727133482694626,
      "num_tokens": 45357156.0,
      "step": 17890
    },
    {
      "entropy": 6.275757837295532,
      "epoch": 2.065204847085978,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045816680561726716,
      "loss": 5.9526,
      "mean_token_accuracy": 0.17532005161046982,
      "num_tokens": 45369627.0,
      "step": 17895
    },
    {
      "entropy": 6.222644901275634,
      "epoch": 2.0657818811309867,
      "grad_norm": 1.125,
      "learning_rate": 0.0004581428378442886,
      "loss": 5.9016,
      "mean_token_accuracy": 0.1732523813843727,
      "num_tokens": 45380603.0,
      "step": 17900
    },
    {
      "entropy": 6.197103118896484,
      "epoch": 2.0663589151759956,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045811886391139173,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1877364158630371,
      "num_tokens": 45394923.0,
      "step": 17905
    },
    {
      "entropy": 6.229747629165649,
      "epoch": 2.066935949221004,
      "grad_norm": 1.78125,
      "learning_rate": 0.00045809488381938284,
      "loss": 5.9232,
      "mean_token_accuracy": 0.17425263226032256,
      "num_tokens": 45407351.0,
      "step": 17910
    },
    {
      "entropy": 6.097294044494629,
      "epoch": 2.067512983266013,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004580708975690684,
      "loss": 5.9082,
      "mean_token_accuracy": 0.17824243754148483,
      "num_tokens": 45420783.0,
      "step": 17915
    },
    {
      "entropy": 6.217814922332764,
      "epoch": 2.0680900173110213,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004580469051612554,
      "loss": 5.8777,
      "mean_token_accuracy": 0.17622975260019302,
      "num_tokens": 45433208.0,
      "step": 17920
    },
    {
      "entropy": 6.249103021621704,
      "epoch": 2.06866705135603,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045802290659675057,
      "loss": 6.0379,
      "mean_token_accuracy": 0.16410011649131775,
      "num_tokens": 45445872.0,
      "step": 17925
    },
    {
      "entropy": 6.2258580207824705,
      "epoch": 2.0692440854010385,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045799890187636123,
      "loss": 5.9794,
      "mean_token_accuracy": 0.17026836425065994,
      "num_tokens": 45458844.0,
      "step": 17930
    },
    {
      "entropy": 6.201646947860718,
      "epoch": 2.0698211194460474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045797489100089464,
      "loss": 5.957,
      "mean_token_accuracy": 0.17284089624881743,
      "num_tokens": 45471919.0,
      "step": 17935
    },
    {
      "entropy": 6.218793487548828,
      "epoch": 2.070398153491056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004579508739711585,
      "loss": 5.943,
      "mean_token_accuracy": 0.17847752720117568,
      "num_tokens": 45483769.0,
      "step": 17940
    },
    {
      "entropy": 6.286684274673462,
      "epoch": 2.0709751875360647,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045792685078796075,
      "loss": 6.0211,
      "mean_token_accuracy": 0.1702808380126953,
      "num_tokens": 45497077.0,
      "step": 17945
    },
    {
      "entropy": 6.243750429153442,
      "epoch": 2.071552221581073,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000457902821452109,
      "loss": 5.9398,
      "mean_token_accuracy": 0.1820717990398407,
      "num_tokens": 45509229.0,
      "step": 17950
    },
    {
      "entropy": 6.238589191436768,
      "epoch": 2.072129255626082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045787878596441193,
      "loss": 5.9315,
      "mean_token_accuracy": 0.1709604248404503,
      "num_tokens": 45521654.0,
      "step": 17955
    },
    {
      "entropy": 6.228718662261963,
      "epoch": 2.0727062896710904,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004578547443256776,
      "loss": 5.9517,
      "mean_token_accuracy": 0.17131188362836838,
      "num_tokens": 45533488.0,
      "step": 17960
    },
    {
      "entropy": 6.198116397857666,
      "epoch": 2.0732833237160992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004578306965367148,
      "loss": 5.8737,
      "mean_token_accuracy": 0.17840521186590194,
      "num_tokens": 45544938.0,
      "step": 17965
    },
    {
      "entropy": 6.221617031097412,
      "epoch": 2.073860357761108,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045780664259833235,
      "loss": 5.9548,
      "mean_token_accuracy": 0.17204647958278657,
      "num_tokens": 45557085.0,
      "step": 17970
    },
    {
      "entropy": 6.249954557418823,
      "epoch": 2.0744373918061165,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045778258251133924,
      "loss": 5.8955,
      "mean_token_accuracy": 0.17703141123056412,
      "num_tokens": 45570174.0,
      "step": 17975
    },
    {
      "entropy": 6.208803415298462,
      "epoch": 2.0750144258511254,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045775851627654474,
      "loss": 5.9011,
      "mean_token_accuracy": 0.17877745032310485,
      "num_tokens": 45582174.0,
      "step": 17980
    },
    {
      "entropy": 6.200427007675171,
      "epoch": 2.075591459896134,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004577344438947584,
      "loss": 5.992,
      "mean_token_accuracy": 0.17164078503847122,
      "num_tokens": 45595732.0,
      "step": 17985
    },
    {
      "entropy": 6.15403265953064,
      "epoch": 2.0761684939411427,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045771036536678984,
      "loss": 5.9319,
      "mean_token_accuracy": 0.17511217743158342,
      "num_tokens": 45609675.0,
      "step": 17990
    },
    {
      "entropy": 6.279201984405518,
      "epoch": 2.076745527986151,
      "grad_norm": 2.0625,
      "learning_rate": 0.00045768628069344885,
      "loss": 5.9654,
      "mean_token_accuracy": 0.17805370539426804,
      "num_tokens": 45622467.0,
      "step": 17995
    },
    {
      "entropy": 6.270020294189453,
      "epoch": 2.07732256203116,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004576621898755455,
      "loss": 6.0249,
      "mean_token_accuracy": 0.16146025508642198,
      "num_tokens": 45636246.0,
      "step": 18000
    },
    {
      "epoch": 2.07732256203116,
      "eval_entropy": 6.116413441577485,
      "eval_loss": 6.092830181121826,
      "eval_mean_token_accuracy": 0.1729923587462968,
      "eval_num_tokens": 45636246.0,
      "eval_runtime": 22.6545,
      "eval_samples_per_second": 1241.96,
      "eval_steps_per_second": 155.245,
      "step": 18000
    },
    {
      "entropy": 6.242395353317261,
      "epoch": 2.0778995960761684,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045763809291389024,
      "loss": 5.8829,
      "mean_token_accuracy": 0.17664990276098252,
      "num_tokens": 45649127.0,
      "step": 18005
    },
    {
      "entropy": 6.251932859420776,
      "epoch": 2.0784766301211772,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004576139898092933,
      "loss": 5.93,
      "mean_token_accuracy": 0.17574980109930038,
      "num_tokens": 45661633.0,
      "step": 18010
    },
    {
      "entropy": 6.1648228645324705,
      "epoch": 2.0790536641661856,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004575898805625657,
      "loss": 5.8679,
      "mean_token_accuracy": 0.18012765049934387,
      "num_tokens": 45673643.0,
      "step": 18015
    },
    {
      "entropy": 6.276806163787842,
      "epoch": 2.0796306982111945,
      "grad_norm": 2.109375,
      "learning_rate": 0.00045756576517451804,
      "loss": 5.9974,
      "mean_token_accuracy": 0.1685441628098488,
      "num_tokens": 45686125.0,
      "step": 18020
    },
    {
      "entropy": 6.196450567245483,
      "epoch": 2.080207732256203,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045754164364596156,
      "loss": 5.8703,
      "mean_token_accuracy": 0.17436889410018921,
      "num_tokens": 45697919.0,
      "step": 18025
    },
    {
      "entropy": 6.212252378463745,
      "epoch": 2.080784766301212,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004575175159777076,
      "loss": 5.9339,
      "mean_token_accuracy": 0.17510858327150344,
      "num_tokens": 45709450.0,
      "step": 18030
    },
    {
      "entropy": 6.226101922988891,
      "epoch": 2.0813618003462206,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004574933821705676,
      "loss": 5.903,
      "mean_token_accuracy": 0.1775021031498909,
      "num_tokens": 45721464.0,
      "step": 18035
    },
    {
      "entropy": 6.239909267425537,
      "epoch": 2.081938834391229,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004574692422253534,
      "loss": 5.997,
      "mean_token_accuracy": 0.17450278103351594,
      "num_tokens": 45733397.0,
      "step": 18040
    },
    {
      "entropy": 6.160855865478515,
      "epoch": 2.082515868436238,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045744509614287687,
      "loss": 5.9495,
      "mean_token_accuracy": 0.17414658963680268,
      "num_tokens": 45744953.0,
      "step": 18045
    },
    {
      "entropy": 6.224947929382324,
      "epoch": 2.0830929024812463,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004574209439239501,
      "loss": 5.8894,
      "mean_token_accuracy": 0.17783963531255723,
      "num_tokens": 45756914.0,
      "step": 18050
    },
    {
      "entropy": 6.22959337234497,
      "epoch": 2.083669936526255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045739678556938563,
      "loss": 5.9758,
      "mean_token_accuracy": 0.1755364254117012,
      "num_tokens": 45770888.0,
      "step": 18055
    },
    {
      "entropy": 6.221868658065796,
      "epoch": 2.0842469705712636,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045737262107999575,
      "loss": 5.9405,
      "mean_token_accuracy": 0.17491403818130494,
      "num_tokens": 45783551.0,
      "step": 18060
    },
    {
      "entropy": 6.1950257301330565,
      "epoch": 2.0848240046162725,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004573484504565934,
      "loss": 5.9148,
      "mean_token_accuracy": 0.1824825018644333,
      "num_tokens": 45795700.0,
      "step": 18065
    },
    {
      "entropy": 6.157202672958374,
      "epoch": 2.085401038661281,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004573242736999915,
      "loss": 5.8836,
      "mean_token_accuracy": 0.1819426953792572,
      "num_tokens": 45808414.0,
      "step": 18070
    },
    {
      "entropy": 6.22504997253418,
      "epoch": 2.0859780727062898,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045730009081100314,
      "loss": 5.8613,
      "mean_token_accuracy": 0.17845331579446794,
      "num_tokens": 45821710.0,
      "step": 18075
    },
    {
      "entropy": 6.160791349411011,
      "epoch": 2.086555106751298,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045727590179044195,
      "loss": 5.912,
      "mean_token_accuracy": 0.1739562124013901,
      "num_tokens": 45835812.0,
      "step": 18080
    },
    {
      "entropy": 6.191755867004394,
      "epoch": 2.087132140796307,
      "grad_norm": 1.25,
      "learning_rate": 0.0004572517066391211,
      "loss": 5.9026,
      "mean_token_accuracy": 0.1821553200483322,
      "num_tokens": 45849076.0,
      "step": 18085
    },
    {
      "entropy": 6.258313703536987,
      "epoch": 2.0877091748413155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045722750535785484,
      "loss": 5.9761,
      "mean_token_accuracy": 0.1702700823545456,
      "num_tokens": 45861375.0,
      "step": 18090
    },
    {
      "entropy": 6.212756967544555,
      "epoch": 2.0882862088863243,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00045720329794745685,
      "loss": 5.9076,
      "mean_token_accuracy": 0.17553171664476394,
      "num_tokens": 45872096.0,
      "step": 18095
    },
    {
      "entropy": 6.15145845413208,
      "epoch": 2.0888632429313327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004571790844087415,
      "loss": 5.8789,
      "mean_token_accuracy": 0.17751434594392776,
      "num_tokens": 45884843.0,
      "step": 18100
    },
    {
      "entropy": 6.237578344345093,
      "epoch": 2.0894402769763416,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004571548647425231,
      "loss": 5.9045,
      "mean_token_accuracy": 0.17498090863227844,
      "num_tokens": 45896644.0,
      "step": 18105
    },
    {
      "entropy": 6.285304403305053,
      "epoch": 2.0900173110213505,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004571306389496164,
      "loss": 6.0258,
      "mean_token_accuracy": 0.16776983588933944,
      "num_tokens": 45910204.0,
      "step": 18110
    },
    {
      "entropy": 6.244596290588379,
      "epoch": 2.090594345066359,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045710640703083594,
      "loss": 5.9348,
      "mean_token_accuracy": 0.1719875767827034,
      "num_tokens": 45923332.0,
      "step": 18115
    },
    {
      "entropy": 6.222869157791138,
      "epoch": 2.0911713791113677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045708216898699707,
      "loss": 5.9541,
      "mean_token_accuracy": 0.176614148914814,
      "num_tokens": 45937322.0,
      "step": 18120
    },
    {
      "entropy": 6.245453834533691,
      "epoch": 2.091748413156376,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045705792481891483,
      "loss": 5.9062,
      "mean_token_accuracy": 0.17339543551206588,
      "num_tokens": 45950369.0,
      "step": 18125
    },
    {
      "entropy": 6.272603368759155,
      "epoch": 2.092325447201385,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045703367452740477,
      "loss": 6.0512,
      "mean_token_accuracy": 0.17232666164636612,
      "num_tokens": 45963402.0,
      "step": 18130
    },
    {
      "entropy": 6.25339183807373,
      "epoch": 2.0929024812463934,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045700941811328247,
      "loss": 5.9672,
      "mean_token_accuracy": 0.1721173793077469,
      "num_tokens": 45976979.0,
      "step": 18135
    },
    {
      "entropy": 6.207638740539551,
      "epoch": 2.0934795152914023,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045698515557736374,
      "loss": 5.9717,
      "mean_token_accuracy": 0.17337723970413207,
      "num_tokens": 45990334.0,
      "step": 18140
    },
    {
      "entropy": 6.171046018600464,
      "epoch": 2.0940565493364107,
      "grad_norm": 1.125,
      "learning_rate": 0.00045696088692046477,
      "loss": 5.8522,
      "mean_token_accuracy": 0.18572404831647873,
      "num_tokens": 46002322.0,
      "step": 18145
    },
    {
      "entropy": 6.209211778640747,
      "epoch": 2.0946335833814196,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045693661214340174,
      "loss": 5.9573,
      "mean_token_accuracy": 0.17103371322154998,
      "num_tokens": 46014136.0,
      "step": 18150
    },
    {
      "entropy": 6.296702432632446,
      "epoch": 2.095210617426428,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045691233124699117,
      "loss": 5.9871,
      "mean_token_accuracy": 0.17024188190698625,
      "num_tokens": 46026686.0,
      "step": 18155
    },
    {
      "entropy": 6.278897476196289,
      "epoch": 2.095787651471437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004568880442320497,
      "loss": 5.9099,
      "mean_token_accuracy": 0.17562095820903778,
      "num_tokens": 46038462.0,
      "step": 18160
    },
    {
      "entropy": 6.190706062316894,
      "epoch": 2.0963646855164453,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045686375109939417,
      "loss": 5.9664,
      "mean_token_accuracy": 0.171871317923069,
      "num_tokens": 46051549.0,
      "step": 18165
    },
    {
      "entropy": 6.214040946960449,
      "epoch": 2.096941719561454,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004568394518498417,
      "loss": 5.9751,
      "mean_token_accuracy": 0.17257556617259978,
      "num_tokens": 46065282.0,
      "step": 18170
    },
    {
      "entropy": 6.249269962310791,
      "epoch": 2.0975187536064626,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045681514648420965,
      "loss": 5.9601,
      "mean_token_accuracy": 0.17436347603797914,
      "num_tokens": 46077238.0,
      "step": 18175
    },
    {
      "entropy": 6.293581247329712,
      "epoch": 2.0980957876514714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004567908350033154,
      "loss": 5.9766,
      "mean_token_accuracy": 0.1720675766468048,
      "num_tokens": 46089044.0,
      "step": 18180
    },
    {
      "entropy": 6.234652948379517,
      "epoch": 2.0986728216964803,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004567665174079767,
      "loss": 6.0156,
      "mean_token_accuracy": 0.17009917348623277,
      "num_tokens": 46101246.0,
      "step": 18185
    },
    {
      "entropy": 6.269860029220581,
      "epoch": 2.0992498557414887,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045674219369901153,
      "loss": 5.9875,
      "mean_token_accuracy": 0.17488809376955033,
      "num_tokens": 46112735.0,
      "step": 18190
    },
    {
      "entropy": 6.214660596847534,
      "epoch": 2.0998268897864976,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567178638772379,
      "loss": 6.0302,
      "mean_token_accuracy": 0.17096078544855117,
      "num_tokens": 46125547.0,
      "step": 18195
    },
    {
      "entropy": 6.347840452194214,
      "epoch": 2.100403923831506,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004566935279434742,
      "loss": 5.9981,
      "mean_token_accuracy": 0.16747518330812455,
      "num_tokens": 46137356.0,
      "step": 18200
    },
    {
      "entropy": 6.1926312923431395,
      "epoch": 2.100980957876515,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045666918589853895,
      "loss": 5.9094,
      "mean_token_accuracy": 0.17773739844560624,
      "num_tokens": 46149935.0,
      "step": 18205
    },
    {
      "entropy": 6.149626016616821,
      "epoch": 2.1015579919215233,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045664483774325094,
      "loss": 5.8524,
      "mean_token_accuracy": 0.1826179087162018,
      "num_tokens": 46162186.0,
      "step": 18210
    },
    {
      "entropy": 6.1907673358917235,
      "epoch": 2.102135025966532,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004566204834784289,
      "loss": 5.8694,
      "mean_token_accuracy": 0.17748078852891921,
      "num_tokens": 46175221.0,
      "step": 18215
    },
    {
      "entropy": 6.201510906219482,
      "epoch": 2.1027120600115405,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045659612310489225,
      "loss": 5.9586,
      "mean_token_accuracy": 0.18019836395978928,
      "num_tokens": 46187898.0,
      "step": 18220
    },
    {
      "entropy": 6.209728860855103,
      "epoch": 2.1032890940565494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045657175662346014,
      "loss": 5.8822,
      "mean_token_accuracy": 0.1802334815263748,
      "num_tokens": 46200792.0,
      "step": 18225
    },
    {
      "entropy": 6.159025526046753,
      "epoch": 2.103866128101558,
      "grad_norm": 3.09375,
      "learning_rate": 0.0004565473840349522,
      "loss": 5.9392,
      "mean_token_accuracy": 0.17072780430316925,
      "num_tokens": 46213791.0,
      "step": 18230
    },
    {
      "entropy": 6.192187929153443,
      "epoch": 2.1044431621465667,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045652300534018816,
      "loss": 5.9233,
      "mean_token_accuracy": 0.17740607112646103,
      "num_tokens": 46226229.0,
      "step": 18235
    },
    {
      "entropy": 6.192289400100708,
      "epoch": 2.105020196191575,
      "grad_norm": 1.265625,
      "learning_rate": 0.000456498620539988,
      "loss": 5.926,
      "mean_token_accuracy": 0.17601072639226914,
      "num_tokens": 46239253.0,
      "step": 18240
    },
    {
      "entropy": 6.2401594638824465,
      "epoch": 2.105597230236584,
      "grad_norm": 1.375,
      "learning_rate": 0.0004564742296351719,
      "loss": 5.9742,
      "mean_token_accuracy": 0.1719544380903244,
      "num_tokens": 46251578.0,
      "step": 18245
    },
    {
      "entropy": 6.280898237228394,
      "epoch": 2.106174264281593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045644983262656014,
      "loss": 5.965,
      "mean_token_accuracy": 0.17226995527744293,
      "num_tokens": 46265421.0,
      "step": 18250
    },
    {
      "entropy": 6.250201892852783,
      "epoch": 2.1067512983266012,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004564254295149735,
      "loss": 5.9651,
      "mean_token_accuracy": 0.1741169959306717,
      "num_tokens": 46277480.0,
      "step": 18255
    },
    {
      "entropy": 6.202625846862793,
      "epoch": 2.10732833237161,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00045640102030123264,
      "loss": 6.0156,
      "mean_token_accuracy": 0.170202574133873,
      "num_tokens": 46289793.0,
      "step": 18260
    },
    {
      "entropy": 6.236398601531983,
      "epoch": 2.1079053664166185,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045637660498615865,
      "loss": 5.9274,
      "mean_token_accuracy": 0.1808223232626915,
      "num_tokens": 46301653.0,
      "step": 18265
    },
    {
      "entropy": 6.08497109413147,
      "epoch": 2.1084824004616274,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004563521835705725,
      "loss": 5.8335,
      "mean_token_accuracy": 0.18415203988552092,
      "num_tokens": 46314368.0,
      "step": 18270
    },
    {
      "entropy": 6.258514070510865,
      "epoch": 2.109059434506636,
      "grad_norm": 2.25,
      "learning_rate": 0.00045632775605529587,
      "loss": 5.9696,
      "mean_token_accuracy": 0.16779158264398575,
      "num_tokens": 46327446.0,
      "step": 18275
    },
    {
      "entropy": 6.274258184432983,
      "epoch": 2.1096364685516447,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004563033224411501,
      "loss": 5.967,
      "mean_token_accuracy": 0.16915316730737687,
      "num_tokens": 46340370.0,
      "step": 18280
    },
    {
      "entropy": 6.2087993144989015,
      "epoch": 2.110213502596653,
      "grad_norm": 1.125,
      "learning_rate": 0.0004562788827289572,
      "loss": 5.9568,
      "mean_token_accuracy": 0.17465278506278992,
      "num_tokens": 46352675.0,
      "step": 18285
    },
    {
      "entropy": 6.150845050811768,
      "epoch": 2.110790536641662,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045625443691953914,
      "loss": 5.8823,
      "mean_token_accuracy": 0.1770549714565277,
      "num_tokens": 46365404.0,
      "step": 18290
    },
    {
      "entropy": 6.214556503295898,
      "epoch": 2.1113675706866704,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004562299850137181,
      "loss": 5.8913,
      "mean_token_accuracy": 0.17711358964443208,
      "num_tokens": 46377544.0,
      "step": 18295
    },
    {
      "entropy": 6.227174139022827,
      "epoch": 2.111944604731679,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004562055270123166,
      "loss": 5.9894,
      "mean_token_accuracy": 0.1742199331521988,
      "num_tokens": 46390425.0,
      "step": 18300
    },
    {
      "entropy": 6.196693086624146,
      "epoch": 2.1125216387766876,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045618106291615715,
      "loss": 5.9197,
      "mean_token_accuracy": 0.17214723229408263,
      "num_tokens": 46403359.0,
      "step": 18305
    },
    {
      "entropy": 6.224980401992798,
      "epoch": 2.1130986728216965,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004561565927260627,
      "loss": 5.9763,
      "mean_token_accuracy": 0.17695431560277938,
      "num_tokens": 46414700.0,
      "step": 18310
    },
    {
      "entropy": 6.23972864151001,
      "epoch": 2.1136757068667054,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004561321164428561,
      "loss": 5.9649,
      "mean_token_accuracy": 0.1728884235024452,
      "num_tokens": 46427199.0,
      "step": 18315
    },
    {
      "entropy": 6.2846475601196286,
      "epoch": 2.114252740911714,
      "grad_norm": 1.125,
      "learning_rate": 0.0004561076340673609,
      "loss": 5.936,
      "mean_token_accuracy": 0.17691826820373535,
      "num_tokens": 46440143.0,
      "step": 18320
    },
    {
      "entropy": 6.24670181274414,
      "epoch": 2.1148297749567226,
      "grad_norm": 1.0,
      "learning_rate": 0.0004560831456004003,
      "loss": 5.9165,
      "mean_token_accuracy": 0.17078517973423005,
      "num_tokens": 46452868.0,
      "step": 18325
    },
    {
      "entropy": 6.2097993850708,
      "epoch": 2.115406809001731,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004560586510427981,
      "loss": 5.8911,
      "mean_token_accuracy": 0.179876109957695,
      "num_tokens": 46465648.0,
      "step": 18330
    },
    {
      "entropy": 6.155358743667603,
      "epoch": 2.11598384304674,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004560341503953781,
      "loss": 5.9482,
      "mean_token_accuracy": 0.17687012404203414,
      "num_tokens": 46478282.0,
      "step": 18335
    },
    {
      "entropy": 6.1971207618713375,
      "epoch": 2.1165608770917483,
      "grad_norm": 1.0,
      "learning_rate": 0.0004560096436589643,
      "loss": 5.8409,
      "mean_token_accuracy": 0.18354799300432206,
      "num_tokens": 46490299.0,
      "step": 18340
    },
    {
      "entropy": 6.157772064208984,
      "epoch": 2.117137911136757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045598513083438115,
      "loss": 5.8665,
      "mean_token_accuracy": 0.18088093400001526,
      "num_tokens": 46502165.0,
      "step": 18345
    },
    {
      "entropy": 6.207816171646118,
      "epoch": 2.1177149451817656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045596061192245297,
      "loss": 6.0133,
      "mean_token_accuracy": 0.16898939311504363,
      "num_tokens": 46514954.0,
      "step": 18350
    },
    {
      "entropy": 6.271923685073853,
      "epoch": 2.1182919792267745,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004559360869240045,
      "loss": 6.0332,
      "mean_token_accuracy": 0.1673020102083683,
      "num_tokens": 46527088.0,
      "step": 18355
    },
    {
      "entropy": 6.276559066772461,
      "epoch": 2.118869013271783,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004559115558398606,
      "loss": 5.9809,
      "mean_token_accuracy": 0.1680253580212593,
      "num_tokens": 46538737.0,
      "step": 18360
    },
    {
      "entropy": 6.214357948303222,
      "epoch": 2.1194460473167918,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004558870186708465,
      "loss": 5.9074,
      "mean_token_accuracy": 0.1808343708515167,
      "num_tokens": 46551963.0,
      "step": 18365
    },
    {
      "entropy": 6.22217869758606,
      "epoch": 2.1200230813618,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045586247541778724,
      "loss": 6.0179,
      "mean_token_accuracy": 0.17336802631616594,
      "num_tokens": 46564247.0,
      "step": 18370
    },
    {
      "entropy": 6.195797967910766,
      "epoch": 2.120600115406809,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004558379260815085,
      "loss": 5.959,
      "mean_token_accuracy": 0.1755570724606514,
      "num_tokens": 46578491.0,
      "step": 18375
    },
    {
      "entropy": 6.248219442367554,
      "epoch": 2.1211771494518175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004558133706628359,
      "loss": 5.9531,
      "mean_token_accuracy": 0.17279623150825502,
      "num_tokens": 46591528.0,
      "step": 18380
    },
    {
      "entropy": 6.26521782875061,
      "epoch": 2.1217541834968263,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045578880916259554,
      "loss": 6.0107,
      "mean_token_accuracy": 0.16866602301597594,
      "num_tokens": 46605028.0,
      "step": 18385
    },
    {
      "entropy": 6.189346647262573,
      "epoch": 2.122331217541835,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004557642415816132,
      "loss": 5.896,
      "mean_token_accuracy": 0.18203431665897368,
      "num_tokens": 46617713.0,
      "step": 18390
    },
    {
      "entropy": 6.227155113220215,
      "epoch": 2.1229082515868436,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004557396679207155,
      "loss": 6.0157,
      "mean_token_accuracy": 0.16739506721496583,
      "num_tokens": 46631897.0,
      "step": 18395
    },
    {
      "entropy": 6.265323686599731,
      "epoch": 2.1234852856318525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045571508818072887,
      "loss": 6.0151,
      "mean_token_accuracy": 0.16782745271921157,
      "num_tokens": 46645457.0,
      "step": 18400
    },
    {
      "entropy": 6.281725883483887,
      "epoch": 2.124062319676861,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004556905023624799,
      "loss": 5.9885,
      "mean_token_accuracy": 0.17250981628894807,
      "num_tokens": 46657779.0,
      "step": 18405
    },
    {
      "entropy": 6.201862144470215,
      "epoch": 2.1246393537218697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045566591046679577,
      "loss": 5.9359,
      "mean_token_accuracy": 0.1725264832377434,
      "num_tokens": 46670590.0,
      "step": 18410
    },
    {
      "entropy": 6.193232536315918,
      "epoch": 2.125216387766878,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045564131249450343,
      "loss": 5.9053,
      "mean_token_accuracy": 0.17696632593870162,
      "num_tokens": 46683369.0,
      "step": 18415
    },
    {
      "entropy": 6.242049694061279,
      "epoch": 2.125793421811887,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004556167084464302,
      "loss": 5.9065,
      "mean_token_accuracy": 0.1703455090522766,
      "num_tokens": 46695371.0,
      "step": 18420
    },
    {
      "entropy": 6.178325033187866,
      "epoch": 2.1263704558568954,
      "grad_norm": 2.40625,
      "learning_rate": 0.0004555920983234038,
      "loss": 5.9094,
      "mean_token_accuracy": 0.16967119425535201,
      "num_tokens": 46708422.0,
      "step": 18425
    },
    {
      "entropy": 6.190279960632324,
      "epoch": 2.1269474899019043,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00045556748212625183,
      "loss": 5.9167,
      "mean_token_accuracy": 0.16888023316860198,
      "num_tokens": 46722236.0,
      "step": 18430
    },
    {
      "entropy": 6.287196636199951,
      "epoch": 2.1275245239469127,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004555428598558023,
      "loss": 5.9249,
      "mean_token_accuracy": 0.17670912742614747,
      "num_tokens": 46733084.0,
      "step": 18435
    },
    {
      "entropy": 6.139206981658935,
      "epoch": 2.1281015579919216,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004555182315128833,
      "loss": 5.9405,
      "mean_token_accuracy": 0.170323446393013,
      "num_tokens": 46745174.0,
      "step": 18440
    },
    {
      "entropy": 6.298870611190796,
      "epoch": 2.12867859203693,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004554935970983234,
      "loss": 5.9951,
      "mean_token_accuracy": 0.17619017660617828,
      "num_tokens": 46758134.0,
      "step": 18445
    },
    {
      "entropy": 6.195008039474487,
      "epoch": 2.129255626081939,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004554689566129509,
      "loss": 5.8611,
      "mean_token_accuracy": 0.1771741986274719,
      "num_tokens": 46770941.0,
      "step": 18450
    },
    {
      "entropy": 6.199351644515991,
      "epoch": 2.1298326601269473,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045544431005759467,
      "loss": 5.9685,
      "mean_token_accuracy": 0.1785759299993515,
      "num_tokens": 46782532.0,
      "step": 18455
    },
    {
      "entropy": 6.145861291885376,
      "epoch": 2.130409694171956,
      "grad_norm": 1.125,
      "learning_rate": 0.00045541965743308376,
      "loss": 5.9046,
      "mean_token_accuracy": 0.18034725040197372,
      "num_tokens": 46794636.0,
      "step": 18460
    },
    {
      "entropy": 6.2272556781768795,
      "epoch": 2.130986728216965,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004553949987402473,
      "loss": 5.9299,
      "mean_token_accuracy": 0.1736836090683937,
      "num_tokens": 46807797.0,
      "step": 18465
    },
    {
      "entropy": 6.243286943435669,
      "epoch": 2.1315637622619734,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004553703339799146,
      "loss": 5.9213,
      "mean_token_accuracy": 0.17754804641008376,
      "num_tokens": 46819838.0,
      "step": 18470
    },
    {
      "entropy": 6.128636407852173,
      "epoch": 2.1321407963069823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004553456631529154,
      "loss": 5.8695,
      "mean_token_accuracy": 0.18137176036834718,
      "num_tokens": 46832172.0,
      "step": 18475
    },
    {
      "entropy": 6.265114498138428,
      "epoch": 2.1327178303519907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004553209862600794,
      "loss": 5.9937,
      "mean_token_accuracy": 0.16853537559509277,
      "num_tokens": 46844457.0,
      "step": 18480
    },
    {
      "entropy": 6.165822982788086,
      "epoch": 2.1332948643969996,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004552963033022365,
      "loss": 5.8619,
      "mean_token_accuracy": 0.1780229240655899,
      "num_tokens": 46857256.0,
      "step": 18485
    },
    {
      "entropy": 6.2471390724182125,
      "epoch": 2.133871898442008,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045527161428021706,
      "loss": 5.9491,
      "mean_token_accuracy": 0.17000543624162673,
      "num_tokens": 46870971.0,
      "step": 18490
    },
    {
      "entropy": 6.2426300048828125,
      "epoch": 2.134448932487017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004552469191948514,
      "loss": 5.9197,
      "mean_token_accuracy": 0.17992283403873444,
      "num_tokens": 46882911.0,
      "step": 18495
    },
    {
      "entropy": 6.23798975944519,
      "epoch": 2.1350259665320253,
      "grad_norm": 2.109375,
      "learning_rate": 0.0004552222180469702,
      "loss": 5.931,
      "mean_token_accuracy": 0.17585217952728271,
      "num_tokens": 46895927.0,
      "step": 18500
    },
    {
      "entropy": 6.237430667877197,
      "epoch": 2.135603000577034,
      "grad_norm": 2.859375,
      "learning_rate": 0.00045519751083740413,
      "loss": 5.8958,
      "mean_token_accuracy": 0.17870725244283675,
      "num_tokens": 46908210.0,
      "step": 18505
    },
    {
      "entropy": 6.239075136184693,
      "epoch": 2.1361800346220425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045517279756698435,
      "loss": 5.9672,
      "mean_token_accuracy": 0.18029032796621322,
      "num_tokens": 46920197.0,
      "step": 18510
    },
    {
      "entropy": 6.254026937484741,
      "epoch": 2.1367570686670514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045514807823654203,
      "loss": 5.8944,
      "mean_token_accuracy": 0.18046819418668747,
      "num_tokens": 46932628.0,
      "step": 18515
    },
    {
      "entropy": 6.215092468261719,
      "epoch": 2.13733410271206,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004551233528469086,
      "loss": 5.9229,
      "mean_token_accuracy": 0.16904639899730683,
      "num_tokens": 46945479.0,
      "step": 18520
    },
    {
      "entropy": 6.200603246688843,
      "epoch": 2.1379111367570687,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045509862139891553,
      "loss": 5.9086,
      "mean_token_accuracy": 0.1746812269091606,
      "num_tokens": 46958178.0,
      "step": 18525
    },
    {
      "entropy": 6.278514242172241,
      "epoch": 2.138488170802077,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045507388389339496,
      "loss": 6.0653,
      "mean_token_accuracy": 0.16110795587301255,
      "num_tokens": 46972959.0,
      "step": 18530
    },
    {
      "entropy": 6.296849870681763,
      "epoch": 2.139065204847086,
      "grad_norm": 1.125,
      "learning_rate": 0.00045504914033117866,
      "loss": 5.9954,
      "mean_token_accuracy": 0.17448684722185134,
      "num_tokens": 46985506.0,
      "step": 18535
    },
    {
      "entropy": 6.149590063095093,
      "epoch": 2.139642238892095,
      "grad_norm": 1.0,
      "learning_rate": 0.00045502439071309897,
      "loss": 5.9199,
      "mean_token_accuracy": 0.17310387939214705,
      "num_tokens": 46997928.0,
      "step": 18540
    },
    {
      "entropy": 6.111760044097901,
      "epoch": 2.1402192729371032,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045499963503998835,
      "loss": 5.7738,
      "mean_token_accuracy": 0.1954570844769478,
      "num_tokens": 47010097.0,
      "step": 18545
    },
    {
      "entropy": 6.226894521713257,
      "epoch": 2.140796306982112,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004549748733126794,
      "loss": 5.9459,
      "mean_token_accuracy": 0.1732877790927887,
      "num_tokens": 47021230.0,
      "step": 18550
    },
    {
      "entropy": 6.2623724937438965,
      "epoch": 2.1413733410271205,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000454950105532005,
      "loss": 5.9899,
      "mean_token_accuracy": 0.17608155757188798,
      "num_tokens": 47034433.0,
      "step": 18555
    },
    {
      "entropy": 6.264982652664185,
      "epoch": 2.1419503750721294,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045492533169879816,
      "loss": 5.9194,
      "mean_token_accuracy": 0.17457908540964126,
      "num_tokens": 47046824.0,
      "step": 18560
    },
    {
      "entropy": 6.1724879264831545,
      "epoch": 2.142527409117138,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045490055181389216,
      "loss": 5.8902,
      "mean_token_accuracy": 0.18058053702116011,
      "num_tokens": 47058868.0,
      "step": 18565
    },
    {
      "entropy": 6.206031465530396,
      "epoch": 2.1431044431621467,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045487576587812046,
      "loss": 5.9608,
      "mean_token_accuracy": 0.16733982115983964,
      "num_tokens": 47070898.0,
      "step": 18570
    },
    {
      "entropy": 6.247266578674316,
      "epoch": 2.143681477207155,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045485097389231675,
      "loss": 5.9567,
      "mean_token_accuracy": 0.17399024218320847,
      "num_tokens": 47083739.0,
      "step": 18575
    },
    {
      "entropy": 6.179333114624024,
      "epoch": 2.144258511252164,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004548261758573149,
      "loss": 5.8709,
      "mean_token_accuracy": 0.17647884041070938,
      "num_tokens": 47095595.0,
      "step": 18580
    },
    {
      "entropy": 6.256615877151489,
      "epoch": 2.1448355452971724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004548013717739489,
      "loss": 6.0437,
      "mean_token_accuracy": 0.16919712126255035,
      "num_tokens": 47108326.0,
      "step": 18585
    },
    {
      "entropy": 6.241568183898925,
      "epoch": 2.145412579342181,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004547765616430531,
      "loss": 5.9366,
      "mean_token_accuracy": 0.1774327889084816,
      "num_tokens": 47120494.0,
      "step": 18590
    },
    {
      "entropy": 6.225147438049317,
      "epoch": 2.14598961338719,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004547517454654619,
      "loss": 5.9671,
      "mean_token_accuracy": 0.17375273406505584,
      "num_tokens": 47133116.0,
      "step": 18595
    },
    {
      "entropy": 6.193332386016846,
      "epoch": 2.1465666474321985,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00045472692324201005,
      "loss": 5.905,
      "mean_token_accuracy": 0.17857620865106583,
      "num_tokens": 47145208.0,
      "step": 18600
    },
    {
      "entropy": 6.228554677963257,
      "epoch": 2.1471436814772074,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045470209497353243,
      "loss": 5.9368,
      "mean_token_accuracy": 0.17507996857166291,
      "num_tokens": 47159217.0,
      "step": 18605
    },
    {
      "entropy": 6.249907732009888,
      "epoch": 2.1477207155222158,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004546772606608641,
      "loss": 5.9713,
      "mean_token_accuracy": 0.17326943725347518,
      "num_tokens": 47171781.0,
      "step": 18610
    },
    {
      "entropy": 6.234317064285278,
      "epoch": 2.1482977495672246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546524203048404,
      "loss": 5.9683,
      "mean_token_accuracy": 0.17872635424137115,
      "num_tokens": 47184045.0,
      "step": 18615
    },
    {
      "entropy": 6.179287052154541,
      "epoch": 2.148874783612233,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004546275739062967,
      "loss": 5.8855,
      "mean_token_accuracy": 0.1790724664926529,
      "num_tokens": 47195511.0,
      "step": 18620
    },
    {
      "entropy": 6.163905191421509,
      "epoch": 2.149451817657242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004546027214660688,
      "loss": 5.8806,
      "mean_token_accuracy": 0.18099177926778792,
      "num_tokens": 47206977.0,
      "step": 18625
    },
    {
      "entropy": 6.228483390808106,
      "epoch": 2.1500288517022503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004545778629849926,
      "loss": 5.9068,
      "mean_token_accuracy": 0.17070182263851166,
      "num_tokens": 47218748.0,
      "step": 18630
    },
    {
      "entropy": 6.293354749679565,
      "epoch": 2.150605885747259,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004545529984639042,
      "loss": 6.0288,
      "mean_token_accuracy": 0.1641404002904892,
      "num_tokens": 47230651.0,
      "step": 18635
    },
    {
      "entropy": 6.173253059387207,
      "epoch": 2.1511829197922676,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004545281279036398,
      "loss": 5.8956,
      "mean_token_accuracy": 0.1823137879371643,
      "num_tokens": 47242959.0,
      "step": 18640
    },
    {
      "entropy": 6.168142986297608,
      "epoch": 2.1517599538372765,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004545032513050361,
      "loss": 5.9054,
      "mean_token_accuracy": 0.17626358568668365,
      "num_tokens": 47255113.0,
      "step": 18645
    },
    {
      "entropy": 6.200266122817993,
      "epoch": 2.152336987882285,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004544783686689296,
      "loss": 6.0133,
      "mean_token_accuracy": 0.17102801501750947,
      "num_tokens": 47266873.0,
      "step": 18650
    },
    {
      "entropy": 6.3203010082244875,
      "epoch": 2.1529140219272938,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045445347999615736,
      "loss": 5.9827,
      "mean_token_accuracy": 0.1752153605222702,
      "num_tokens": 47278685.0,
      "step": 18655
    },
    {
      "entropy": 6.213234853744507,
      "epoch": 2.153491055972302,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045442858528755653,
      "loss": 5.9814,
      "mean_token_accuracy": 0.18181101679801942,
      "num_tokens": 47291553.0,
      "step": 18660
    },
    {
      "entropy": 6.207487154006958,
      "epoch": 2.154068090017311,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045440368454396435,
      "loss": 5.8891,
      "mean_token_accuracy": 0.17823161482810973,
      "num_tokens": 47303890.0,
      "step": 18665
    },
    {
      "entropy": 6.203156280517578,
      "epoch": 2.15464512406232,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004543787777662183,
      "loss": 5.8939,
      "mean_token_accuracy": 0.1782517820596695,
      "num_tokens": 47316990.0,
      "step": 18670
    },
    {
      "entropy": 6.206212997436523,
      "epoch": 2.1552221581073283,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045435386495515617,
      "loss": 5.9626,
      "mean_token_accuracy": 0.17625533044338226,
      "num_tokens": 47330568.0,
      "step": 18675
    },
    {
      "entropy": 6.177602052688599,
      "epoch": 2.155799192152337,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004543289461116159,
      "loss": 5.9156,
      "mean_token_accuracy": 0.17385042011737822,
      "num_tokens": 47343563.0,
      "step": 18680
    },
    {
      "entropy": 6.258643531799317,
      "epoch": 2.1563762261973456,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004543040212364356,
      "loss": 5.9313,
      "mean_token_accuracy": 0.17511464655399323,
      "num_tokens": 47356272.0,
      "step": 18685
    },
    {
      "entropy": 6.222632789611817,
      "epoch": 2.1569532602423545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004542790903304536,
      "loss": 5.954,
      "mean_token_accuracy": 0.1678279832005501,
      "num_tokens": 47370244.0,
      "step": 18690
    },
    {
      "entropy": 6.234720897674561,
      "epoch": 2.157530294287363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004542541533945085,
      "loss": 5.9204,
      "mean_token_accuracy": 0.18038647323846818,
      "num_tokens": 47383120.0,
      "step": 18695
    },
    {
      "entropy": 6.2574015140533445,
      "epoch": 2.1581073283323717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045422921042943885,
      "loss": 6.026,
      "mean_token_accuracy": 0.1677708551287651,
      "num_tokens": 47397274.0,
      "step": 18700
    },
    {
      "entropy": 6.178472709655762,
      "epoch": 2.15868436237738,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000454204261436084,
      "loss": 5.8899,
      "mean_token_accuracy": 0.17316479980945587,
      "num_tokens": 47409611.0,
      "step": 18705
    },
    {
      "entropy": 6.203462886810303,
      "epoch": 2.159261396422389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045417930641528255,
      "loss": 5.9202,
      "mean_token_accuracy": 0.1778599888086319,
      "num_tokens": 47421435.0,
      "step": 18710
    },
    {
      "entropy": 6.276810979843139,
      "epoch": 2.1598384304673974,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045415434536787424,
      "loss": 5.9995,
      "mean_token_accuracy": 0.17613883763551713,
      "num_tokens": 47434749.0,
      "step": 18715
    },
    {
      "entropy": 6.20872278213501,
      "epoch": 2.1604154645124063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004541293782946985,
      "loss": 5.9431,
      "mean_token_accuracy": 0.17725253254175186,
      "num_tokens": 47446477.0,
      "step": 18720
    },
    {
      "entropy": 6.117638778686524,
      "epoch": 2.1609924985574147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004541044051965952,
      "loss": 5.8292,
      "mean_token_accuracy": 0.18287974447011948,
      "num_tokens": 47459547.0,
      "step": 18725
    },
    {
      "entropy": 6.269925165176391,
      "epoch": 2.1615695326024236,
      "grad_norm": 1.125,
      "learning_rate": 0.0004540794260744041,
      "loss": 6.0191,
      "mean_token_accuracy": 0.16641358882188798,
      "num_tokens": 47472294.0,
      "step": 18730
    },
    {
      "entropy": 6.23092622756958,
      "epoch": 2.162146566647432,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004540544409289655,
      "loss": 5.9058,
      "mean_token_accuracy": 0.17685572654008866,
      "num_tokens": 47484240.0,
      "step": 18735
    },
    {
      "entropy": 6.214578866958618,
      "epoch": 2.162723600692441,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004540294497611197,
      "loss": 5.9856,
      "mean_token_accuracy": 0.17463434487581253,
      "num_tokens": 47498179.0,
      "step": 18740
    },
    {
      "entropy": 6.20415768623352,
      "epoch": 2.1633006347374497,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045400445257170725,
      "loss": 5.9266,
      "mean_token_accuracy": 0.17393019199371337,
      "num_tokens": 47511354.0,
      "step": 18745
    },
    {
      "entropy": 6.272518587112427,
      "epoch": 2.163877668782458,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004539794493615689,
      "loss": 6.0156,
      "mean_token_accuracy": 0.1667408227920532,
      "num_tokens": 47523589.0,
      "step": 18750
    },
    {
      "entropy": 6.136549711227417,
      "epoch": 2.164454702827467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004539544401315458,
      "loss": 5.9327,
      "mean_token_accuracy": 0.17985440343618392,
      "num_tokens": 47536784.0,
      "step": 18755
    },
    {
      "entropy": 6.226091432571411,
      "epoch": 2.1650317368724754,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000453929424882479,
      "loss": 5.9236,
      "mean_token_accuracy": 0.17071688771247864,
      "num_tokens": 47550560.0,
      "step": 18760
    },
    {
      "entropy": 6.227948236465454,
      "epoch": 2.1656087709174843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045390440361520987,
      "loss": 5.895,
      "mean_token_accuracy": 0.17807429581880568,
      "num_tokens": 47564069.0,
      "step": 18765
    },
    {
      "entropy": 6.242389440536499,
      "epoch": 2.1661858049624927,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004538793763305799,
      "loss": 5.9733,
      "mean_token_accuracy": 0.16201029866933822,
      "num_tokens": 47577570.0,
      "step": 18770
    },
    {
      "entropy": 6.212618494033814,
      "epoch": 2.1667628390075016,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045385434302943104,
      "loss": 6.0142,
      "mean_token_accuracy": 0.17255715876817704,
      "num_tokens": 47590199.0,
      "step": 18775
    },
    {
      "entropy": 6.259509897232055,
      "epoch": 2.16733987305251,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004538293037126052,
      "loss": 5.9149,
      "mean_token_accuracy": 0.17407530844211577,
      "num_tokens": 47601697.0,
      "step": 18780
    },
    {
      "entropy": 6.272859239578247,
      "epoch": 2.167916907097519,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045380425838094444,
      "loss": 5.9577,
      "mean_token_accuracy": 0.1750097766518593,
      "num_tokens": 47615331.0,
      "step": 18785
    },
    {
      "entropy": 6.176296091079712,
      "epoch": 2.1684939411425272,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045377920703529133,
      "loss": 5.9098,
      "mean_token_accuracy": 0.17534357011318208,
      "num_tokens": 47628455.0,
      "step": 18790
    },
    {
      "entropy": 6.254702377319336,
      "epoch": 2.169070975187536,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004537541496764885,
      "loss": 5.9253,
      "mean_token_accuracy": 0.17553993165493012,
      "num_tokens": 47642078.0,
      "step": 18795
    },
    {
      "entropy": 6.242267417907715,
      "epoch": 2.1696480092325445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004537290863053786,
      "loss": 5.9378,
      "mean_token_accuracy": 0.1725284069776535,
      "num_tokens": 47653672.0,
      "step": 18800
    },
    {
      "entropy": 6.090953302383423,
      "epoch": 2.1702250432775534,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00045370401692280455,
      "loss": 5.7931,
      "mean_token_accuracy": 0.18746485859155654,
      "num_tokens": 47666085.0,
      "step": 18805
    },
    {
      "entropy": 6.25534143447876,
      "epoch": 2.170802077322562,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045367894152960976,
      "loss": 6.0072,
      "mean_token_accuracy": 0.1746744379401207,
      "num_tokens": 47678180.0,
      "step": 18810
    },
    {
      "entropy": 6.231589221954346,
      "epoch": 2.1713791113675707,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045365386012663744,
      "loss": 5.9397,
      "mean_token_accuracy": 0.17218563407659532,
      "num_tokens": 47690631.0,
      "step": 18815
    },
    {
      "entropy": 6.2311194896697994,
      "epoch": 2.1719561454125795,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045362877271473137,
      "loss": 5.8993,
      "mean_token_accuracy": 0.17505298256874086,
      "num_tokens": 47702577.0,
      "step": 18820
    },
    {
      "entropy": 6.198976993560791,
      "epoch": 2.172533179457588,
      "grad_norm": 1.25,
      "learning_rate": 0.00045360367929473517,
      "loss": 5.9298,
      "mean_token_accuracy": 0.18071759790182113,
      "num_tokens": 47714683.0,
      "step": 18825
    },
    {
      "entropy": 6.210054349899292,
      "epoch": 2.173110213502597,
      "grad_norm": 1.0,
      "learning_rate": 0.000453578579867493,
      "loss": 5.8748,
      "mean_token_accuracy": 0.17670900523662567,
      "num_tokens": 47726202.0,
      "step": 18830
    },
    {
      "entropy": 6.163262987136841,
      "epoch": 2.1736872475476052,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045355347443384896,
      "loss": 5.914,
      "mean_token_accuracy": 0.17876358926296235,
      "num_tokens": 47740178.0,
      "step": 18835
    },
    {
      "entropy": 6.221033239364624,
      "epoch": 2.174264281592614,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045352836299464755,
      "loss": 5.9323,
      "mean_token_accuracy": 0.17617103904485704,
      "num_tokens": 47752710.0,
      "step": 18840
    },
    {
      "entropy": 6.205242776870728,
      "epoch": 2.1748413156376225,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004535032455507333,
      "loss": 5.8558,
      "mean_token_accuracy": 0.17509319633245468,
      "num_tokens": 47765292.0,
      "step": 18845
    },
    {
      "entropy": 6.215845823287964,
      "epoch": 2.1754183496826314,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045347812210295107,
      "loss": 5.9082,
      "mean_token_accuracy": 0.17581940293312073,
      "num_tokens": 47777799.0,
      "step": 18850
    },
    {
      "entropy": 6.217535352706909,
      "epoch": 2.17599538372764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045345299265214583,
      "loss": 5.9047,
      "mean_token_accuracy": 0.1748897075653076,
      "num_tokens": 47790705.0,
      "step": 18855
    },
    {
      "entropy": 6.152811431884766,
      "epoch": 2.1765724177726486,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045342785719916284,
      "loss": 5.8459,
      "mean_token_accuracy": 0.17450116872787474,
      "num_tokens": 47802169.0,
      "step": 18860
    },
    {
      "entropy": 6.069148826599121,
      "epoch": 2.177149451817657,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045340271574484755,
      "loss": 5.8052,
      "mean_token_accuracy": 0.18310066014528276,
      "num_tokens": 47815832.0,
      "step": 18865
    },
    {
      "entropy": 6.290272855758667,
      "epoch": 2.177726485862666,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004533775682900454,
      "loss": 6.014,
      "mean_token_accuracy": 0.17042950987815858,
      "num_tokens": 47828967.0,
      "step": 18870
    },
    {
      "entropy": 6.222682809829712,
      "epoch": 2.1783035199076743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004533524148356025,
      "loss": 5.8551,
      "mean_token_accuracy": 0.1862866297364235,
      "num_tokens": 47842114.0,
      "step": 18875
    },
    {
      "entropy": 6.076592254638672,
      "epoch": 2.178880553952683,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004533272553823646,
      "loss": 5.8291,
      "mean_token_accuracy": 0.17668368220329284,
      "num_tokens": 47855808.0,
      "step": 18880
    },
    {
      "entropy": 6.250540828704834,
      "epoch": 2.179457587997692,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045330208993117816,
      "loss": 5.9459,
      "mean_token_accuracy": 0.17007622569799424,
      "num_tokens": 47866866.0,
      "step": 18885
    },
    {
      "entropy": 6.22902479171753,
      "epoch": 2.1800346220427005,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004532769184828894,
      "loss": 5.9179,
      "mean_token_accuracy": 0.17572437971830368,
      "num_tokens": 47878823.0,
      "step": 18890
    },
    {
      "entropy": 6.200373554229737,
      "epoch": 2.1806116560877093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004532517410383451,
      "loss": 5.9678,
      "mean_token_accuracy": 0.17269928604364396,
      "num_tokens": 47890512.0,
      "step": 18895
    },
    {
      "entropy": 6.2269518852233885,
      "epoch": 2.1811886901327178,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000453226557598392,
      "loss": 5.9764,
      "mean_token_accuracy": 0.16899292171001434,
      "num_tokens": 47903728.0,
      "step": 18900
    },
    {
      "entropy": 6.238075494766235,
      "epoch": 2.1817657241777266,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004532013681638771,
      "loss": 5.9799,
      "mean_token_accuracy": 0.17065757811069487,
      "num_tokens": 47914742.0,
      "step": 18905
    },
    {
      "entropy": 6.262040901184082,
      "epoch": 2.182342758222735,
      "grad_norm": 1.875,
      "learning_rate": 0.0004531761727356478,
      "loss": 5.9637,
      "mean_token_accuracy": 0.1727651610970497,
      "num_tokens": 47926737.0,
      "step": 18910
    },
    {
      "entropy": 6.238742256164551,
      "epoch": 2.182919792267744,
      "grad_norm": 1.0,
      "learning_rate": 0.0004531509713145514,
      "loss": 5.9314,
      "mean_token_accuracy": 0.174980790913105,
      "num_tokens": 47941207.0,
      "step": 18915
    },
    {
      "entropy": 6.208298969268799,
      "epoch": 2.1834968263127523,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045312576390143557,
      "loss": 5.9828,
      "mean_token_accuracy": 0.16858302056789398,
      "num_tokens": 47953640.0,
      "step": 18920
    },
    {
      "entropy": 6.228661108016968,
      "epoch": 2.184073860357761,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045310055049714815,
      "loss": 5.9659,
      "mean_token_accuracy": 0.17834703028202056,
      "num_tokens": 47966399.0,
      "step": 18925
    },
    {
      "entropy": 6.241978883743286,
      "epoch": 2.1846508944027696,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045307533110253726,
      "loss": 5.9175,
      "mean_token_accuracy": 0.17427961975336076,
      "num_tokens": 47979521.0,
      "step": 18930
    },
    {
      "entropy": 6.270679616928101,
      "epoch": 2.1852279284477785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045305010571845096,
      "loss": 5.9875,
      "mean_token_accuracy": 0.17076631486415864,
      "num_tokens": 47992780.0,
      "step": 18935
    },
    {
      "entropy": 6.319071340560913,
      "epoch": 2.185804962492787,
      "grad_norm": 1.375,
      "learning_rate": 0.0004530248743457378,
      "loss": 6.0078,
      "mean_token_accuracy": 0.17148044854402542,
      "num_tokens": 48005548.0,
      "step": 18940
    },
    {
      "entropy": 6.200914764404297,
      "epoch": 2.1863819965377957,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529996369852465,
      "loss": 5.8982,
      "mean_token_accuracy": 0.17534489631652833,
      "num_tokens": 48017352.0,
      "step": 18945
    },
    {
      "entropy": 6.07051453590393,
      "epoch": 2.1869590305828046,
      "grad_norm": 1.0,
      "learning_rate": 0.00045297439363782576,
      "loss": 5.7113,
      "mean_token_accuracy": 0.19109055697917937,
      "num_tokens": 48029252.0,
      "step": 18950
    },
    {
      "entropy": 6.199999475479126,
      "epoch": 2.187536064627813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529491443043247,
      "loss": 5.9418,
      "mean_token_accuracy": 0.1842118799686432,
      "num_tokens": 48041534.0,
      "step": 18955
    },
    {
      "entropy": 6.1737473487854,
      "epoch": 2.188113098672822,
      "grad_norm": 1.125,
      "learning_rate": 0.0004529238889855926,
      "loss": 5.9869,
      "mean_token_accuracy": 0.1713000401854515,
      "num_tokens": 48054396.0,
      "step": 18960
    },
    {
      "entropy": 6.286145353317261,
      "epoch": 2.1886901327178303,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004528986276824789,
      "loss": 5.9492,
      "mean_token_accuracy": 0.17444169223308564,
      "num_tokens": 48066785.0,
      "step": 18965
    },
    {
      "entropy": 6.1863932609558105,
      "epoch": 2.189267166762839,
      "grad_norm": 1.0,
      "learning_rate": 0.0004528733603958331,
      "loss": 5.9085,
      "mean_token_accuracy": 0.18538776636123658,
      "num_tokens": 48079837.0,
      "step": 18970
    },
    {
      "entropy": 6.194906759262085,
      "epoch": 2.1898442008078476,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004528480871265053,
      "loss": 5.87,
      "mean_token_accuracy": 0.17983012944459914,
      "num_tokens": 48091670.0,
      "step": 18975
    },
    {
      "entropy": 6.255014896392822,
      "epoch": 2.1904212348528564,
      "grad_norm": 2.15625,
      "learning_rate": 0.00045282280787534537,
      "loss": 5.99,
      "mean_token_accuracy": 0.17368623316287995,
      "num_tokens": 48105261.0,
      "step": 18980
    },
    {
      "entropy": 6.2829742431640625,
      "epoch": 2.190998268897865,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004527975226432036,
      "loss": 5.9153,
      "mean_token_accuracy": 0.17215203493833542,
      "num_tokens": 48118272.0,
      "step": 18985
    },
    {
      "entropy": 6.190290689468384,
      "epoch": 2.1915753029428737,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045277223143093057,
      "loss": 5.9195,
      "mean_token_accuracy": 0.17416684180498124,
      "num_tokens": 48130413.0,
      "step": 18990
    },
    {
      "entropy": 6.1797760963439945,
      "epoch": 2.192152336987882,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045274693423937674,
      "loss": 5.8944,
      "mean_token_accuracy": 0.1710223600268364,
      "num_tokens": 48143308.0,
      "step": 18995
    },
    {
      "entropy": 6.119444990158081,
      "epoch": 2.192729371032891,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045272163106939314,
      "loss": 5.8411,
      "mean_token_accuracy": 0.18233658671379088,
      "num_tokens": 48155832.0,
      "step": 19000
    },
    {
      "entropy": 6.218946266174316,
      "epoch": 2.1933064050778994,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045269632192183076,
      "loss": 6.0152,
      "mean_token_accuracy": 0.1638078957796097,
      "num_tokens": 48168390.0,
      "step": 19005
    },
    {
      "entropy": 6.267431020736694,
      "epoch": 2.1938834391229083,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045267100679754075,
      "loss": 5.9599,
      "mean_token_accuracy": 0.17242570966482162,
      "num_tokens": 48180918.0,
      "step": 19010
    },
    {
      "entropy": 6.216310119628906,
      "epoch": 2.1944604731679167,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004526456856973748,
      "loss": 5.9549,
      "mean_token_accuracy": 0.17398718148469924,
      "num_tokens": 48194251.0,
      "step": 19015
    },
    {
      "entropy": 6.266015577316284,
      "epoch": 2.1950375072129256,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004526203586221844,
      "loss": 5.9513,
      "mean_token_accuracy": 0.17341589480638503,
      "num_tokens": 48206704.0,
      "step": 19020
    },
    {
      "entropy": 6.222143745422363,
      "epoch": 2.1956145412579344,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045259502557282145,
      "loss": 5.9304,
      "mean_token_accuracy": 0.17723745554685594,
      "num_tokens": 48219390.0,
      "step": 19025
    },
    {
      "entropy": 6.134809350967407,
      "epoch": 2.196191575302943,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004525696865501381,
      "loss": 5.8949,
      "mean_token_accuracy": 0.17256021797657012,
      "num_tokens": 48231555.0,
      "step": 19030
    },
    {
      "entropy": 6.243925762176514,
      "epoch": 2.1967686093479517,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004525443415549865,
      "loss": 5.9766,
      "mean_token_accuracy": 0.17290915101766585,
      "num_tokens": 48243763.0,
      "step": 19035
    },
    {
      "entropy": 6.211654376983643,
      "epoch": 2.19734564339296,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00045251899058821915,
      "loss": 5.9834,
      "mean_token_accuracy": 0.17231059968471527,
      "num_tokens": 48257567.0,
      "step": 19040
    },
    {
      "entropy": 6.19317660331726,
      "epoch": 2.197922677437969,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004524936336506887,
      "loss": 5.8512,
      "mean_token_accuracy": 0.18214067071676254,
      "num_tokens": 48270313.0,
      "step": 19045
    },
    {
      "entropy": 6.189791822433472,
      "epoch": 2.1984997114829774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004524682707432482,
      "loss": 5.8669,
      "mean_token_accuracy": 0.17424180805683137,
      "num_tokens": 48282768.0,
      "step": 19050
    },
    {
      "entropy": 6.139924001693726,
      "epoch": 2.1990767455279863,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045244290186675034,
      "loss": 5.8626,
      "mean_token_accuracy": 0.17862701565027236,
      "num_tokens": 48293861.0,
      "step": 19055
    },
    {
      "entropy": 6.194202089309693,
      "epoch": 2.1996537795729947,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004524175270220489,
      "loss": 5.8946,
      "mean_token_accuracy": 0.17505065798759462,
      "num_tokens": 48306062.0,
      "step": 19060
    },
    {
      "entropy": 6.251491737365723,
      "epoch": 2.2002308136180035,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045239214620999683,
      "loss": 5.9766,
      "mean_token_accuracy": 0.16668111830949783,
      "num_tokens": 48318777.0,
      "step": 19065
    },
    {
      "entropy": 6.207370710372925,
      "epoch": 2.200807847663012,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004523667594314481,
      "loss": 5.9334,
      "mean_token_accuracy": 0.17182350158691406,
      "num_tokens": 48330575.0,
      "step": 19070
    },
    {
      "entropy": 6.115412998199463,
      "epoch": 2.201384881708021,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045234136668725655,
      "loss": 5.7773,
      "mean_token_accuracy": 0.18870446383953093,
      "num_tokens": 48343980.0,
      "step": 19075
    },
    {
      "entropy": 6.213305044174194,
      "epoch": 2.2019619157530292,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045231596797827616,
      "loss": 5.8388,
      "mean_token_accuracy": 0.1812451496720314,
      "num_tokens": 48356140.0,
      "step": 19080
    },
    {
      "entropy": 6.174299716949463,
      "epoch": 2.202538949798038,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045229056330536134,
      "loss": 5.9042,
      "mean_token_accuracy": 0.18119534999132156,
      "num_tokens": 48367756.0,
      "step": 19085
    },
    {
      "entropy": 6.168049669265747,
      "epoch": 2.2031159838430465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045226515266936644,
      "loss": 5.9393,
      "mean_token_accuracy": 0.16965750306844712,
      "num_tokens": 48381637.0,
      "step": 19090
    },
    {
      "entropy": 6.230480575561524,
      "epoch": 2.2036930178880554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004522397360711462,
      "loss": 5.8748,
      "mean_token_accuracy": 0.17553682923316954,
      "num_tokens": 48394593.0,
      "step": 19095
    },
    {
      "entropy": 6.165048265457154,
      "epoch": 2.2042700519330642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004522143135115555,
      "loss": 5.8574,
      "mean_token_accuracy": 0.1800720989704132,
      "num_tokens": 48408555.0,
      "step": 19100
    },
    {
      "entropy": 6.1731829166412355,
      "epoch": 2.2048470859780727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045218888499144933,
      "loss": 5.8946,
      "mean_token_accuracy": 0.1775633677840233,
      "num_tokens": 48420359.0,
      "step": 19105
    },
    {
      "entropy": 6.254141187667846,
      "epoch": 2.2054241200230815,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045216345051168314,
      "loss": 6.0368,
      "mean_token_accuracy": 0.16812050044536592,
      "num_tokens": 48432516.0,
      "step": 19110
    },
    {
      "entropy": 6.192972087860108,
      "epoch": 2.20600115406809,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004521380100731122,
      "loss": 5.9272,
      "mean_token_accuracy": 0.17607006281614304,
      "num_tokens": 48444355.0,
      "step": 19115
    },
    {
      "entropy": 6.170709800720215,
      "epoch": 2.206578188113099,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045211256367659234,
      "loss": 5.8752,
      "mean_token_accuracy": 0.18138357400894164,
      "num_tokens": 48458341.0,
      "step": 19120
    },
    {
      "entropy": 6.17162766456604,
      "epoch": 2.207155222158107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004520871113229793,
      "loss": 5.9257,
      "mean_token_accuracy": 0.17600769847631453,
      "num_tokens": 48471473.0,
      "step": 19125
    },
    {
      "entropy": 6.188865756988525,
      "epoch": 2.207732256203116,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045206165301312927,
      "loss": 5.9205,
      "mean_token_accuracy": 0.1756698504090309,
      "num_tokens": 48483149.0,
      "step": 19130
    },
    {
      "entropy": 6.243536329269409,
      "epoch": 2.2083092902481245,
      "grad_norm": 1.25,
      "learning_rate": 0.0004520361887478985,
      "loss": 5.9099,
      "mean_token_accuracy": 0.17231071293354033,
      "num_tokens": 48495617.0,
      "step": 19135
    },
    {
      "entropy": 6.098401594161987,
      "epoch": 2.2088863242931334,
      "grad_norm": 3.1875,
      "learning_rate": 0.0004520107185281435,
      "loss": 5.8987,
      "mean_token_accuracy": 0.17789782136678695,
      "num_tokens": 48509391.0,
      "step": 19140
    },
    {
      "entropy": 6.089808416366577,
      "epoch": 2.209463358338142,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004519852423547208,
      "loss": 5.8373,
      "mean_token_accuracy": 0.18326293528079987,
      "num_tokens": 48523127.0,
      "step": 19145
    },
    {
      "entropy": 6.135551261901855,
      "epoch": 2.2100403923831506,
      "grad_norm": 2.65625,
      "learning_rate": 0.0004519597602284875,
      "loss": 5.8662,
      "mean_token_accuracy": 0.1830893412232399,
      "num_tokens": 48536797.0,
      "step": 19150
    },
    {
      "entropy": 6.254802131652832,
      "epoch": 2.210617426428159,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519342721503005,
      "loss": 5.9156,
      "mean_token_accuracy": 0.1770019993185997,
      "num_tokens": 48549088.0,
      "step": 19155
    },
    {
      "entropy": 6.185900688171387,
      "epoch": 2.211194460473168,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004519087781210171,
      "loss": 5.8464,
      "mean_token_accuracy": 0.1776304841041565,
      "num_tokens": 48560451.0,
      "step": 19160
    },
    {
      "entropy": 6.117572355270386,
      "epoch": 2.2117714945181763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000451883278141495,
      "loss": 5.8671,
      "mean_token_accuracy": 0.17668378055095674,
      "num_tokens": 48572870.0,
      "step": 19165
    },
    {
      "entropy": 6.198688268661499,
      "epoch": 2.212348528563185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045185777221259157,
      "loss": 5.9052,
      "mean_token_accuracy": 0.1720116391777992,
      "num_tokens": 48585909.0,
      "step": 19170
    },
    {
      "entropy": 6.227552270889282,
      "epoch": 2.212925562608194,
      "grad_norm": 1.125,
      "learning_rate": 0.0004518322603351648,
      "loss": 5.9127,
      "mean_token_accuracy": 0.18233373910188674,
      "num_tokens": 48599157.0,
      "step": 19175
    },
    {
      "entropy": 6.1833775520324705,
      "epoch": 2.2135025966532025,
      "grad_norm": 1.5,
      "learning_rate": 0.00045180674251007287,
      "loss": 5.8566,
      "mean_token_accuracy": 0.1842576578259468,
      "num_tokens": 48611740.0,
      "step": 19180
    },
    {
      "entropy": 6.152484560012818,
      "epoch": 2.2140796306982113,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045178121873817395,
      "loss": 5.9721,
      "mean_token_accuracy": 0.1699468493461609,
      "num_tokens": 48624893.0,
      "step": 19185
    },
    {
      "entropy": 6.189026546478272,
      "epoch": 2.2146566647432198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004517556890203265,
      "loss": 5.9497,
      "mean_token_accuracy": 0.16636952012777328,
      "num_tokens": 48636414.0,
      "step": 19190
    },
    {
      "entropy": 6.176068592071533,
      "epoch": 2.2152336987882286,
      "grad_norm": 1.90625,
      "learning_rate": 0.00045173015335738925,
      "loss": 5.8886,
      "mean_token_accuracy": 0.17687400579452514,
      "num_tokens": 48649936.0,
      "step": 19195
    },
    {
      "entropy": 6.219943332672119,
      "epoch": 2.215810732833237,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045170461175022116,
      "loss": 5.9104,
      "mean_token_accuracy": 0.1829262226819992,
      "num_tokens": 48662857.0,
      "step": 19200
    },
    {
      "entropy": 6.223324632644653,
      "epoch": 2.216387766878246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004516790641996812,
      "loss": 5.8683,
      "mean_token_accuracy": 0.18157433569431305,
      "num_tokens": 48675087.0,
      "step": 19205
    },
    {
      "entropy": 6.131745100021362,
      "epoch": 2.2169648009232543,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004516535107066286,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1740524262189865,
      "num_tokens": 48687575.0,
      "step": 19210
    },
    {
      "entropy": 6.168126440048217,
      "epoch": 2.217541834968263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045162795127192296,
      "loss": 5.8762,
      "mean_token_accuracy": 0.17217035144567489,
      "num_tokens": 48701115.0,
      "step": 19215
    },
    {
      "entropy": 6.251860284805298,
      "epoch": 2.2181188690132716,
      "grad_norm": 1.109375,
      "learning_rate": 0.000451602385896424,
      "loss": 6.0011,
      "mean_token_accuracy": 0.16828219592571259,
      "num_tokens": 48713738.0,
      "step": 19220
    },
    {
      "entropy": 6.2112242698669435,
      "epoch": 2.2186959030582805,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045157681458099145,
      "loss": 5.8076,
      "mean_token_accuracy": 0.18412246257066728,
      "num_tokens": 48725282.0,
      "step": 19225
    },
    {
      "entropy": 6.170522260665893,
      "epoch": 2.2192729371032893,
      "grad_norm": 1.125,
      "learning_rate": 0.0004515512373264854,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17799303233623504,
      "num_tokens": 48738476.0,
      "step": 19230
    },
    {
      "entropy": 6.189772510528565,
      "epoch": 2.2198499711482977,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045152565413376623,
      "loss": 5.9089,
      "mean_token_accuracy": 0.17495458871126174,
      "num_tokens": 48751447.0,
      "step": 19235
    },
    {
      "entropy": 6.193010902404785,
      "epoch": 2.2204270051933066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004515000650036944,
      "loss": 5.9375,
      "mean_token_accuracy": 0.17232024669647217,
      "num_tokens": 48764124.0,
      "step": 19240
    },
    {
      "entropy": 6.22163953781128,
      "epoch": 2.221004039238315,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004514744699371305,
      "loss": 5.9958,
      "mean_token_accuracy": 0.1685050144791603,
      "num_tokens": 48776414.0,
      "step": 19245
    },
    {
      "entropy": 6.254934215545655,
      "epoch": 2.221581073283324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045144886893493547,
      "loss": 5.9314,
      "mean_token_accuracy": 0.17561885863542556,
      "num_tokens": 48789318.0,
      "step": 19250
    },
    {
      "entropy": 6.1196118831634525,
      "epoch": 2.2221581073283323,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045142326199797047,
      "loss": 5.8587,
      "mean_token_accuracy": 0.17800619453191757,
      "num_tokens": 48801773.0,
      "step": 19255
    },
    {
      "entropy": 6.239759874343872,
      "epoch": 2.222735141373341,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045139764912709656,
      "loss": 5.861,
      "mean_token_accuracy": 0.17993413656949997,
      "num_tokens": 48814137.0,
      "step": 19260
    },
    {
      "entropy": 6.19697527885437,
      "epoch": 2.2233121754183496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004513720303231754,
      "loss": 5.8682,
      "mean_token_accuracy": 0.17760514318943024,
      "num_tokens": 48826129.0,
      "step": 19265
    },
    {
      "entropy": 6.195125102996826,
      "epoch": 2.2238892094633584,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045134640558706864,
      "loss": 5.9495,
      "mean_token_accuracy": 0.17445218861103057,
      "num_tokens": 48838930.0,
      "step": 19270
    },
    {
      "entropy": 6.245221185684204,
      "epoch": 2.224466243508367,
      "grad_norm": 1.109375,
      "learning_rate": 0.000451320774919638,
      "loss": 5.9168,
      "mean_token_accuracy": 0.17168458849191665,
      "num_tokens": 48851506.0,
      "step": 19275
    },
    {
      "entropy": 6.202529716491699,
      "epoch": 2.2250432775533757,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045129513832174587,
      "loss": 5.8892,
      "mean_token_accuracy": 0.17213271409273148,
      "num_tokens": 48865267.0,
      "step": 19280
    },
    {
      "entropy": 6.183564043045044,
      "epoch": 2.225620311598384,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045126949579425414,
      "loss": 5.9589,
      "mean_token_accuracy": 0.17022151350975037,
      "num_tokens": 48877642.0,
      "step": 19285
    },
    {
      "entropy": 6.328790330886841,
      "epoch": 2.226197345643393,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045124384733802563,
      "loss": 6.0025,
      "mean_token_accuracy": 0.1685243621468544,
      "num_tokens": 48891135.0,
      "step": 19290
    },
    {
      "entropy": 6.246169519424439,
      "epoch": 2.2267743796884014,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004512181929539228,
      "loss": 5.9617,
      "mean_token_accuracy": 0.17567284852266313,
      "num_tokens": 48904361.0,
      "step": 19295
    },
    {
      "entropy": 6.1482072353363035,
      "epoch": 2.2273514137334103,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045119253264280855,
      "loss": 5.9456,
      "mean_token_accuracy": 0.17884223759174347,
      "num_tokens": 48916497.0,
      "step": 19300
    },
    {
      "entropy": 6.253855752944946,
      "epoch": 2.227928447778419,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045116686640554607,
      "loss": 5.9073,
      "mean_token_accuracy": 0.17755719721317292,
      "num_tokens": 48928476.0,
      "step": 19305
    },
    {
      "entropy": 6.225940561294555,
      "epoch": 2.2285054818234276,
      "grad_norm": 1.4375,
      "learning_rate": 0.00045114119424299843,
      "loss": 5.958,
      "mean_token_accuracy": 0.17427153140306473,
      "num_tokens": 48942153.0,
      "step": 19310
    },
    {
      "entropy": 6.248000001907348,
      "epoch": 2.2290825158684364,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004511155161560293,
      "loss": 5.9774,
      "mean_token_accuracy": 0.17155166268348693,
      "num_tokens": 48955313.0,
      "step": 19315
    },
    {
      "entropy": 6.198637628555298,
      "epoch": 2.229659549913445,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045108983214550225,
      "loss": 5.8672,
      "mean_token_accuracy": 0.18007466942071915,
      "num_tokens": 48967875.0,
      "step": 19320
    },
    {
      "entropy": 6.279212284088135,
      "epoch": 2.2302365839584537,
      "grad_norm": 1.0,
      "learning_rate": 0.0004510641422122811,
      "loss": 5.9204,
      "mean_token_accuracy": 0.17138898074626924,
      "num_tokens": 48979534.0,
      "step": 19325
    },
    {
      "entropy": 6.255653953552246,
      "epoch": 2.230813618003462,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045103844635723005,
      "loss": 5.9646,
      "mean_token_accuracy": 0.17417775243520736,
      "num_tokens": 48991425.0,
      "step": 19330
    },
    {
      "entropy": 6.114951229095459,
      "epoch": 2.231390652048471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045101274458121335,
      "loss": 5.8303,
      "mean_token_accuracy": 0.18759170174598694,
      "num_tokens": 49003783.0,
      "step": 19335
    },
    {
      "entropy": 6.261584711074829,
      "epoch": 2.2319676860934794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045098703688509537,
      "loss": 5.9823,
      "mean_token_accuracy": 0.1674163371324539,
      "num_tokens": 49017998.0,
      "step": 19340
    },
    {
      "entropy": 6.276994848251343,
      "epoch": 2.2325447201384883,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004509613232697408,
      "loss": 5.9988,
      "mean_token_accuracy": 0.167509825527668,
      "num_tokens": 49030429.0,
      "step": 19345
    },
    {
      "entropy": 6.2080472946167,
      "epoch": 2.2331217541834967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004509356037360145,
      "loss": 5.9082,
      "mean_token_accuracy": 0.17921029180288314,
      "num_tokens": 49043619.0,
      "step": 19350
    },
    {
      "entropy": 6.251058197021484,
      "epoch": 2.2336987882285055,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004509098782847816,
      "loss": 5.9337,
      "mean_token_accuracy": 0.17255532145500183,
      "num_tokens": 49056674.0,
      "step": 19355
    },
    {
      "entropy": 6.231851053237915,
      "epoch": 2.234275822273514,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004508841469169073,
      "loss": 5.952,
      "mean_token_accuracy": 0.176688514649868,
      "num_tokens": 49069616.0,
      "step": 19360
    },
    {
      "entropy": 6.162151050567627,
      "epoch": 2.234852856318523,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045085840963325716,
      "loss": 5.902,
      "mean_token_accuracy": 0.17886948734521865,
      "num_tokens": 49082001.0,
      "step": 19365
    },
    {
      "entropy": 6.209615707397461,
      "epoch": 2.2354298903635312,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508326664346967,
      "loss": 5.8871,
      "mean_token_accuracy": 0.1740911200642586,
      "num_tokens": 49094433.0,
      "step": 19370
    },
    {
      "entropy": 6.174746799468994,
      "epoch": 2.23600692440854,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045080691732209176,
      "loss": 5.8633,
      "mean_token_accuracy": 0.18208640515804292,
      "num_tokens": 49106152.0,
      "step": 19375
    },
    {
      "entropy": 6.178849172592163,
      "epoch": 2.236583958453549,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045078116229630864,
      "loss": 5.9521,
      "mean_token_accuracy": 0.18309886753559113,
      "num_tokens": 49120402.0,
      "step": 19380
    },
    {
      "entropy": 6.317041301727295,
      "epoch": 2.2371609924985574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045075540135821343,
      "loss": 6.0472,
      "mean_token_accuracy": 0.16531797200441362,
      "num_tokens": 49133571.0,
      "step": 19385
    },
    {
      "entropy": 6.285482406616211,
      "epoch": 2.2377380265435662,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004507296345086725,
      "loss": 6.0433,
      "mean_token_accuracy": 0.16770367175340653,
      "num_tokens": 49147268.0,
      "step": 19390
    },
    {
      "entropy": 6.216909456253052,
      "epoch": 2.2383150605885747,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004507038617485526,
      "loss": 5.8327,
      "mean_token_accuracy": 0.1768430009484291,
      "num_tokens": 49159043.0,
      "step": 19395
    },
    {
      "entropy": 6.187124443054199,
      "epoch": 2.2388920946335835,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045067808307872064,
      "loss": 5.9344,
      "mean_token_accuracy": 0.17853873074054719,
      "num_tokens": 49172164.0,
      "step": 19400
    },
    {
      "entropy": 6.238157081604004,
      "epoch": 2.239469128678592,
      "grad_norm": 1.125,
      "learning_rate": 0.00045065229850004365,
      "loss": 5.9049,
      "mean_token_accuracy": 0.1760144054889679,
      "num_tokens": 49185023.0,
      "step": 19405
    },
    {
      "entropy": 6.2631641864776615,
      "epoch": 2.240046162723601,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004506265080133888,
      "loss": 5.9316,
      "mean_token_accuracy": 0.1767286479473114,
      "num_tokens": 49198411.0,
      "step": 19410
    },
    {
      "entropy": 6.149572944641113,
      "epoch": 2.240623196768609,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045060071161962364,
      "loss": 5.9208,
      "mean_token_accuracy": 0.17646293342113495,
      "num_tokens": 49211162.0,
      "step": 19415
    },
    {
      "entropy": 6.227586889266968,
      "epoch": 2.241200230813618,
      "grad_norm": 1.25,
      "learning_rate": 0.0004505749093196158,
      "loss": 5.9655,
      "mean_token_accuracy": 0.1740314096212387,
      "num_tokens": 49223808.0,
      "step": 19420
    },
    {
      "entropy": 6.203909540176392,
      "epoch": 2.2417772648586265,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000450549101114233,
      "loss": 5.8892,
      "mean_token_accuracy": 0.1805233746767044,
      "num_tokens": 49236947.0,
      "step": 19425
    },
    {
      "entropy": 6.23789119720459,
      "epoch": 2.2423542989036354,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004505232870043434,
      "loss": 5.9582,
      "mean_token_accuracy": 0.1749293550848961,
      "num_tokens": 49249932.0,
      "step": 19430
    },
    {
      "entropy": 6.18991231918335,
      "epoch": 2.2429313329486438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004504974669908152,
      "loss": 5.9493,
      "mean_token_accuracy": 0.1756827101111412,
      "num_tokens": 49261783.0,
      "step": 19435
    },
    {
      "entropy": 6.327587842941284,
      "epoch": 2.2435083669936526,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045047164107451696,
      "loss": 5.9583,
      "mean_token_accuracy": 0.17252910435199736,
      "num_tokens": 49274495.0,
      "step": 19440
    },
    {
      "entropy": 6.2004626274108885,
      "epoch": 2.244085401038661,
      "grad_norm": 1.125,
      "learning_rate": 0.0004504458092563172,
      "loss": 5.9586,
      "mean_token_accuracy": 0.17035316228866576,
      "num_tokens": 49288450.0,
      "step": 19445
    },
    {
      "entropy": 6.239744567871094,
      "epoch": 2.24466243508367,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045041997153708475,
      "loss": 5.9496,
      "mean_token_accuracy": 0.17644737511873246,
      "num_tokens": 49300612.0,
      "step": 19450
    },
    {
      "entropy": 6.276393747329712,
      "epoch": 2.2452394691286788,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045039412791768877,
      "loss": 5.9832,
      "mean_token_accuracy": 0.17354014962911607,
      "num_tokens": 49313860.0,
      "step": 19455
    },
    {
      "entropy": 6.225166273117066,
      "epoch": 2.245816503173687,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045036827839899833,
      "loss": 5.9455,
      "mean_token_accuracy": 0.17199012637138367,
      "num_tokens": 49326864.0,
      "step": 19460
    },
    {
      "entropy": 6.2819251537323,
      "epoch": 2.246393537218696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045034242298188303,
      "loss": 5.9544,
      "mean_token_accuracy": 0.1678101897239685,
      "num_tokens": 49339444.0,
      "step": 19465
    },
    {
      "entropy": 6.12954158782959,
      "epoch": 2.2469705712637045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004503165616672124,
      "loss": 5.8615,
      "mean_token_accuracy": 0.18276833593845368,
      "num_tokens": 49352999.0,
      "step": 19470
    },
    {
      "entropy": 6.201620101928711,
      "epoch": 2.2475476053087133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004502906944558563,
      "loss": 5.8898,
      "mean_token_accuracy": 0.17764174044132233,
      "num_tokens": 49365600.0,
      "step": 19475
    },
    {
      "entropy": 6.308801078796387,
      "epoch": 2.2481246393537218,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004502648213486848,
      "loss": 5.9954,
      "mean_token_accuracy": 0.17161748558282852,
      "num_tokens": 49378121.0,
      "step": 19480
    },
    {
      "entropy": 6.1782279968261715,
      "epoch": 2.2487016733987306,
      "grad_norm": 1.125,
      "learning_rate": 0.00045023894234656807,
      "loss": 5.9314,
      "mean_token_accuracy": 0.17830360680818558,
      "num_tokens": 49390889.0,
      "step": 19485
    },
    {
      "entropy": 6.2839850902557375,
      "epoch": 2.249278707443739,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004502130574503766,
      "loss": 5.9815,
      "mean_token_accuracy": 0.17739966809749602,
      "num_tokens": 49404613.0,
      "step": 19490
    },
    {
      "entropy": 6.256258344650268,
      "epoch": 2.249855741488748,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004501871666609809,
      "loss": 5.9335,
      "mean_token_accuracy": 0.1765059784054756,
      "num_tokens": 49416142.0,
      "step": 19495
    },
    {
      "entropy": 6.2220173358917235,
      "epoch": 2.2504327755337563,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004501612699792519,
      "loss": 5.88,
      "mean_token_accuracy": 0.16992273181676865,
      "num_tokens": 49427848.0,
      "step": 19500
    },
    {
      "entropy": 6.331051683425903,
      "epoch": 2.251009809578765,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004501353674060606,
      "loss": 6.0521,
      "mean_token_accuracy": 0.16149662882089616,
      "num_tokens": 49442338.0,
      "step": 19505
    },
    {
      "entropy": 6.24295973777771,
      "epoch": 2.251586843623774,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004501094589422782,
      "loss": 5.935,
      "mean_token_accuracy": 0.17748762667179108,
      "num_tokens": 49455565.0,
      "step": 19510
    },
    {
      "entropy": 6.188134336471558,
      "epoch": 2.2521638776687825,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045008354458877614,
      "loss": 5.9141,
      "mean_token_accuracy": 0.17939674258232116,
      "num_tokens": 49468894.0,
      "step": 19515
    },
    {
      "entropy": 6.110340213775634,
      "epoch": 2.252740911713791,
      "grad_norm": 1.0,
      "learning_rate": 0.000450057624346426,
      "loss": 5.7611,
      "mean_token_accuracy": 0.1934901311993599,
      "num_tokens": 49480324.0,
      "step": 19520
    },
    {
      "entropy": 6.247253465652466,
      "epoch": 2.2533179457587997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045003169821609967,
      "loss": 5.9207,
      "mean_token_accuracy": 0.17454594671726226,
      "num_tokens": 49493823.0,
      "step": 19525
    },
    {
      "entropy": 6.233027696609497,
      "epoch": 2.2538949798038086,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045000576619866914,
      "loss": 5.9071,
      "mean_token_accuracy": 0.17930898070335388,
      "num_tokens": 49506512.0,
      "step": 19530
    },
    {
      "entropy": 6.1688103675842285,
      "epoch": 2.254472013848817,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044997982829500657,
      "loss": 5.9167,
      "mean_token_accuracy": 0.17058774083852768,
      "num_tokens": 49518424.0,
      "step": 19535
    },
    {
      "entropy": 6.193355369567871,
      "epoch": 2.255049047893826,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044995388450598436,
      "loss": 5.9605,
      "mean_token_accuracy": 0.17754435241222383,
      "num_tokens": 49531828.0,
      "step": 19540
    },
    {
      "entropy": 6.238674116134644,
      "epoch": 2.2556260819388343,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004499279348324751,
      "loss": 5.8964,
      "mean_token_accuracy": 0.16793029755353928,
      "num_tokens": 49545502.0,
      "step": 19545
    },
    {
      "entropy": 6.149578189849853,
      "epoch": 2.256203115983843,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044990197927535173,
      "loss": 5.8967,
      "mean_token_accuracy": 0.1788370653986931,
      "num_tokens": 49557323.0,
      "step": 19550
    },
    {
      "entropy": 6.139378929138184,
      "epoch": 2.2567801500288516,
      "grad_norm": 1.859375,
      "learning_rate": 0.00044987601783548703,
      "loss": 5.8089,
      "mean_token_accuracy": 0.18495144098997116,
      "num_tokens": 49570149.0,
      "step": 19555
    },
    {
      "entropy": 6.199171733856201,
      "epoch": 2.2573571840738604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004498500505137545,
      "loss": 5.8407,
      "mean_token_accuracy": 0.1889175683259964,
      "num_tokens": 49581866.0,
      "step": 19560
    },
    {
      "entropy": 6.092684745788574,
      "epoch": 2.257934218118869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004498240773110273,
      "loss": 5.8473,
      "mean_token_accuracy": 0.18385370969772338,
      "num_tokens": 49592785.0,
      "step": 19565
    },
    {
      "entropy": 6.244862270355225,
      "epoch": 2.2585112521638777,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004497980982281791,
      "loss": 6.0156,
      "mean_token_accuracy": 0.1703037440776825,
      "num_tokens": 49604925.0,
      "step": 19570
    },
    {
      "entropy": 6.2518209457397464,
      "epoch": 2.259088286208886,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004497721132660837,
      "loss": 5.9509,
      "mean_token_accuracy": 0.17581710517406463,
      "num_tokens": 49618548.0,
      "step": 19575
    },
    {
      "entropy": 6.187556648254395,
      "epoch": 2.259665320253895,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044974612242561516,
      "loss": 5.9415,
      "mean_token_accuracy": 0.17620837986469268,
      "num_tokens": 49631968.0,
      "step": 19580
    },
    {
      "entropy": 6.185222244262695,
      "epoch": 2.260242354298904,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004497201257076475,
      "loss": 5.9315,
      "mean_token_accuracy": 0.18308930695056916,
      "num_tokens": 49645269.0,
      "step": 19585
    },
    {
      "entropy": 6.237845849990845,
      "epoch": 2.2608193883439123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004496941231130552,
      "loss": 5.891,
      "mean_token_accuracy": 0.18141574263572693,
      "num_tokens": 49655903.0,
      "step": 19590
    },
    {
      "entropy": 6.288414096832275,
      "epoch": 2.261396422388921,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004496681146427129,
      "loss": 5.9861,
      "mean_token_accuracy": 0.16647860109806062,
      "num_tokens": 49667982.0,
      "step": 19595
    },
    {
      "entropy": 6.214272165298462,
      "epoch": 2.2619734564339296,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044964210029749523,
      "loss": 5.8577,
      "mean_token_accuracy": 0.1796637073159218,
      "num_tokens": 49681441.0,
      "step": 19600
    },
    {
      "entropy": 6.1274974822998045,
      "epoch": 2.2625504904789384,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044961608007827736,
      "loss": 5.8507,
      "mean_token_accuracy": 0.18508483171463014,
      "num_tokens": 49695805.0,
      "step": 19605
    },
    {
      "entropy": 6.187057113647461,
      "epoch": 2.263127524523947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004495900539859344,
      "loss": 5.973,
      "mean_token_accuracy": 0.17697968930006028,
      "num_tokens": 49708762.0,
      "step": 19610
    },
    {
      "entropy": 6.27952127456665,
      "epoch": 2.2637045585689557,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044956402202134157,
      "loss": 5.9378,
      "mean_token_accuracy": 0.18088844120502473,
      "num_tokens": 49724570.0,
      "step": 19615
    },
    {
      "entropy": 6.275792503356934,
      "epoch": 2.264281592613964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044953798418537465,
      "loss": 5.94,
      "mean_token_accuracy": 0.17699608206748962,
      "num_tokens": 49736060.0,
      "step": 19620
    },
    {
      "entropy": 6.193467664718628,
      "epoch": 2.264858626658973,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004495119404789093,
      "loss": 5.9197,
      "mean_token_accuracy": 0.17421784102916718,
      "num_tokens": 49748520.0,
      "step": 19625
    },
    {
      "entropy": 6.23006501197815,
      "epoch": 2.2654356607039814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004494858909028216,
      "loss": 5.9784,
      "mean_token_accuracy": 0.16674500256776809,
      "num_tokens": 49761097.0,
      "step": 19630
    },
    {
      "entropy": 6.214932680130005,
      "epoch": 2.2660126947489903,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004494598354579875,
      "loss": 5.8704,
      "mean_token_accuracy": 0.18461449444293976,
      "num_tokens": 49773430.0,
      "step": 19635
    },
    {
      "entropy": 6.187216472625733,
      "epoch": 2.2665897287939987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004494337741452836,
      "loss": 5.9325,
      "mean_token_accuracy": 0.17151115089654922,
      "num_tokens": 49785525.0,
      "step": 19640
    },
    {
      "entropy": 6.178369617462158,
      "epoch": 2.2671667628390075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004494077069655863,
      "loss": 5.9369,
      "mean_token_accuracy": 0.1749401494860649,
      "num_tokens": 49796650.0,
      "step": 19645
    },
    {
      "entropy": 6.2306800365448,
      "epoch": 2.267743796884016,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004493816339197724,
      "loss": 5.9266,
      "mean_token_accuracy": 0.1802807092666626,
      "num_tokens": 49808191.0,
      "step": 19650
    },
    {
      "entropy": 6.1881694316864015,
      "epoch": 2.268320830929025,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044935555500871897,
      "loss": 5.9515,
      "mean_token_accuracy": 0.17349853515625,
      "num_tokens": 49820051.0,
      "step": 19655
    },
    {
      "entropy": 6.264400243759155,
      "epoch": 2.2688978649740337,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000449329470233303,
      "loss": 5.9709,
      "mean_token_accuracy": 0.17374078631401063,
      "num_tokens": 49835392.0,
      "step": 19660
    },
    {
      "entropy": 6.1371400356292725,
      "epoch": 2.269474899019042,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044930337959440183,
      "loss": 5.9025,
      "mean_token_accuracy": 0.17800106406211852,
      "num_tokens": 49847344.0,
      "step": 19665
    },
    {
      "entropy": 6.221584177017212,
      "epoch": 2.270051933064051,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004492772830928931,
      "loss": 5.95,
      "mean_token_accuracy": 0.16992736905813216,
      "num_tokens": 49859741.0,
      "step": 19670
    },
    {
      "entropy": 6.215978479385376,
      "epoch": 2.2706289671090594,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044925118072965456,
      "loss": 5.9152,
      "mean_token_accuracy": 0.1744893655180931,
      "num_tokens": 49871565.0,
      "step": 19675
    },
    {
      "entropy": 6.168155670166016,
      "epoch": 2.2712060011540682,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000449225072505564,
      "loss": 5.882,
      "mean_token_accuracy": 0.1758724957704544,
      "num_tokens": 49883405.0,
      "step": 19680
    },
    {
      "entropy": 6.248301076889038,
      "epoch": 2.2717830351990767,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044919895842149976,
      "loss": 5.9506,
      "mean_token_accuracy": 0.17669990807771682,
      "num_tokens": 49895924.0,
      "step": 19685
    },
    {
      "entropy": 6.1905598640441895,
      "epoch": 2.2723600692440855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044917283847834005,
      "loss": 5.8828,
      "mean_token_accuracy": 0.17163273692131042,
      "num_tokens": 49908506.0,
      "step": 19690
    },
    {
      "entropy": 6.2119269371032715,
      "epoch": 2.272937103289094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004491467126769635,
      "loss": 5.9188,
      "mean_token_accuracy": 0.17607008367776872,
      "num_tokens": 49919917.0,
      "step": 19695
    },
    {
      "entropy": 6.207222509384155,
      "epoch": 2.273514137334103,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044912058101824866,
      "loss": 5.904,
      "mean_token_accuracy": 0.1741974800825119,
      "num_tokens": 49932508.0,
      "step": 19700
    },
    {
      "entropy": 6.191767454147339,
      "epoch": 2.274091171379111,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044909444350307463,
      "loss": 5.9344,
      "mean_token_accuracy": 0.1761745110154152,
      "num_tokens": 49945065.0,
      "step": 19705
    },
    {
      "entropy": 6.172743940353394,
      "epoch": 2.27466820542412,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004490683001323205,
      "loss": 5.8509,
      "mean_token_accuracy": 0.17929312735795974,
      "num_tokens": 49957530.0,
      "step": 19710
    },
    {
      "entropy": 6.146152830123901,
      "epoch": 2.2752452394691285,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044904215090686554,
      "loss": 5.8801,
      "mean_token_accuracy": 0.17303403168916703,
      "num_tokens": 49969760.0,
      "step": 19715
    },
    {
      "entropy": 6.258595895767212,
      "epoch": 2.2758222735141374,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044901599582758926,
      "loss": 5.9564,
      "mean_token_accuracy": 0.1772393137216568,
      "num_tokens": 49983151.0,
      "step": 19720
    },
    {
      "entropy": 6.181928110122681,
      "epoch": 2.2763993075591458,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044898983489537143,
      "loss": 5.8337,
      "mean_token_accuracy": 0.1764400839805603,
      "num_tokens": 49995262.0,
      "step": 19725
    },
    {
      "entropy": 6.208969163894653,
      "epoch": 2.2769763416041546,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000448963668111092,
      "loss": 5.8993,
      "mean_token_accuracy": 0.17962197512388228,
      "num_tokens": 50007153.0,
      "step": 19730
    },
    {
      "entropy": 6.162887144088745,
      "epoch": 2.2775533756491635,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00044893749547563085,
      "loss": 5.9366,
      "mean_token_accuracy": 0.17497817277908326,
      "num_tokens": 50020145.0,
      "step": 19735
    },
    {
      "entropy": 6.155488443374634,
      "epoch": 2.278130409694172,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044891131698986846,
      "loss": 5.9078,
      "mean_token_accuracy": 0.17562464624643326,
      "num_tokens": 50032741.0,
      "step": 19740
    },
    {
      "entropy": 6.162566900253296,
      "epoch": 2.2787074437391808,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004488851326546854,
      "loss": 5.9427,
      "mean_token_accuracy": 0.17191180139780043,
      "num_tokens": 50046112.0,
      "step": 19745
    },
    {
      "entropy": 6.242087411880493,
      "epoch": 2.279284477784189,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004488589424709622,
      "loss": 5.8989,
      "mean_token_accuracy": 0.18044222593307496,
      "num_tokens": 50058718.0,
      "step": 19750
    },
    {
      "entropy": 6.147966098785401,
      "epoch": 2.279861511829198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004488327464395799,
      "loss": 5.8655,
      "mean_token_accuracy": 0.17074532955884933,
      "num_tokens": 50071759.0,
      "step": 19755
    },
    {
      "entropy": 6.244132232666016,
      "epoch": 2.2804385458742065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004488065445614195,
      "loss": 5.9034,
      "mean_token_accuracy": 0.17695511877536774,
      "num_tokens": 50083948.0,
      "step": 19760
    },
    {
      "entropy": 6.20736174583435,
      "epoch": 2.2810155799192153,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004487803368373623,
      "loss": 5.9119,
      "mean_token_accuracy": 0.17644120901823043,
      "num_tokens": 50097310.0,
      "step": 19765
    },
    {
      "entropy": 6.1846113204956055,
      "epoch": 2.2815926139642237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004487541232682898,
      "loss": 5.9435,
      "mean_token_accuracy": 0.17838864773511887,
      "num_tokens": 50109939.0,
      "step": 19770
    },
    {
      "entropy": 6.266738367080689,
      "epoch": 2.2821696480092326,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004487279038550836,
      "loss": 5.9815,
      "mean_token_accuracy": 0.1711701899766922,
      "num_tokens": 50120651.0,
      "step": 19775
    },
    {
      "entropy": 6.2213959217071535,
      "epoch": 2.282746682054241,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004487016785986258,
      "loss": 5.9244,
      "mean_token_accuracy": 0.17605694830417634,
      "num_tokens": 50133180.0,
      "step": 19780
    },
    {
      "entropy": 6.146054744720459,
      "epoch": 2.28332371609925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044867544749979817,
      "loss": 5.902,
      "mean_token_accuracy": 0.1776263177394867,
      "num_tokens": 50144808.0,
      "step": 19785
    },
    {
      "entropy": 6.176597976684571,
      "epoch": 2.2839007501442588,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044864921055948317,
      "loss": 5.8998,
      "mean_token_accuracy": 0.17587872147560119,
      "num_tokens": 50157487.0,
      "step": 19790
    },
    {
      "entropy": 6.259101152420044,
      "epoch": 2.284477784189267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044862296777856326,
      "loss": 5.9708,
      "mean_token_accuracy": 0.16480046957731248,
      "num_tokens": 50171837.0,
      "step": 19795
    },
    {
      "entropy": 6.228283309936524,
      "epoch": 2.2850548182342756,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485967191579211,
      "loss": 5.8797,
      "mean_token_accuracy": 0.18508685678243636,
      "num_tokens": 50184802.0,
      "step": 19800
    },
    {
      "entropy": 6.209461212158203,
      "epoch": 2.2856318522792844,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004485704646984394,
      "loss": 5.8914,
      "mean_token_accuracy": 0.17526766955852507,
      "num_tokens": 50196433.0,
      "step": 19805
    },
    {
      "entropy": 6.1189063549041744,
      "epoch": 2.2862088863242933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004485442044010015,
      "loss": 5.835,
      "mean_token_accuracy": 0.19224470853805542,
      "num_tokens": 50209549.0,
      "step": 19810
    },
    {
      "entropy": 6.224236059188843,
      "epoch": 2.2867859203693017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004485179382664904,
      "loss": 5.9863,
      "mean_token_accuracy": 0.17298453003168107,
      "num_tokens": 50220691.0,
      "step": 19815
    },
    {
      "entropy": 6.336439943313598,
      "epoch": 2.2873629544143106,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004484916662957896,
      "loss": 6.0277,
      "mean_token_accuracy": 0.1718783736228943,
      "num_tokens": 50233003.0,
      "step": 19820
    },
    {
      "entropy": 6.160794115066528,
      "epoch": 2.287939988459319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004484653884897829,
      "loss": 5.8572,
      "mean_token_accuracy": 0.18045479953289031,
      "num_tokens": 50244742.0,
      "step": 19825
    },
    {
      "entropy": 6.094951963424682,
      "epoch": 2.288517022504328,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00044843910484935394,
      "loss": 5.8215,
      "mean_token_accuracy": 0.18447613716125488,
      "num_tokens": 50258188.0,
      "step": 19830
    },
    {
      "entropy": 6.253140926361084,
      "epoch": 2.2890940565493363,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004484128153753868,
      "loss": 5.9649,
      "mean_token_accuracy": 0.17857753336429597,
      "num_tokens": 50271155.0,
      "step": 19835
    },
    {
      "entropy": 6.259934520721435,
      "epoch": 2.289671090594345,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044838652006876583,
      "loss": 5.9484,
      "mean_token_accuracy": 0.1786944955587387,
      "num_tokens": 50283518.0,
      "step": 19840
    },
    {
      "entropy": 6.135973882675171,
      "epoch": 2.2902481246393536,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044836021893037537,
      "loss": 5.8368,
      "mean_token_accuracy": 0.18843693882226945,
      "num_tokens": 50296208.0,
      "step": 19845
    },
    {
      "entropy": 6.154265737533569,
      "epoch": 2.2908251586843624,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004483339119611001,
      "loss": 5.8718,
      "mean_token_accuracy": 0.1847282886505127,
      "num_tokens": 50309111.0,
      "step": 19850
    },
    {
      "entropy": 6.208286762237549,
      "epoch": 2.291402192729371,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044830759916182476,
      "loss": 5.9524,
      "mean_token_accuracy": 0.1728263944387436,
      "num_tokens": 50321895.0,
      "step": 19855
    },
    {
      "entropy": 6.20464072227478,
      "epoch": 2.2919792267743797,
      "grad_norm": 1.9921875,
      "learning_rate": 0.0004482812805334344,
      "loss": 5.9917,
      "mean_token_accuracy": 0.16743112802505494,
      "num_tokens": 50335769.0,
      "step": 19860
    },
    {
      "entropy": 6.26806149482727,
      "epoch": 2.2925562608193886,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004482549560768142,
      "loss": 5.9187,
      "mean_token_accuracy": 0.17352110594511033,
      "num_tokens": 50348576.0,
      "step": 19865
    },
    {
      "entropy": 6.221711444854736,
      "epoch": 2.293133294864397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004482286257928497,
      "loss": 5.926,
      "mean_token_accuracy": 0.17436802685260772,
      "num_tokens": 50360652.0,
      "step": 19870
    },
    {
      "entropy": 6.177728652954102,
      "epoch": 2.293710328909406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004482022896824263,
      "loss": 5.9429,
      "mean_token_accuracy": 0.17300977408885956,
      "num_tokens": 50373256.0,
      "step": 19875
    },
    {
      "entropy": 6.2224808692932125,
      "epoch": 2.2942873629544143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044817594774643004,
      "loss": 5.8255,
      "mean_token_accuracy": 0.18255933970212937,
      "num_tokens": 50385245.0,
      "step": 19880
    },
    {
      "entropy": 6.191463851928711,
      "epoch": 2.294864396999423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044814959998574675,
      "loss": 5.9129,
      "mean_token_accuracy": 0.17643292397260665,
      "num_tokens": 50397276.0,
      "step": 19885
    },
    {
      "entropy": 6.273288202285767,
      "epoch": 2.2954414310444315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044812324640126265,
      "loss": 6.0222,
      "mean_token_accuracy": 0.17482039630413054,
      "num_tokens": 50409725.0,
      "step": 19890
    },
    {
      "entropy": 6.262744665145874,
      "epoch": 2.2960184650894404,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004480968869938642,
      "loss": 5.9305,
      "mean_token_accuracy": 0.17356754541397096,
      "num_tokens": 50421987.0,
      "step": 19895
    },
    {
      "entropy": 6.1676928997039795,
      "epoch": 2.296595499134449,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044807052176443793,
      "loss": 5.9042,
      "mean_token_accuracy": 0.1803156927227974,
      "num_tokens": 50436153.0,
      "step": 19900
    },
    {
      "entropy": 6.124884366989136,
      "epoch": 2.2971725331794577,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044804415071387067,
      "loss": 5.9951,
      "mean_token_accuracy": 0.17772471457719802,
      "num_tokens": 50449683.0,
      "step": 19905
    },
    {
      "entropy": 6.200593662261963,
      "epoch": 2.297749567224466,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004480177738430492,
      "loss": 5.7895,
      "mean_token_accuracy": 0.18063700795173646,
      "num_tokens": 50461535.0,
      "step": 19910
    },
    {
      "entropy": 6.235341739654541,
      "epoch": 2.298326601269475,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044799139115286104,
      "loss": 5.9209,
      "mean_token_accuracy": 0.17471311539411544,
      "num_tokens": 50476094.0,
      "step": 19915
    },
    {
      "entropy": 6.181540203094483,
      "epoch": 2.2989036353144834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004479650026441933,
      "loss": 5.8874,
      "mean_token_accuracy": 0.1754192292690277,
      "num_tokens": 50488751.0,
      "step": 19920
    },
    {
      "entropy": 6.220038938522339,
      "epoch": 2.2994806693594922,
      "grad_norm": 1.53125,
      "learning_rate": 0.00044793860831793356,
      "loss": 5.9475,
      "mean_token_accuracy": 0.17229341864585876,
      "num_tokens": 50502272.0,
      "step": 19925
    },
    {
      "entropy": 6.246703433990478,
      "epoch": 2.3000577034045007,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044791220817496963,
      "loss": 5.9187,
      "mean_token_accuracy": 0.17167287766933442,
      "num_tokens": 50514792.0,
      "step": 19930
    },
    {
      "entropy": 6.15828309059143,
      "epoch": 2.3006347374495095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478858022161895,
      "loss": 5.8417,
      "mean_token_accuracy": 0.18575220704078674,
      "num_tokens": 50528484.0,
      "step": 19935
    },
    {
      "entropy": 6.285755968093872,
      "epoch": 2.3012117714945184,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004478593904424813,
      "loss": 5.9919,
      "mean_token_accuracy": 0.17144529074430465,
      "num_tokens": 50540737.0,
      "step": 19940
    },
    {
      "entropy": 6.213376903533936,
      "epoch": 2.301788805539527,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004478329728547334,
      "loss": 5.9239,
      "mean_token_accuracy": 0.17930820733308792,
      "num_tokens": 50553819.0,
      "step": 19945
    },
    {
      "entropy": 6.238288450241089,
      "epoch": 2.3023658395845357,
      "grad_norm": 1.0,
      "learning_rate": 0.00044780654945383424,
      "loss": 5.953,
      "mean_token_accuracy": 0.1771923080086708,
      "num_tokens": 50566275.0,
      "step": 19950
    },
    {
      "entropy": 6.238650465011597,
      "epoch": 2.302942873629544,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044778012024067267,
      "loss": 5.8737,
      "mean_token_accuracy": 0.18007982820272445,
      "num_tokens": 50577653.0,
      "step": 19955
    },
    {
      "entropy": 6.233930253982544,
      "epoch": 2.303519907674553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004477536852161376,
      "loss": 5.9626,
      "mean_token_accuracy": 0.1656869351863861,
      "num_tokens": 50590237.0,
      "step": 19960
    },
    {
      "entropy": 6.168978595733643,
      "epoch": 2.3040969417195614,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004477272443811181,
      "loss": 5.8562,
      "mean_token_accuracy": 0.1712908536195755,
      "num_tokens": 50602333.0,
      "step": 19965
    },
    {
      "entropy": 6.262486791610717,
      "epoch": 2.3046739757645702,
      "grad_norm": 1.0,
      "learning_rate": 0.0004477007977365035,
      "loss": 5.9618,
      "mean_token_accuracy": 0.1690537542104721,
      "num_tokens": 50614232.0,
      "step": 19970
    },
    {
      "entropy": 6.247935342788696,
      "epoch": 2.3052510098095786,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004476743452831834,
      "loss": 5.9541,
      "mean_token_accuracy": 0.17183466255664825,
      "num_tokens": 50627959.0,
      "step": 19975
    },
    {
      "entropy": 6.169540357589722,
      "epoch": 2.3058280438545875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044764788702204747,
      "loss": 5.8152,
      "mean_token_accuracy": 0.1843651071190834,
      "num_tokens": 50639962.0,
      "step": 19980
    },
    {
      "entropy": 6.214004039764404,
      "epoch": 2.306405077899596,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004476214229539856,
      "loss": 5.9304,
      "mean_token_accuracy": 0.17756774574518203,
      "num_tokens": 50652291.0,
      "step": 19985
    },
    {
      "entropy": 6.233112430572509,
      "epoch": 2.306982111944605,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004475949530798879,
      "loss": 5.9011,
      "mean_token_accuracy": 0.17774045169353486,
      "num_tokens": 50664993.0,
      "step": 19990
    },
    {
      "entropy": 6.14248218536377,
      "epoch": 2.307559145989613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044756847740064475,
      "loss": 5.7793,
      "mean_token_accuracy": 0.17513145357370377,
      "num_tokens": 50677846.0,
      "step": 19995
    },
    {
      "entropy": 6.143243598937988,
      "epoch": 2.308136180034622,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004475419959171465,
      "loss": 5.8587,
      "mean_token_accuracy": 0.18726265281438828,
      "num_tokens": 50691068.0,
      "step": 20000
    },
    {
      "entropy": 6.171260356903076,
      "epoch": 2.3087132140796305,
      "grad_norm": 1.0,
      "learning_rate": 0.000447515508630284,
      "loss": 5.931,
      "mean_token_accuracy": 0.177291938662529,
      "num_tokens": 50703663.0,
      "step": 20005
    },
    {
      "entropy": 6.2965131282806395,
      "epoch": 2.3092902481246393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044748901554094806,
      "loss": 5.9033,
      "mean_token_accuracy": 0.1819427266716957,
      "num_tokens": 50715709.0,
      "step": 20010
    },
    {
      "entropy": 6.198210048675537,
      "epoch": 2.309867282169648,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004474625166500297,
      "loss": 5.8687,
      "mean_token_accuracy": 0.18542738258838654,
      "num_tokens": 50727700.0,
      "step": 20015
    },
    {
      "entropy": 6.208634757995606,
      "epoch": 2.3104443162146566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044743601195842026,
      "loss": 5.9009,
      "mean_token_accuracy": 0.17990079522132874,
      "num_tokens": 50740604.0,
      "step": 20020
    },
    {
      "entropy": 6.277386045455932,
      "epoch": 2.3110213502596655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044740950146701127,
      "loss": 5.9924,
      "mean_token_accuracy": 0.17152770310640336,
      "num_tokens": 50753666.0,
      "step": 20025
    },
    {
      "entropy": 6.195950126647949,
      "epoch": 2.311598384304674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473829851766943,
      "loss": 5.9452,
      "mean_token_accuracy": 0.17860945016145707,
      "num_tokens": 50766324.0,
      "step": 20030
    },
    {
      "entropy": 6.265624618530273,
      "epoch": 2.3121754183496828,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004473564630883612,
      "loss": 5.9371,
      "mean_token_accuracy": 0.17894651144742965,
      "num_tokens": 50779037.0,
      "step": 20035
    },
    {
      "entropy": 6.181953954696655,
      "epoch": 2.312752452394691,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004473299352029042,
      "loss": 5.8668,
      "mean_token_accuracy": 0.18175242394208907,
      "num_tokens": 50791089.0,
      "step": 20040
    },
    {
      "entropy": 6.1564857959747314,
      "epoch": 2.3133294864397,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004473034015212154,
      "loss": 5.8547,
      "mean_token_accuracy": 0.1750751867890358,
      "num_tokens": 50803608.0,
      "step": 20045
    },
    {
      "entropy": 6.173952054977417,
      "epoch": 2.3139065204847085,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004472768620441872,
      "loss": 5.909,
      "mean_token_accuracy": 0.17483999580144882,
      "num_tokens": 50815712.0,
      "step": 20050
    },
    {
      "entropy": 6.243170690536499,
      "epoch": 2.3144835545297173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044725031677271234,
      "loss": 5.8446,
      "mean_token_accuracy": 0.18447517454624177,
      "num_tokens": 50827409.0,
      "step": 20055
    },
    {
      "entropy": 6.165426301956177,
      "epoch": 2.3150605885747257,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004472237657076837,
      "loss": 5.8925,
      "mean_token_accuracy": 0.17439614236354828,
      "num_tokens": 50840901.0,
      "step": 20060
    },
    {
      "entropy": 6.258820390701294,
      "epoch": 2.3156376226197346,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004471972088499942,
      "loss": 5.9661,
      "mean_token_accuracy": 0.17297065407037734,
      "num_tokens": 50854453.0,
      "step": 20065
    },
    {
      "entropy": 6.269705677032471,
      "epoch": 2.3162146566647435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004471706462005371,
      "loss": 5.8945,
      "mean_token_accuracy": 0.1765055999159813,
      "num_tokens": 50866566.0,
      "step": 20070
    },
    {
      "entropy": 6.137576532363892,
      "epoch": 2.316791690709752,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004471440777602059,
      "loss": 5.9051,
      "mean_token_accuracy": 0.17659924328327178,
      "num_tokens": 50879695.0,
      "step": 20075
    },
    {
      "entropy": 6.127049732208252,
      "epoch": 2.3173687247547603,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044711750352989407,
      "loss": 5.843,
      "mean_token_accuracy": 0.17433359920978547,
      "num_tokens": 50892350.0,
      "step": 20080
    },
    {
      "entropy": 6.255441665649414,
      "epoch": 2.317945758799769,
      "grad_norm": 1.0,
      "learning_rate": 0.0004470909235104956,
      "loss": 5.9548,
      "mean_token_accuracy": 0.17594036012887954,
      "num_tokens": 50905899.0,
      "step": 20085
    },
    {
      "entropy": 6.242904472351074,
      "epoch": 2.318522792844778,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470643377029043,
      "loss": 5.8895,
      "mean_token_accuracy": 0.18192937225103378,
      "num_tokens": 50919698.0,
      "step": 20090
    },
    {
      "entropy": 6.236587238311768,
      "epoch": 2.3190998268897864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004470377461080145,
      "loss": 5.9695,
      "mean_token_accuracy": 0.17987661361694335,
      "num_tokens": 50932236.0,
      "step": 20095
    },
    {
      "entropy": 6.238249254226685,
      "epoch": 2.3196768609347953,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004470111487267206,
      "loss": 5.9123,
      "mean_token_accuracy": 0.1801683187484741,
      "num_tokens": 50945403.0,
      "step": 20100
    },
    {
      "entropy": 6.208332920074463,
      "epoch": 2.3202538949798037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004469845455599171,
      "loss": 5.9185,
      "mean_token_accuracy": 0.1714116394519806,
      "num_tokens": 50958854.0,
      "step": 20105
    },
    {
      "entropy": 6.259473562240601,
      "epoch": 2.3208309290248126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004469579366084989,
      "loss": 5.9694,
      "mean_token_accuracy": 0.17025109976530076,
      "num_tokens": 50971545.0,
      "step": 20110
    },
    {
      "entropy": 6.2504490375518795,
      "epoch": 2.321407963069821,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044693132187336094,
      "loss": 6.0001,
      "mean_token_accuracy": 0.17513484954833985,
      "num_tokens": 50985282.0,
      "step": 20115
    },
    {
      "entropy": 6.165658330917358,
      "epoch": 2.32198499711483,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004469047013553983,
      "loss": 5.8862,
      "mean_token_accuracy": 0.17963058054447173,
      "num_tokens": 50998063.0,
      "step": 20120
    },
    {
      "entropy": 6.177840852737427,
      "epoch": 2.3225620311598383,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00044687807505550646,
      "loss": 5.9096,
      "mean_token_accuracy": 0.18144043385982514,
      "num_tokens": 51011413.0,
      "step": 20125
    },
    {
      "entropy": 6.221637010574341,
      "epoch": 2.323139065204847,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044685144297458096,
      "loss": 5.8834,
      "mean_token_accuracy": 0.18326039761304855,
      "num_tokens": 51023861.0,
      "step": 20130
    },
    {
      "entropy": 6.232079792022705,
      "epoch": 2.3237160992498556,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044682480511351754,
      "loss": 5.875,
      "mean_token_accuracy": 0.18111318051815034,
      "num_tokens": 51035891.0,
      "step": 20135
    },
    {
      "entropy": 6.149922037124634,
      "epoch": 2.3242931332948644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004467981614732121,
      "loss": 5.8481,
      "mean_token_accuracy": 0.18590421825647355,
      "num_tokens": 51048115.0,
      "step": 20140
    },
    {
      "entropy": 6.196219253540039,
      "epoch": 2.3248701673398733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044677151205456086,
      "loss": 5.9211,
      "mean_token_accuracy": 0.17520425617694854,
      "num_tokens": 51061503.0,
      "step": 20145
    },
    {
      "entropy": 6.238576412200928,
      "epoch": 2.3254472013848817,
      "grad_norm": 2.296875,
      "learning_rate": 0.0004467448568584601,
      "loss": 5.9289,
      "mean_token_accuracy": 0.17929600328207015,
      "num_tokens": 51074052.0,
      "step": 20150
    },
    {
      "entropy": 6.236129093170166,
      "epoch": 2.32602423542989,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004467181958858064,
      "loss": 5.9597,
      "mean_token_accuracy": 0.18050430119037628,
      "num_tokens": 51089456.0,
      "step": 20155
    },
    {
      "entropy": 6.247223806381226,
      "epoch": 2.326601269474899,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004466915291374965,
      "loss": 5.9676,
      "mean_token_accuracy": 0.16788686513900758,
      "num_tokens": 51102835.0,
      "step": 20160
    },
    {
      "entropy": 6.2244714260101315,
      "epoch": 2.327178303519908,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004466648566144273,
      "loss": 5.9395,
      "mean_token_accuracy": 0.1768688216805458,
      "num_tokens": 51115472.0,
      "step": 20165
    },
    {
      "entropy": 6.2383575439453125,
      "epoch": 2.3277553375649163,
      "grad_norm": 1.28125,
      "learning_rate": 0.000446638178317496,
      "loss": 5.9246,
      "mean_token_accuracy": 0.1741773083806038,
      "num_tokens": 51129768.0,
      "step": 20170
    },
    {
      "entropy": 6.202746438980102,
      "epoch": 2.328332371609925,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044661149424759974,
      "loss": 5.8963,
      "mean_token_accuracy": 0.17682316452264785,
      "num_tokens": 51143517.0,
      "step": 20175
    },
    {
      "entropy": 6.201585912704468,
      "epoch": 2.3289094056549335,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004465848044056361,
      "loss": 5.784,
      "mean_token_accuracy": 0.18313243687152864,
      "num_tokens": 51157235.0,
      "step": 20180
    },
    {
      "entropy": 6.241024971008301,
      "epoch": 2.3294864396999424,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004465581087925028,
      "loss": 5.9012,
      "mean_token_accuracy": 0.17613800168037413,
      "num_tokens": 51169061.0,
      "step": 20185
    },
    {
      "entropy": 6.1859945297241214,
      "epoch": 2.330063473744951,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004465314074090978,
      "loss": 5.9147,
      "mean_token_accuracy": 0.17369749397039413,
      "num_tokens": 51181063.0,
      "step": 20190
    },
    {
      "entropy": 6.255963897705078,
      "epoch": 2.3306405077899597,
      "grad_norm": 1.0,
      "learning_rate": 0.00044650470025631904,
      "loss": 5.8258,
      "mean_token_accuracy": 0.18363078087568283,
      "num_tokens": 51193536.0,
      "step": 20195
    },
    {
      "entropy": 6.172869110107422,
      "epoch": 2.331217541834968,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004464779873350649,
      "loss": 5.8457,
      "mean_token_accuracy": 0.18686456829309464,
      "num_tokens": 51205960.0,
      "step": 20200
    },
    {
      "entropy": 6.235818147659302,
      "epoch": 2.331794575879977,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004464512686462339,
      "loss": 6.0352,
      "mean_token_accuracy": 0.17198680937290192,
      "num_tokens": 51219227.0,
      "step": 20205
    },
    {
      "entropy": 6.26986722946167,
      "epoch": 2.3323716099249854,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044642454419072455,
      "loss": 5.8917,
      "mean_token_accuracy": 0.17737701833248137,
      "num_tokens": 51232781.0,
      "step": 20210
    },
    {
      "entropy": 6.1620416164398195,
      "epoch": 2.3329486439699942,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004463978139694358,
      "loss": 5.8911,
      "mean_token_accuracy": 0.18242465257644652,
      "num_tokens": 51245574.0,
      "step": 20215
    },
    {
      "entropy": 6.181765127182007,
      "epoch": 2.333525678015003,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044637107798326675,
      "loss": 5.9094,
      "mean_token_accuracy": 0.17442320585250853,
      "num_tokens": 51258710.0,
      "step": 20220
    },
    {
      "entropy": 6.303627157211304,
      "epoch": 2.3341027120600115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004463443362331166,
      "loss": 5.9982,
      "mean_token_accuracy": 0.17676437944173812,
      "num_tokens": 51270382.0,
      "step": 20225
    },
    {
      "entropy": 6.296655225753784,
      "epoch": 2.3346797461050204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044631758871988486,
      "loss": 6.0015,
      "mean_token_accuracy": 0.17527170479297638,
      "num_tokens": 51283923.0,
      "step": 20230
    },
    {
      "entropy": 6.296395015716553,
      "epoch": 2.335256780150029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004462908354444711,
      "loss": 5.949,
      "mean_token_accuracy": 0.1749487653374672,
      "num_tokens": 51297671.0,
      "step": 20235
    },
    {
      "entropy": 6.188365316390991,
      "epoch": 2.3358338141950377,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004462640764077751,
      "loss": 5.893,
      "mean_token_accuracy": 0.17709133625030518,
      "num_tokens": 51309339.0,
      "step": 20240
    },
    {
      "entropy": 6.241654634475708,
      "epoch": 2.336410848240046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004462373116106971,
      "loss": 5.9949,
      "mean_token_accuracy": 0.17668516039848328,
      "num_tokens": 51321479.0,
      "step": 20245
    },
    {
      "entropy": 6.22584114074707,
      "epoch": 2.336987882285055,
      "grad_norm": 1.25,
      "learning_rate": 0.00044621054105413704,
      "loss": 5.9345,
      "mean_token_accuracy": 0.17172120362520218,
      "num_tokens": 51335135.0,
      "step": 20250
    },
    {
      "entropy": 6.133741807937622,
      "epoch": 2.3375649163300634,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044618376473899555,
      "loss": 5.8699,
      "mean_token_accuracy": 0.1795585796236992,
      "num_tokens": 51346556.0,
      "step": 20255
    },
    {
      "entropy": 6.123044729232788,
      "epoch": 2.338141950375072,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004461569826661732,
      "loss": 5.8347,
      "mean_token_accuracy": 0.18230682015419006,
      "num_tokens": 51359692.0,
      "step": 20260
    },
    {
      "entropy": 6.258272743225097,
      "epoch": 2.3387189844200806,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004461301948365707,
      "loss": 5.944,
      "mean_token_accuracy": 0.16859951466321946,
      "num_tokens": 51371682.0,
      "step": 20265
    },
    {
      "entropy": 6.238030576705933,
      "epoch": 2.3392960184650895,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004461034012510891,
      "loss": 5.9245,
      "mean_token_accuracy": 0.1813396081328392,
      "num_tokens": 51383533.0,
      "step": 20270
    },
    {
      "entropy": 6.156278944015503,
      "epoch": 2.339873052510098,
      "grad_norm": 1.125,
      "learning_rate": 0.00044607660191062963,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17602861672639847,
      "num_tokens": 51395696.0,
      "step": 20275
    },
    {
      "entropy": 6.193694400787353,
      "epoch": 2.340450086555107,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044604979681609364,
      "loss": 5.8663,
      "mean_token_accuracy": 0.18097577840089799,
      "num_tokens": 51408460.0,
      "step": 20280
    },
    {
      "entropy": 6.147213506698608,
      "epoch": 2.341027120600115,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044602298596838264,
      "loss": 5.8235,
      "mean_token_accuracy": 0.18585553616285325,
      "num_tokens": 51420768.0,
      "step": 20285
    },
    {
      "entropy": 6.229567384719848,
      "epoch": 2.341604154645124,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044599616936839853,
      "loss": 5.9126,
      "mean_token_accuracy": 0.17630589306354522,
      "num_tokens": 51434429.0,
      "step": 20290
    },
    {
      "entropy": 6.2479852676391605,
      "epoch": 2.342181188690133,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004459693470170432,
      "loss": 5.964,
      "mean_token_accuracy": 0.17569663673639296,
      "num_tokens": 51447146.0,
      "step": 20295
    },
    {
      "entropy": 6.163450908660889,
      "epoch": 2.3427582227351413,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004459425189152187,
      "loss": 5.8526,
      "mean_token_accuracy": 0.18089221566915512,
      "num_tokens": 51458656.0,
      "step": 20300
    },
    {
      "entropy": 6.1185956478118895,
      "epoch": 2.34333525678015,
      "grad_norm": 1.875,
      "learning_rate": 0.00044591568506382757,
      "loss": 5.7917,
      "mean_token_accuracy": 0.18526783734560012,
      "num_tokens": 51473350.0,
      "step": 20305
    },
    {
      "entropy": 6.210563611984253,
      "epoch": 2.3439122908251586,
      "grad_norm": 1.125,
      "learning_rate": 0.00044588884546377226,
      "loss": 5.8819,
      "mean_token_accuracy": 0.18186958581209184,
      "num_tokens": 51486801.0,
      "step": 20310
    },
    {
      "entropy": 6.221357297897339,
      "epoch": 2.3444893248701675,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004458620001159555,
      "loss": 5.9124,
      "mean_token_accuracy": 0.18402253985404968,
      "num_tokens": 51499217.0,
      "step": 20315
    },
    {
      "entropy": 6.258128213882446,
      "epoch": 2.345066358915176,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004458351490212803,
      "loss": 5.9787,
      "mean_token_accuracy": 0.17318946570158006,
      "num_tokens": 51511714.0,
      "step": 20320
    },
    {
      "entropy": 6.1194176197052,
      "epoch": 2.3456433929601848,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044580829218064964,
      "loss": 5.7943,
      "mean_token_accuracy": 0.18501977026462554,
      "num_tokens": 51522725.0,
      "step": 20325
    },
    {
      "entropy": 6.2714615821838375,
      "epoch": 2.346220427005193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000445781429594967,
      "loss": 5.9029,
      "mean_token_accuracy": 0.18235889971256256,
      "num_tokens": 51535178.0,
      "step": 20330
    },
    {
      "entropy": 6.195360422134399,
      "epoch": 2.346797461050202,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004457545612651357,
      "loss": 6.0008,
      "mean_token_accuracy": 0.17095666080713273,
      "num_tokens": 51546859.0,
      "step": 20335
    },
    {
      "entropy": 6.213983869552612,
      "epoch": 2.3473744950952105,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044572768719205964,
      "loss": 5.9098,
      "mean_token_accuracy": 0.17400958389043808,
      "num_tokens": 51558719.0,
      "step": 20340
    },
    {
      "entropy": 6.2138770580291744,
      "epoch": 2.3479515291402193,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044570080737664264,
      "loss": 5.9327,
      "mean_token_accuracy": 0.17676707804203035,
      "num_tokens": 51570630.0,
      "step": 20345
    },
    {
      "entropy": 6.228191184997558,
      "epoch": 2.3485285631852277,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044567392181978874,
      "loss": 5.8876,
      "mean_token_accuracy": 0.17099658101797105,
      "num_tokens": 51583463.0,
      "step": 20350
    },
    {
      "entropy": 6.2143758773803714,
      "epoch": 2.3491055972302366,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00044564703052240223,
      "loss": 5.9353,
      "mean_token_accuracy": 0.17696345150470733,
      "num_tokens": 51597090.0,
      "step": 20355
    },
    {
      "entropy": 6.219122362136841,
      "epoch": 2.349682631275245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004456201334853876,
      "loss": 5.8587,
      "mean_token_accuracy": 0.18090834617614746,
      "num_tokens": 51609288.0,
      "step": 20360
    },
    {
      "entropy": 6.1154890060424805,
      "epoch": 2.350259665320254,
      "grad_norm": 1.84375,
      "learning_rate": 0.00044559323070964956,
      "loss": 5.8541,
      "mean_token_accuracy": 0.1837046980857849,
      "num_tokens": 51622269.0,
      "step": 20365
    },
    {
      "entropy": 6.238179540634155,
      "epoch": 2.3508366993652627,
      "grad_norm": 1.078125,
      "learning_rate": 0.000445566322196093,
      "loss": 5.9468,
      "mean_token_accuracy": 0.1666563108563423,
      "num_tokens": 51634222.0,
      "step": 20370
    },
    {
      "entropy": 6.270152044296265,
      "epoch": 2.351413733410271,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004455394079456228,
      "loss": 5.9197,
      "mean_token_accuracy": 0.17454564124345778,
      "num_tokens": 51646933.0,
      "step": 20375
    },
    {
      "entropy": 6.2179069995880125,
      "epoch": 2.35199076745528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044551248795914446,
      "loss": 5.9337,
      "mean_token_accuracy": 0.17222847640514374,
      "num_tokens": 51658987.0,
      "step": 20380
    },
    {
      "entropy": 6.222844505310059,
      "epoch": 2.3525678015002884,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004454855622375632,
      "loss": 5.9092,
      "mean_token_accuracy": 0.17403256595134736,
      "num_tokens": 51669513.0,
      "step": 20385
    },
    {
      "entropy": 6.2505049228668215,
      "epoch": 2.3531448355452973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004454586307817848,
      "loss": 5.9076,
      "mean_token_accuracy": 0.17644617408514024,
      "num_tokens": 51682084.0,
      "step": 20390
    },
    {
      "entropy": 6.1657110214233395,
      "epoch": 2.3537218695903057,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000445431693592715,
      "loss": 5.9156,
      "mean_token_accuracy": 0.18160975426435472,
      "num_tokens": 51695335.0,
      "step": 20395
    },
    {
      "entropy": 6.303590488433838,
      "epoch": 2.3542989036353146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004454047506712598,
      "loss": 6.0019,
      "mean_token_accuracy": 0.16991567760705947,
      "num_tokens": 51707856.0,
      "step": 20400
    },
    {
      "entropy": 6.257295179367065,
      "epoch": 2.354875937680323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044537780201832545,
      "loss": 5.8853,
      "mean_token_accuracy": 0.1759468987584114,
      "num_tokens": 51719716.0,
      "step": 20405
    },
    {
      "entropy": 6.120660829544067,
      "epoch": 2.355452971725332,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004453508476348184,
      "loss": 5.8576,
      "mean_token_accuracy": 0.1873482272028923,
      "num_tokens": 51733263.0,
      "step": 20410
    },
    {
      "entropy": 6.171567296981811,
      "epoch": 2.3560300057703403,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004453238875216452,
      "loss": 5.8262,
      "mean_token_accuracy": 0.18494777530431747,
      "num_tokens": 51745143.0,
      "step": 20415
    },
    {
      "entropy": 6.264321565628052,
      "epoch": 2.356607039815349,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004452969216797127,
      "loss": 5.9456,
      "mean_token_accuracy": 0.17102426737546922,
      "num_tokens": 51757155.0,
      "step": 20420
    },
    {
      "entropy": 6.178544187545777,
      "epoch": 2.357184073860358,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004452699501099277,
      "loss": 5.8602,
      "mean_token_accuracy": 0.1825183629989624,
      "num_tokens": 51769170.0,
      "step": 20425
    },
    {
      "entropy": 6.224378490447998,
      "epoch": 2.3577611079053664,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044524297281319766,
      "loss": 5.9331,
      "mean_token_accuracy": 0.17603640556335448,
      "num_tokens": 51781254.0,
      "step": 20430
    },
    {
      "entropy": 6.27843451499939,
      "epoch": 2.358338141950375,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044521598979042963,
      "loss": 5.9701,
      "mean_token_accuracy": 0.17067469358444215,
      "num_tokens": 51793425.0,
      "step": 20435
    },
    {
      "entropy": 6.208112382888794,
      "epoch": 2.3589151759953837,
      "grad_norm": 1.125,
      "learning_rate": 0.00044518900104253154,
      "loss": 5.8863,
      "mean_token_accuracy": 0.1794102519750595,
      "num_tokens": 51806196.0,
      "step": 20440
    },
    {
      "entropy": 6.08141679763794,
      "epoch": 2.3594922100403926,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004451620065704108,
      "loss": 5.909,
      "mean_token_accuracy": 0.1794336900115013,
      "num_tokens": 51819137.0,
      "step": 20445
    },
    {
      "entropy": 6.230327129364014,
      "epoch": 2.360069244085401,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044513500637497546,
      "loss": 5.9676,
      "mean_token_accuracy": 0.17364402115345,
      "num_tokens": 51830924.0,
      "step": 20450
    },
    {
      "entropy": 6.2120137214660645,
      "epoch": 2.36064627813041,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044510800045713373,
      "loss": 5.8613,
      "mean_token_accuracy": 0.17585181295871735,
      "num_tokens": 51843503.0,
      "step": 20455
    },
    {
      "entropy": 6.174780321121216,
      "epoch": 2.3612233121754183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044508098881779396,
      "loss": 5.9327,
      "mean_token_accuracy": 0.17654947489500045,
      "num_tokens": 51855818.0,
      "step": 20460
    },
    {
      "entropy": 6.192159175872803,
      "epoch": 2.361800346220427,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004450539714578645,
      "loss": 5.9015,
      "mean_token_accuracy": 0.18018826544284822,
      "num_tokens": 51867033.0,
      "step": 20465
    },
    {
      "entropy": 6.2804539680480955,
      "epoch": 2.3623773802654355,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004450269483782543,
      "loss": 5.9246,
      "mean_token_accuracy": 0.1715424582362175,
      "num_tokens": 51880845.0,
      "step": 20470
    },
    {
      "entropy": 6.175959062576294,
      "epoch": 2.3629544143104444,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004449999195798721,
      "loss": 5.9509,
      "mean_token_accuracy": 0.1731112152338028,
      "num_tokens": 51893761.0,
      "step": 20475
    },
    {
      "entropy": 6.216841125488282,
      "epoch": 2.363531448355453,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044497288506362706,
      "loss": 5.9011,
      "mean_token_accuracy": 0.1775978684425354,
      "num_tokens": 51907478.0,
      "step": 20480
    },
    {
      "entropy": 6.25024037361145,
      "epoch": 2.3641084824004617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004449458448304284,
      "loss": 5.8986,
      "mean_token_accuracy": 0.18205091655254363,
      "num_tokens": 51919242.0,
      "step": 20485
    },
    {
      "entropy": 6.244975900650024,
      "epoch": 2.36468551644547,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044491879888118574,
      "loss": 5.9593,
      "mean_token_accuracy": 0.17636770755052567,
      "num_tokens": 51930791.0,
      "step": 20490
    },
    {
      "entropy": 6.109720230102539,
      "epoch": 2.365262550490479,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004448917472168087,
      "loss": 5.7711,
      "mean_token_accuracy": 0.1822209522128105,
      "num_tokens": 51943848.0,
      "step": 20495
    },
    {
      "entropy": 6.246545028686524,
      "epoch": 2.365839584535488,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044486468983820707,
      "loss": 5.9557,
      "mean_token_accuracy": 0.1686646580696106,
      "num_tokens": 51956648.0,
      "step": 20500
    },
    {
      "entropy": 6.234007215499878,
      "epoch": 2.3664166185804962,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000444837626746291,
      "loss": 5.9126,
      "mean_token_accuracy": 0.17662867456674575,
      "num_tokens": 51969068.0,
      "step": 20505
    },
    {
      "entropy": 6.211513042449951,
      "epoch": 2.366993652625505,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004448105579419707,
      "loss": 5.9168,
      "mean_token_accuracy": 0.17582010626792907,
      "num_tokens": 51981856.0,
      "step": 20510
    },
    {
      "entropy": 6.257792949676514,
      "epoch": 2.3675706866705135,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004447834834261567,
      "loss": 5.9791,
      "mean_token_accuracy": 0.16998671889305114,
      "num_tokens": 51994428.0,
      "step": 20515
    },
    {
      "entropy": 6.1701915740966795,
      "epoch": 2.3681477207155224,
      "grad_norm": 1.0,
      "learning_rate": 0.0004447564031997595,
      "loss": 5.8249,
      "mean_token_accuracy": 0.1855105385184288,
      "num_tokens": 52007875.0,
      "step": 20520
    },
    {
      "entropy": 6.1876523971557615,
      "epoch": 2.368724754760531,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044472931726369,
      "loss": 5.8072,
      "mean_token_accuracy": 0.18526231050491332,
      "num_tokens": 52020876.0,
      "step": 20525
    },
    {
      "entropy": 6.313043832778931,
      "epoch": 2.3693017888055397,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044470222561885926,
      "loss": 5.9765,
      "mean_token_accuracy": 0.16895418912172316,
      "num_tokens": 52032056.0,
      "step": 20530
    },
    {
      "entropy": 6.229685974121094,
      "epoch": 2.369878822850548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044467512826617845,
      "loss": 5.9194,
      "mean_token_accuracy": 0.17216948568820953,
      "num_tokens": 52044488.0,
      "step": 20535
    },
    {
      "entropy": 6.194803714752197,
      "epoch": 2.370455856895557,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044464802520655897,
      "loss": 5.918,
      "mean_token_accuracy": 0.17740179896354674,
      "num_tokens": 52056186.0,
      "step": 20540
    },
    {
      "entropy": 6.27158522605896,
      "epoch": 2.3710328909405654,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004446209164409124,
      "loss": 5.8929,
      "mean_token_accuracy": 0.17821502685546875,
      "num_tokens": 52068218.0,
      "step": 20545
    },
    {
      "entropy": 6.1447595119476315,
      "epoch": 2.371609924985574,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004445938019701506,
      "loss": 5.8197,
      "mean_token_accuracy": 0.19037968665361404,
      "num_tokens": 52081854.0,
      "step": 20550
    },
    {
      "entropy": 6.138282728195191,
      "epoch": 2.3721869590305826,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004445666817951855,
      "loss": 5.8201,
      "mean_token_accuracy": 0.1810576945543289,
      "num_tokens": 52094022.0,
      "step": 20555
    },
    {
      "entropy": 6.2049354076385494,
      "epoch": 2.3727639930755915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004445395559169293,
      "loss": 5.8504,
      "mean_token_accuracy": 0.18278271555900574,
      "num_tokens": 52105677.0,
      "step": 20560
    },
    {
      "entropy": 6.20401782989502,
      "epoch": 2.3733410271206,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004445124243362943,
      "loss": 5.9219,
      "mean_token_accuracy": 0.1751824989914894,
      "num_tokens": 52118039.0,
      "step": 20565
    },
    {
      "entropy": 6.217830038070678,
      "epoch": 2.3739180611656088,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004444852870541932,
      "loss": 5.8951,
      "mean_token_accuracy": 0.18027360439300538,
      "num_tokens": 52131519.0,
      "step": 20570
    },
    {
      "entropy": 6.223957586288452,
      "epoch": 2.3744950952106176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004444581440715386,
      "loss": 5.9052,
      "mean_token_accuracy": 0.17486371994018554,
      "num_tokens": 52143240.0,
      "step": 20575
    },
    {
      "entropy": 6.156794452667237,
      "epoch": 2.375072129255626,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044443099538924347,
      "loss": 5.8348,
      "mean_token_accuracy": 0.1791414126753807,
      "num_tokens": 52155977.0,
      "step": 20580
    },
    {
      "entropy": 6.174764204025268,
      "epoch": 2.375649163300635,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004444038410082211,
      "loss": 5.8923,
      "mean_token_accuracy": 0.1771896630525589,
      "num_tokens": 52169388.0,
      "step": 20585
    },
    {
      "entropy": 6.1451273441314695,
      "epoch": 2.3762261973456433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004443766809293846,
      "loss": 5.8173,
      "mean_token_accuracy": 0.1791497439146042,
      "num_tokens": 52181830.0,
      "step": 20590
    },
    {
      "entropy": 6.145556116104126,
      "epoch": 2.376803231390652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004443495151536475,
      "loss": 5.7948,
      "mean_token_accuracy": 0.18753628581762313,
      "num_tokens": 52194255.0,
      "step": 20595
    },
    {
      "entropy": 6.218704557418823,
      "epoch": 2.3773802654356606,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004443223436819237,
      "loss": 5.9057,
      "mean_token_accuracy": 0.17673536837100984,
      "num_tokens": 52205646.0,
      "step": 20600
    },
    {
      "entropy": 6.216388416290283,
      "epoch": 2.3779572994806695,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044429516651512687,
      "loss": 5.8749,
      "mean_token_accuracy": 0.17649720460176468,
      "num_tokens": 52217519.0,
      "step": 20605
    },
    {
      "entropy": 6.181417036056518,
      "epoch": 2.378534333525678,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044426798365417133,
      "loss": 5.9024,
      "mean_token_accuracy": 0.18006114661693573,
      "num_tokens": 52230139.0,
      "step": 20610
    },
    {
      "entropy": 6.159763383865356,
      "epoch": 2.3791113675706868,
      "grad_norm": 1.578125,
      "learning_rate": 0.00044424079509997104,
      "loss": 6.0185,
      "mean_token_accuracy": 0.1719463661313057,
      "num_tokens": 52244049.0,
      "step": 20615
    },
    {
      "entropy": 6.236967277526856,
      "epoch": 2.379688401615695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004442136008534409,
      "loss": 5.902,
      "mean_token_accuracy": 0.18061581999063492,
      "num_tokens": 52257180.0,
      "step": 20620
    },
    {
      "entropy": 6.223541164398194,
      "epoch": 2.380265435660704,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044418640091549525,
      "loss": 5.9095,
      "mean_token_accuracy": 0.1799600303173065,
      "num_tokens": 52270260.0,
      "step": 20625
    },
    {
      "entropy": 6.080956268310547,
      "epoch": 2.3808424697057124,
      "grad_norm": 1.078125,
      "learning_rate": 0.000444159195287049,
      "loss": 5.8505,
      "mean_token_accuracy": 0.18108619451522828,
      "num_tokens": 52282335.0,
      "step": 20630
    },
    {
      "entropy": 6.247951030731201,
      "epoch": 2.3814195037507213,
      "grad_norm": 1.125,
      "learning_rate": 0.0004441319839690173,
      "loss": 5.8854,
      "mean_token_accuracy": 0.17612518668174743,
      "num_tokens": 52296547.0,
      "step": 20635
    },
    {
      "entropy": 6.2390968799591064,
      "epoch": 2.3819965377957297,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004441047669623153,
      "loss": 5.933,
      "mean_token_accuracy": 0.1712224617600441,
      "num_tokens": 52309523.0,
      "step": 20640
    },
    {
      "entropy": 6.192690086364746,
      "epoch": 2.3825735718407386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044407754426785845,
      "loss": 5.8802,
      "mean_token_accuracy": 0.1796162709593773,
      "num_tokens": 52322476.0,
      "step": 20645
    },
    {
      "entropy": 6.12495059967041,
      "epoch": 2.3831506058857475,
      "grad_norm": 2.125,
      "learning_rate": 0.0004440503158865625,
      "loss": 5.8206,
      "mean_token_accuracy": 0.18461779803037642,
      "num_tokens": 52335919.0,
      "step": 20650
    },
    {
      "entropy": 6.224644613265991,
      "epoch": 2.383727639930756,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044402308181934295,
      "loss": 5.8806,
      "mean_token_accuracy": 0.1802664205431938,
      "num_tokens": 52347636.0,
      "step": 20655
    },
    {
      "entropy": 6.275594902038574,
      "epoch": 2.3843046739757647,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004439958420671162,
      "loss": 5.8984,
      "mean_token_accuracy": 0.1733878329396248,
      "num_tokens": 52360181.0,
      "step": 20660
    },
    {
      "entropy": 6.190026378631591,
      "epoch": 2.384881708020773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044396859663079814,
      "loss": 5.9203,
      "mean_token_accuracy": 0.18247169107198716,
      "num_tokens": 52373525.0,
      "step": 20665
    },
    {
      "entropy": 6.126813983917236,
      "epoch": 2.385458742065782,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044394134551130533,
      "loss": 5.7935,
      "mean_token_accuracy": 0.18269770294427873,
      "num_tokens": 52386252.0,
      "step": 20670
    },
    {
      "entropy": 6.168920612335205,
      "epoch": 2.3860357761107904,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004439140887095542,
      "loss": 5.8237,
      "mean_token_accuracy": 0.18038810044527054,
      "num_tokens": 52398056.0,
      "step": 20675
    },
    {
      "entropy": 6.160497331619263,
      "epoch": 2.3866128101557993,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044388682622646165,
      "loss": 5.8866,
      "mean_token_accuracy": 0.18339675068855285,
      "num_tokens": 52409861.0,
      "step": 20680
    },
    {
      "entropy": 6.153534173965454,
      "epoch": 2.3871898442008077,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004438595580629446,
      "loss": 5.8642,
      "mean_token_accuracy": 0.17689687460660936,
      "num_tokens": 52423252.0,
      "step": 20685
    },
    {
      "entropy": 6.231155490875244,
      "epoch": 2.3877668782458166,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004438322842199202,
      "loss": 5.8119,
      "mean_token_accuracy": 0.18488748669624328,
      "num_tokens": 52435950.0,
      "step": 20690
    },
    {
      "entropy": 6.181946468353272,
      "epoch": 2.388343912290825,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004438050046983057,
      "loss": 5.8772,
      "mean_token_accuracy": 0.18326539695262908,
      "num_tokens": 52449257.0,
      "step": 20695
    },
    {
      "entropy": 6.129793930053711,
      "epoch": 2.388920946335834,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00044377771949901876,
      "loss": 5.8776,
      "mean_token_accuracy": 0.1826947495341301,
      "num_tokens": 52461543.0,
      "step": 20700
    },
    {
      "entropy": 6.034094667434692,
      "epoch": 2.3894979803808427,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044375042862297703,
      "loss": 5.7827,
      "mean_token_accuracy": 0.18628005534410477,
      "num_tokens": 52473752.0,
      "step": 20705
    },
    {
      "entropy": 6.250870418548584,
      "epoch": 2.390075014425851,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044372313207109856,
      "loss": 5.936,
      "mean_token_accuracy": 0.17118626981973648,
      "num_tokens": 52485885.0,
      "step": 20710
    },
    {
      "entropy": 6.232921504974366,
      "epoch": 2.3906520484708595,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044369582984430127,
      "loss": 5.8967,
      "mean_token_accuracy": 0.17732545882463455,
      "num_tokens": 52498389.0,
      "step": 20715
    },
    {
      "entropy": 6.161101770401001,
      "epoch": 2.3912290825158684,
      "grad_norm": 1.125,
      "learning_rate": 0.00044366852194350354,
      "loss": 5.9196,
      "mean_token_accuracy": 0.17828254997730256,
      "num_tokens": 52511176.0,
      "step": 20720
    },
    {
      "entropy": 6.2019850730896,
      "epoch": 2.3918061165608773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004436412083696239,
      "loss": 5.858,
      "mean_token_accuracy": 0.17637231349945068,
      "num_tokens": 52523173.0,
      "step": 20725
    },
    {
      "entropy": 6.180507469177246,
      "epoch": 2.3923831506058857,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044361388912358095,
      "loss": 5.8576,
      "mean_token_accuracy": 0.1753206104040146,
      "num_tokens": 52535776.0,
      "step": 20730
    },
    {
      "entropy": 6.0130421161651615,
      "epoch": 2.3929601846508946,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004435865642062936,
      "loss": 5.6563,
      "mean_token_accuracy": 0.19541964679956436,
      "num_tokens": 52549416.0,
      "step": 20735
    },
    {
      "entropy": 6.154394340515137,
      "epoch": 2.393537218695903,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004435592336186809,
      "loss": 5.8559,
      "mean_token_accuracy": 0.18084726482629776,
      "num_tokens": 52562300.0,
      "step": 20740
    },
    {
      "entropy": 6.264912223815918,
      "epoch": 2.394114252740912,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004435318973616622,
      "loss": 5.9503,
      "mean_token_accuracy": 0.17170136272907258,
      "num_tokens": 52573906.0,
      "step": 20745
    },
    {
      "entropy": 6.226292419433594,
      "epoch": 2.3946912867859202,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044350455543615665,
      "loss": 5.848,
      "mean_token_accuracy": 0.18819532096385955,
      "num_tokens": 52586138.0,
      "step": 20750
    },
    {
      "entropy": 6.221322345733642,
      "epoch": 2.395268320830929,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004434772078430843,
      "loss": 5.9211,
      "mean_token_accuracy": 0.17501968443393706,
      "num_tokens": 52599705.0,
      "step": 20755
    },
    {
      "entropy": 6.20050082206726,
      "epoch": 2.3958453548759375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004434498545833646,
      "loss": 5.9694,
      "mean_token_accuracy": 0.17441977709531784,
      "num_tokens": 52612041.0,
      "step": 20760
    },
    {
      "entropy": 6.275346374511718,
      "epoch": 2.3964223889209464,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004434224956579177,
      "loss": 5.9354,
      "mean_token_accuracy": 0.1756373316049576,
      "num_tokens": 52625634.0,
      "step": 20765
    },
    {
      "entropy": 6.119609260559082,
      "epoch": 2.396999422965955,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004433951310676639,
      "loss": 5.7395,
      "mean_token_accuracy": 0.19740791916847228,
      "num_tokens": 52638706.0,
      "step": 20770
    },
    {
      "entropy": 6.067688941955566,
      "epoch": 2.3975764570109637,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044336776081352347,
      "loss": 5.8935,
      "mean_token_accuracy": 0.18710660338401794,
      "num_tokens": 52650876.0,
      "step": 20775
    },
    {
      "entropy": 6.181147623062134,
      "epoch": 2.3981534910559725,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044334038489641706,
      "loss": 5.8468,
      "mean_token_accuracy": 0.18153303414583205,
      "num_tokens": 52663664.0,
      "step": 20780
    },
    {
      "entropy": 6.249605894088745,
      "epoch": 2.398730525100981,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00044331300331726544,
      "loss": 5.969,
      "mean_token_accuracy": 0.17528854310512543,
      "num_tokens": 52675870.0,
      "step": 20785
    },
    {
      "entropy": 6.19515962600708,
      "epoch": 2.3993075591459894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044328561607698947,
      "loss": 5.9009,
      "mean_token_accuracy": 0.17321067303419113,
      "num_tokens": 52687029.0,
      "step": 20790
    },
    {
      "entropy": 6.1399908542633055,
      "epoch": 2.3998845931909982,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004432582231765105,
      "loss": 5.7928,
      "mean_token_accuracy": 0.1765931084752083,
      "num_tokens": 52700242.0,
      "step": 20795
    },
    {
      "entropy": 6.125870227813721,
      "epoch": 2.400461627236007,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044323082461674974,
      "loss": 5.8552,
      "mean_token_accuracy": 0.1790474236011505,
      "num_tokens": 52711561.0,
      "step": 20800
    },
    {
      "entropy": 6.098082399368286,
      "epoch": 2.4010386612810155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432034203986287,
      "loss": 5.8224,
      "mean_token_accuracy": 0.179676416516304,
      "num_tokens": 52725101.0,
      "step": 20805
    },
    {
      "entropy": 6.260610103607178,
      "epoch": 2.4016156953260244,
      "grad_norm": 1.25,
      "learning_rate": 0.0004431760105230693,
      "loss": 5.9327,
      "mean_token_accuracy": 0.1707353785634041,
      "num_tokens": 52738315.0,
      "step": 20810
    },
    {
      "entropy": 6.253695631027222,
      "epoch": 2.402192729371033,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044314859499099325,
      "loss": 6.0138,
      "mean_token_accuracy": 0.1738351672887802,
      "num_tokens": 52752547.0,
      "step": 20815
    },
    {
      "entropy": 6.238799381256103,
      "epoch": 2.4027697634160416,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044312117380332274,
      "loss": 5.9741,
      "mean_token_accuracy": 0.17198569625616072,
      "num_tokens": 52765831.0,
      "step": 20820
    },
    {
      "entropy": 6.203983211517334,
      "epoch": 2.40334679746105,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044309374696098,
      "loss": 5.8477,
      "mean_token_accuracy": 0.1774840086698532,
      "num_tokens": 52779144.0,
      "step": 20825
    },
    {
      "entropy": 6.17804536819458,
      "epoch": 2.403923831506059,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004430663144648876,
      "loss": 5.8744,
      "mean_token_accuracy": 0.18458810150623323,
      "num_tokens": 52790956.0,
      "step": 20830
    },
    {
      "entropy": 6.212011384963989,
      "epoch": 2.4045008655510673,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044303887631596823,
      "loss": 5.9118,
      "mean_token_accuracy": 0.1782439947128296,
      "num_tokens": 52804057.0,
      "step": 20835
    },
    {
      "entropy": 6.2021739959716795,
      "epoch": 2.405077899596076,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004430114325151447,
      "loss": 5.8207,
      "mean_token_accuracy": 0.17667302638292312,
      "num_tokens": 52817197.0,
      "step": 20840
    },
    {
      "entropy": 6.252050304412842,
      "epoch": 2.4056549336410846,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004429839830633401,
      "loss": 5.9263,
      "mean_token_accuracy": 0.17479386329650878,
      "num_tokens": 52830162.0,
      "step": 20845
    },
    {
      "entropy": 6.147500848770141,
      "epoch": 2.4062319676860935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004429565279614777,
      "loss": 5.9127,
      "mean_token_accuracy": 0.17931033670902252,
      "num_tokens": 52843681.0,
      "step": 20850
    },
    {
      "entropy": 6.217716884613037,
      "epoch": 2.4068090017311023,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044292906721048094,
      "loss": 5.9099,
      "mean_token_accuracy": 0.18140170574188233,
      "num_tokens": 52857297.0,
      "step": 20855
    },
    {
      "entropy": 6.04101710319519,
      "epoch": 2.4073860357761108,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004429016008112733,
      "loss": 5.6437,
      "mean_token_accuracy": 0.1920620620250702,
      "num_tokens": 52870551.0,
      "step": 20860
    },
    {
      "entropy": 6.204928207397461,
      "epoch": 2.4079630698211196,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004428741287647788,
      "loss": 5.8678,
      "mean_token_accuracy": 0.18151758164167403,
      "num_tokens": 52884628.0,
      "step": 20865
    },
    {
      "entropy": 6.219751262664795,
      "epoch": 2.408540103866128,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044284665107192136,
      "loss": 5.8465,
      "mean_token_accuracy": 0.18571862429380417,
      "num_tokens": 52897079.0,
      "step": 20870
    },
    {
      "entropy": 6.2039085865020756,
      "epoch": 2.409117137911137,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004428191677336251,
      "loss": 5.9341,
      "mean_token_accuracy": 0.1741613671183586,
      "num_tokens": 52909990.0,
      "step": 20875
    },
    {
      "entropy": 6.224388694763183,
      "epoch": 2.4096941719561453,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004427916787508146,
      "loss": 5.9164,
      "mean_token_accuracy": 0.17620352655649185,
      "num_tokens": 52922732.0,
      "step": 20880
    },
    {
      "entropy": 6.319819927215576,
      "epoch": 2.410271206001154,
      "grad_norm": 1.0,
      "learning_rate": 0.0004427641841244144,
      "loss": 5.9392,
      "mean_token_accuracy": 0.17118202596902848,
      "num_tokens": 52934648.0,
      "step": 20885
    },
    {
      "entropy": 6.240547323226929,
      "epoch": 2.4108482400461626,
      "grad_norm": 1.625,
      "learning_rate": 0.000442736683855349,
      "loss": 5.9929,
      "mean_token_accuracy": 0.1720513328909874,
      "num_tokens": 52947913.0,
      "step": 20890
    },
    {
      "entropy": 6.202064085006714,
      "epoch": 2.4114252740911715,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004427091779445437,
      "loss": 5.9067,
      "mean_token_accuracy": 0.17776024490594863,
      "num_tokens": 52960491.0,
      "step": 20895
    },
    {
      "entropy": 6.2297093868255615,
      "epoch": 2.41200230813618,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004426816663929235,
      "loss": 5.8968,
      "mean_token_accuracy": 0.17991008013486862,
      "num_tokens": 52972813.0,
      "step": 20900
    },
    {
      "entropy": 6.256739854812622,
      "epoch": 2.4125793421811887,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044265414920141366,
      "loss": 5.9157,
      "mean_token_accuracy": 0.17677846252918245,
      "num_tokens": 52985548.0,
      "step": 20905
    },
    {
      "entropy": 6.209710121154785,
      "epoch": 2.413156376226197,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044262662637093987,
      "loss": 5.8355,
      "mean_token_accuracy": 0.18242569118738175,
      "num_tokens": 52998911.0,
      "step": 20910
    },
    {
      "entropy": 6.094181299209595,
      "epoch": 2.413733410271206,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00044259909790242776,
      "loss": 5.8137,
      "mean_token_accuracy": 0.18979533314704894,
      "num_tokens": 53011205.0,
      "step": 20915
    },
    {
      "entropy": 6.189552450180054,
      "epoch": 2.4143104443162144,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004425715637968032,
      "loss": 5.8986,
      "mean_token_accuracy": 0.17822712957859038,
      "num_tokens": 53023428.0,
      "step": 20920
    },
    {
      "entropy": 6.286620140075684,
      "epoch": 2.4148874783612233,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004425440240549923,
      "loss": 5.9922,
      "mean_token_accuracy": 0.16845131367444993,
      "num_tokens": 53035488.0,
      "step": 20925
    },
    {
      "entropy": 6.119616508483887,
      "epoch": 2.415464512406232,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044251647867792147,
      "loss": 5.7615,
      "mean_token_accuracy": 0.19196383357048036,
      "num_tokens": 53048416.0,
      "step": 20930
    },
    {
      "entropy": 6.206809568405151,
      "epoch": 2.4160415464512406,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044248892766651706,
      "loss": 6.012,
      "mean_token_accuracy": 0.17298592031002044,
      "num_tokens": 53061763.0,
      "step": 20935
    },
    {
      "entropy": 6.2102940559387205,
      "epoch": 2.4166185804962494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004424613710217057,
      "loss": 5.8226,
      "mean_token_accuracy": 0.1838536038994789,
      "num_tokens": 53073666.0,
      "step": 20940
    },
    {
      "entropy": 6.109739542007446,
      "epoch": 2.417195614541258,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044243380874441437,
      "loss": 5.7945,
      "mean_token_accuracy": 0.1887991577386856,
      "num_tokens": 53087155.0,
      "step": 20945
    },
    {
      "entropy": 6.188466739654541,
      "epoch": 2.4177726485862667,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044240624083557,
      "loss": 5.9308,
      "mean_token_accuracy": 0.17866060733795167,
      "num_tokens": 53099199.0,
      "step": 20950
    },
    {
      "entropy": 6.232250165939331,
      "epoch": 2.418349682631275,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044237866729609994,
      "loss": 5.9363,
      "mean_token_accuracy": 0.17470151782035828,
      "num_tokens": 53111997.0,
      "step": 20955
    },
    {
      "entropy": 6.067967510223388,
      "epoch": 2.418926716676284,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004423510881269315,
      "loss": 5.8507,
      "mean_token_accuracy": 0.1857800379395485,
      "num_tokens": 53123812.0,
      "step": 20960
    },
    {
      "entropy": 6.243478584289551,
      "epoch": 2.4195037507212924,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004423235033289924,
      "loss": 5.9855,
      "mean_token_accuracy": 0.1670878440141678,
      "num_tokens": 53137232.0,
      "step": 20965
    },
    {
      "entropy": 6.27320122718811,
      "epoch": 2.4200807847663013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004422959129032103,
      "loss": 5.9064,
      "mean_token_accuracy": 0.17410949617624283,
      "num_tokens": 53149183.0,
      "step": 20970
    },
    {
      "entropy": 6.081487798690796,
      "epoch": 2.4206578188113097,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044226831685051333,
      "loss": 5.8092,
      "mean_token_accuracy": 0.18246926218271256,
      "num_tokens": 53163187.0,
      "step": 20975
    },
    {
      "entropy": 6.272024440765381,
      "epoch": 2.4212348528563186,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004422407151718296,
      "loss": 5.912,
      "mean_token_accuracy": 0.18224563598632812,
      "num_tokens": 53178441.0,
      "step": 20980
    },
    {
      "entropy": 6.255713939666748,
      "epoch": 2.4218118869013274,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044221310786808746,
      "loss": 5.8991,
      "mean_token_accuracy": 0.17704823911190032,
      "num_tokens": 53190042.0,
      "step": 20985
    },
    {
      "entropy": 6.235786724090576,
      "epoch": 2.422388920946336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004421854949402155,
      "loss": 5.887,
      "mean_token_accuracy": 0.17682436257600784,
      "num_tokens": 53201724.0,
      "step": 20990
    },
    {
      "entropy": 6.165446186065674,
      "epoch": 2.4229659549913443,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044215787638914245,
      "loss": 5.867,
      "mean_token_accuracy": 0.17937190979719161,
      "num_tokens": 53213415.0,
      "step": 20995
    },
    {
      "entropy": 6.219665288925171,
      "epoch": 2.423542989036353,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004421302522157973,
      "loss": 5.9,
      "mean_token_accuracy": 0.1843603178858757,
      "num_tokens": 53227324.0,
      "step": 21000
    },
    {
      "epoch": 2.423542989036353,
      "eval_entropy": 6.076491791109894,
      "eval_loss": 6.019885540008545,
      "eval_mean_token_accuracy": 0.17840918350434112,
      "eval_num_tokens": 53227324.0,
      "eval_runtime": 22.7497,
      "eval_samples_per_second": 1236.761,
      "eval_steps_per_second": 154.595,
      "step": 21000
    },
    {
      "entropy": 6.230165576934814,
      "epoch": 2.424120023081362,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004421026224211091,
      "loss": 5.934,
      "mean_token_accuracy": 0.17129454314708709,
      "num_tokens": 53240325.0,
      "step": 21005
    },
    {
      "entropy": 6.134762001037598,
      "epoch": 2.4246970571263704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044207498700600724,
      "loss": 5.7844,
      "mean_token_accuracy": 0.18430345356464387,
      "num_tokens": 53251857.0,
      "step": 21010
    },
    {
      "entropy": 6.222596931457519,
      "epoch": 2.4252740911713793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044204734597142115,
      "loss": 5.9991,
      "mean_token_accuracy": 0.17335744947195053,
      "num_tokens": 53266309.0,
      "step": 21015
    },
    {
      "entropy": 6.2669469833374025,
      "epoch": 2.4258511252163877,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044201969931828057,
      "loss": 5.8677,
      "mean_token_accuracy": 0.18234310448169708,
      "num_tokens": 53278085.0,
      "step": 21020
    },
    {
      "entropy": 6.2235143184661865,
      "epoch": 2.4264281592613965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004419920470475154,
      "loss": 5.8775,
      "mean_token_accuracy": 0.1721951812505722,
      "num_tokens": 53291395.0,
      "step": 21025
    },
    {
      "entropy": 6.277893877029419,
      "epoch": 2.427005193306405,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004419643891600556,
      "loss": 6.0218,
      "mean_token_accuracy": 0.17307724058628082,
      "num_tokens": 53304735.0,
      "step": 21030
    },
    {
      "entropy": 6.2512092113494875,
      "epoch": 2.427582227351414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044193672565683153,
      "loss": 5.9068,
      "mean_token_accuracy": 0.18121398389339446,
      "num_tokens": 53316467.0,
      "step": 21035
    },
    {
      "entropy": 6.135804843902588,
      "epoch": 2.4281592613964222,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004419090565387736,
      "loss": 5.8901,
      "mean_token_accuracy": 0.1738771915435791,
      "num_tokens": 53328591.0,
      "step": 21040
    },
    {
      "entropy": 6.252773952484131,
      "epoch": 2.428736295441431,
      "grad_norm": 1.34375,
      "learning_rate": 0.00044188138180681255,
      "loss": 5.9323,
      "mean_token_accuracy": 0.17573420852422714,
      "num_tokens": 53340640.0,
      "step": 21045
    },
    {
      "entropy": 6.271492004394531,
      "epoch": 2.4293133294864395,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000441853701461879,
      "loss": 5.9015,
      "mean_token_accuracy": 0.17789920717477797,
      "num_tokens": 53351798.0,
      "step": 21050
    },
    {
      "entropy": 6.239266872406006,
      "epoch": 2.4298903635314484,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044182601550490414,
      "loss": 5.9497,
      "mean_token_accuracy": 0.17646777331829072,
      "num_tokens": 53364994.0,
      "step": 21055
    },
    {
      "entropy": 6.318173408508301,
      "epoch": 2.4304673975764572,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004417983239368192,
      "loss": 6.0036,
      "mean_token_accuracy": 0.17207963168621063,
      "num_tokens": 53376729.0,
      "step": 21060
    },
    {
      "entropy": 6.222643661499023,
      "epoch": 2.4310444316214657,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044177062675855534,
      "loss": 5.9576,
      "mean_token_accuracy": 0.17806044220924377,
      "num_tokens": 53389723.0,
      "step": 21065
    },
    {
      "entropy": 6.229648637771606,
      "epoch": 2.431621465666474,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004417429239710444,
      "loss": 5.9615,
      "mean_token_accuracy": 0.17081690430641175,
      "num_tokens": 53402102.0,
      "step": 21070
    },
    {
      "entropy": 6.1297372341156,
      "epoch": 2.432198499711483,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004417152155752179,
      "loss": 5.7789,
      "mean_token_accuracy": 0.19271715134382247,
      "num_tokens": 53416246.0,
      "step": 21075
    },
    {
      "entropy": 6.21894941329956,
      "epoch": 2.432775533756492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004416875015720081,
      "loss": 5.8315,
      "mean_token_accuracy": 0.17722507417201996,
      "num_tokens": 53430666.0,
      "step": 21080
    },
    {
      "entropy": 6.166200828552246,
      "epoch": 2.4333525678015,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004416597819623469,
      "loss": 5.8011,
      "mean_token_accuracy": 0.18992147594690323,
      "num_tokens": 53442302.0,
      "step": 21085
    },
    {
      "entropy": 6.184221410751343,
      "epoch": 2.433929601846509,
      "grad_norm": 1.0,
      "learning_rate": 0.00044163205674716666,
      "loss": 5.9174,
      "mean_token_accuracy": 0.17275547981262207,
      "num_tokens": 53454288.0,
      "step": 21090
    },
    {
      "entropy": 6.289410352706909,
      "epoch": 2.4345066358915175,
      "grad_norm": 1.0,
      "learning_rate": 0.00044160432592740015,
      "loss": 5.9595,
      "mean_token_accuracy": 0.17322028130292894,
      "num_tokens": 53465307.0,
      "step": 21095
    },
    {
      "entropy": 6.198148584365844,
      "epoch": 2.4350836699365264,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044157658950397965,
      "loss": 5.828,
      "mean_token_accuracy": 0.18212511390447617,
      "num_tokens": 53478465.0,
      "step": 21100
    },
    {
      "entropy": 6.217931413650513,
      "epoch": 2.435660703981535,
      "grad_norm": 1.125,
      "learning_rate": 0.00044154884747783844,
      "loss": 5.9445,
      "mean_token_accuracy": 0.17746067941188812,
      "num_tokens": 53490783.0,
      "step": 21105
    },
    {
      "entropy": 6.186709356307984,
      "epoch": 2.4362377380265436,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00044152109984990954,
      "loss": 5.824,
      "mean_token_accuracy": 0.18825887739658356,
      "num_tokens": 53502716.0,
      "step": 21110
    },
    {
      "entropy": 6.182549142837525,
      "epoch": 2.436814772071552,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000441493346621126,
      "loss": 5.9456,
      "mean_token_accuracy": 0.17565467059612275,
      "num_tokens": 53514962.0,
      "step": 21115
    },
    {
      "entropy": 6.263493394851684,
      "epoch": 2.437391806116561,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044146558779242155,
      "loss": 5.9457,
      "mean_token_accuracy": 0.17664034068584442,
      "num_tokens": 53527741.0,
      "step": 21120
    },
    {
      "entropy": 6.155855655670166,
      "epoch": 2.4379688401615693,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004414378233647298,
      "loss": 5.8943,
      "mean_token_accuracy": 0.17725570648908615,
      "num_tokens": 53539705.0,
      "step": 21125
    },
    {
      "entropy": 6.167147397994995,
      "epoch": 2.438545874206578,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00044141005333898437,
      "loss": 5.8583,
      "mean_token_accuracy": 0.1794601321220398,
      "num_tokens": 53552310.0,
      "step": 21130
    },
    {
      "entropy": 6.176887559890747,
      "epoch": 2.439122908251587,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004413822777161196,
      "loss": 5.8445,
      "mean_token_accuracy": 0.1860383003950119,
      "num_tokens": 53565553.0,
      "step": 21135
    },
    {
      "entropy": 6.2276452541351315,
      "epoch": 2.4396999422965955,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004413544964970695,
      "loss": 5.9259,
      "mean_token_accuracy": 0.17372694462537766,
      "num_tokens": 53578837.0,
      "step": 21140
    },
    {
      "entropy": 6.206234121322632,
      "epoch": 2.4402769763416043,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004413267096827686,
      "loss": 5.8569,
      "mean_token_accuracy": 0.18036649823188783,
      "num_tokens": 53591286.0,
      "step": 21145
    },
    {
      "entropy": 6.228763246536255,
      "epoch": 2.4408540103866128,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004412989172741514,
      "loss": 5.9031,
      "mean_token_accuracy": 0.1840852752327919,
      "num_tokens": 53603395.0,
      "step": 21150
    },
    {
      "entropy": 6.123088312149048,
      "epoch": 2.4414310444316216,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044127111927215267,
      "loss": 5.8742,
      "mean_token_accuracy": 0.17809473276138305,
      "num_tokens": 53616134.0,
      "step": 21155
    },
    {
      "entropy": 6.17396650314331,
      "epoch": 2.44200807847663,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00044124331567770746,
      "loss": 5.8321,
      "mean_token_accuracy": 0.18286345899105072,
      "num_tokens": 53628736.0,
      "step": 21160
    },
    {
      "entropy": 6.247747707366943,
      "epoch": 2.442585112521639,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004412155064917509,
      "loss": 5.88,
      "mean_token_accuracy": 0.1826551675796509,
      "num_tokens": 53640921.0,
      "step": 21165
    },
    {
      "entropy": 6.150192737579346,
      "epoch": 2.4431621465666473,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044118769171521836,
      "loss": 5.8688,
      "mean_token_accuracy": 0.1746382847428322,
      "num_tokens": 53653483.0,
      "step": 21170
    },
    {
      "entropy": 6.260478067398071,
      "epoch": 2.443739180611656,
      "grad_norm": 3.859375,
      "learning_rate": 0.00044115987134904543,
      "loss": 5.8602,
      "mean_token_accuracy": 0.18295249938964844,
      "num_tokens": 53666175.0,
      "step": 21175
    },
    {
      "entropy": 6.1524045944213865,
      "epoch": 2.4443162146566646,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044113204539416776,
      "loss": 5.8163,
      "mean_token_accuracy": 0.1850590541958809,
      "num_tokens": 53678509.0,
      "step": 21180
    },
    {
      "entropy": 6.195194101333618,
      "epoch": 2.4448932487016735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004411042138515212,
      "loss": 5.8969,
      "mean_token_accuracy": 0.18193794637918473,
      "num_tokens": 53690825.0,
      "step": 21185
    },
    {
      "entropy": 6.138634300231933,
      "epoch": 2.445470282746682,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004410763767220419,
      "loss": 5.8419,
      "mean_token_accuracy": 0.1866929441690445,
      "num_tokens": 53703730.0,
      "step": 21190
    },
    {
      "entropy": 6.23060359954834,
      "epoch": 2.4460473167916907,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004410485340066662,
      "loss": 5.9313,
      "mean_token_accuracy": 0.17497013807296752,
      "num_tokens": 53715467.0,
      "step": 21195
    },
    {
      "entropy": 6.259818363189697,
      "epoch": 2.446624350836699,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004410206857063305,
      "loss": 5.8612,
      "mean_token_accuracy": 0.17825368344783782,
      "num_tokens": 53728559.0,
      "step": 21200
    },
    {
      "entropy": 6.22233829498291,
      "epoch": 2.447201384881708,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004409928318219715,
      "loss": 5.9014,
      "mean_token_accuracy": 0.17439157664775848,
      "num_tokens": 53740862.0,
      "step": 21205
    },
    {
      "entropy": 6.141938924789429,
      "epoch": 2.447778418926717,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004409649723545262,
      "loss": 5.8574,
      "mean_token_accuracy": 0.17874789983034134,
      "num_tokens": 53753026.0,
      "step": 21210
    },
    {
      "entropy": 6.2839446544647215,
      "epoch": 2.4483554529717253,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004409371073049313,
      "loss": 5.9209,
      "mean_token_accuracy": 0.174986369907856,
      "num_tokens": 53765318.0,
      "step": 21215
    },
    {
      "entropy": 6.210834074020386,
      "epoch": 2.448932487016734,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004409092366741243,
      "loss": 5.8834,
      "mean_token_accuracy": 0.1839582920074463,
      "num_tokens": 53777199.0,
      "step": 21220
    },
    {
      "entropy": 6.107977199554443,
      "epoch": 2.4495095210617426,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004408813604630425,
      "loss": 5.8635,
      "mean_token_accuracy": 0.1781278893351555,
      "num_tokens": 53790229.0,
      "step": 21225
    },
    {
      "entropy": 6.149947452545166,
      "epoch": 2.4500865551067514,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004408534786726236,
      "loss": 5.8052,
      "mean_token_accuracy": 0.1819358617067337,
      "num_tokens": 53803194.0,
      "step": 21230
    },
    {
      "entropy": 6.183941984176636,
      "epoch": 2.45066358915176,
      "grad_norm": 1.125,
      "learning_rate": 0.0004408255913038053,
      "loss": 5.8736,
      "mean_token_accuracy": 0.1778736636042595,
      "num_tokens": 53815710.0,
      "step": 21235
    },
    {
      "entropy": 6.265126419067383,
      "epoch": 2.4512406231967687,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004407976983575256,
      "loss": 6.0046,
      "mean_token_accuracy": 0.17137757688760757,
      "num_tokens": 53829427.0,
      "step": 21240
    },
    {
      "entropy": 6.212850141525268,
      "epoch": 2.451817657241777,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004407697998347227,
      "loss": 5.8438,
      "mean_token_accuracy": 0.17882610708475113,
      "num_tokens": 53842515.0,
      "step": 21245
    },
    {
      "entropy": 6.20806736946106,
      "epoch": 2.452394691286786,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004407418957363349,
      "loss": 5.9061,
      "mean_token_accuracy": 0.17825556695461273,
      "num_tokens": 53855913.0,
      "step": 21250
    },
    {
      "entropy": 6.199302816390992,
      "epoch": 2.4529717253317944,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044071398606330075,
      "loss": 5.8625,
      "mean_token_accuracy": 0.18881430327892304,
      "num_tokens": 53869354.0,
      "step": 21255
    },
    {
      "entropy": 6.149003553390503,
      "epoch": 2.4535487593768033,
      "grad_norm": 1.015625,
      "learning_rate": 0.000440686070816559,
      "loss": 5.8544,
      "mean_token_accuracy": 0.18218761384487153,
      "num_tokens": 53881149.0,
      "step": 21260
    },
    {
      "entropy": 6.206202411651612,
      "epoch": 2.4541257934218117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004406581499970486,
      "loss": 5.8998,
      "mean_token_accuracy": 0.18592076152563095,
      "num_tokens": 53893930.0,
      "step": 21265
    },
    {
      "entropy": 6.190030717849732,
      "epoch": 2.4547028274668206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004406302236057086,
      "loss": 5.8637,
      "mean_token_accuracy": 0.18284722715616225,
      "num_tokens": 53905306.0,
      "step": 21270
    },
    {
      "entropy": 6.270929956436158,
      "epoch": 2.455279861511829,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044060229164347826,
      "loss": 5.9625,
      "mean_token_accuracy": 0.1789903849363327,
      "num_tokens": 53918405.0,
      "step": 21275
    },
    {
      "entropy": 6.164856338500977,
      "epoch": 2.455856895556838,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044057435411129714,
      "loss": 5.8212,
      "mean_token_accuracy": 0.18167843520641327,
      "num_tokens": 53932931.0,
      "step": 21280
    },
    {
      "entropy": 6.2761815071105955,
      "epoch": 2.4564339296018467,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044054641101010485,
      "loss": 5.9392,
      "mean_token_accuracy": 0.17176334857940673,
      "num_tokens": 53945438.0,
      "step": 21285
    },
    {
      "entropy": 6.227966785430908,
      "epoch": 2.457010963646855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044051846234084127,
      "loss": 5.9342,
      "mean_token_accuracy": 0.17109089940786362,
      "num_tokens": 53958552.0,
      "step": 21290
    },
    {
      "entropy": 6.188142347335815,
      "epoch": 2.457587997691864,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004404905081044464,
      "loss": 5.8369,
      "mean_token_accuracy": 0.18699333667755128,
      "num_tokens": 53970776.0,
      "step": 21295
    },
    {
      "entropy": 6.17668604850769,
      "epoch": 2.4581650317368724,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004404625483018605,
      "loss": 5.9755,
      "mean_token_accuracy": 0.18093385994434358,
      "num_tokens": 53982667.0,
      "step": 21300
    },
    {
      "entropy": 6.251619434356689,
      "epoch": 2.4587420657818813,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000440434582934024,
      "loss": 5.8748,
      "mean_token_accuracy": 0.18301041275262833,
      "num_tokens": 53994603.0,
      "step": 21305
    },
    {
      "entropy": 6.218372440338134,
      "epoch": 2.4593190998268897,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044040661200187746,
      "loss": 5.9255,
      "mean_token_accuracy": 0.17349501103162765,
      "num_tokens": 54007499.0,
      "step": 21310
    },
    {
      "entropy": 6.168102264404297,
      "epoch": 2.4598961338718985,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044037863550636173,
      "loss": 5.8876,
      "mean_token_accuracy": 0.1771981120109558,
      "num_tokens": 54019471.0,
      "step": 21315
    },
    {
      "entropy": 6.195329380035401,
      "epoch": 2.460473167916907,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044035065344841766,
      "loss": 5.856,
      "mean_token_accuracy": 0.18363291770219803,
      "num_tokens": 54032352.0,
      "step": 21320
    },
    {
      "entropy": 6.205634355545044,
      "epoch": 2.461050201961916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044032266582898655,
      "loss": 5.908,
      "mean_token_accuracy": 0.18235446214675904,
      "num_tokens": 54046284.0,
      "step": 21325
    },
    {
      "entropy": 6.1461199760437015,
      "epoch": 2.4616272360069242,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004402946726490097,
      "loss": 5.8277,
      "mean_token_accuracy": 0.18235756158828736,
      "num_tokens": 54058867.0,
      "step": 21330
    },
    {
      "entropy": 6.152193260192871,
      "epoch": 2.462204270051933,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044026667390942867,
      "loss": 5.791,
      "mean_token_accuracy": 0.18168945908546447,
      "num_tokens": 54070587.0,
      "step": 21335
    },
    {
      "entropy": 6.104036474227906,
      "epoch": 2.462781304096942,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004402386696111851,
      "loss": 5.7838,
      "mean_token_accuracy": 0.18518714904785155,
      "num_tokens": 54083707.0,
      "step": 21340
    },
    {
      "entropy": 6.217674112319946,
      "epoch": 2.4633583381419504,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044021065975522096,
      "loss": 5.8232,
      "mean_token_accuracy": 0.1845399856567383,
      "num_tokens": 54095166.0,
      "step": 21345
    },
    {
      "entropy": 6.189757442474365,
      "epoch": 2.463935372186959,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004401826443424784,
      "loss": 5.9071,
      "mean_token_accuracy": 0.17897473275661469,
      "num_tokens": 54108329.0,
      "step": 21350
    },
    {
      "entropy": 6.2229245662689205,
      "epoch": 2.4645124062319677,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00044015462337389954,
      "loss": 5.9053,
      "mean_token_accuracy": 0.18149892538785933,
      "num_tokens": 54120799.0,
      "step": 21355
    },
    {
      "entropy": 6.24553804397583,
      "epoch": 2.4650894402769765,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044012659685042704,
      "loss": 5.9401,
      "mean_token_accuracy": 0.17447718232870102,
      "num_tokens": 54133346.0,
      "step": 21360
    },
    {
      "entropy": 6.2007505893707275,
      "epoch": 2.465666474321985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004400985647730034,
      "loss": 5.8715,
      "mean_token_accuracy": 0.18146896362304688,
      "num_tokens": 54146633.0,
      "step": 21365
    },
    {
      "entropy": 6.175316715240479,
      "epoch": 2.466243508366994,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044007052714257165,
      "loss": 5.7918,
      "mean_token_accuracy": 0.18526008278131484,
      "num_tokens": 54158708.0,
      "step": 21370
    },
    {
      "entropy": 6.082137632369995,
      "epoch": 2.466820542412002,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004400424839600747,
      "loss": 5.7923,
      "mean_token_accuracy": 0.18420593589544296,
      "num_tokens": 54171707.0,
      "step": 21375
    },
    {
      "entropy": 6.174465322494507,
      "epoch": 2.467397576457011,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00044001443522645575,
      "loss": 5.8629,
      "mean_token_accuracy": 0.17182135730981826,
      "num_tokens": 54184778.0,
      "step": 21380
    },
    {
      "entropy": 6.181794452667236,
      "epoch": 2.4679746105020195,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004399863809426582,
      "loss": 5.888,
      "mean_token_accuracy": 0.17853653877973558,
      "num_tokens": 54197199.0,
      "step": 21385
    },
    {
      "entropy": 6.217908239364624,
      "epoch": 2.4685516445470284,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043995832110962576,
      "loss": 5.8481,
      "mean_token_accuracy": 0.18075670748949052,
      "num_tokens": 54211277.0,
      "step": 21390
    },
    {
      "entropy": 6.2485222816467285,
      "epoch": 2.4691286785920368,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043993025572830203,
      "loss": 5.8977,
      "mean_token_accuracy": 0.17664981335401536,
      "num_tokens": 54224342.0,
      "step": 21395
    },
    {
      "entropy": 6.212941360473633,
      "epoch": 2.4697057126370456,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043990218479963115,
      "loss": 5.8589,
      "mean_token_accuracy": 0.18220894932746887,
      "num_tokens": 54236115.0,
      "step": 21400
    },
    {
      "entropy": 6.113332319259643,
      "epoch": 2.470282746682054,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043987410832455715,
      "loss": 5.7566,
      "mean_token_accuracy": 0.18675069510936737,
      "num_tokens": 54248665.0,
      "step": 21405
    },
    {
      "entropy": 6.130006456375122,
      "epoch": 2.470859780727063,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043984602630402447,
      "loss": 5.8323,
      "mean_token_accuracy": 0.18561057597398758,
      "num_tokens": 54260280.0,
      "step": 21410
    },
    {
      "entropy": 6.234626388549804,
      "epoch": 2.4714368147720718,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043981793873897756,
      "loss": 5.8595,
      "mean_token_accuracy": 0.18214363157749175,
      "num_tokens": 54271057.0,
      "step": 21415
    },
    {
      "entropy": 6.160151624679566,
      "epoch": 2.47201384881708,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004397898456303612,
      "loss": 5.7658,
      "mean_token_accuracy": 0.18412299305200577,
      "num_tokens": 54284568.0,
      "step": 21420
    },
    {
      "entropy": 6.041882467269898,
      "epoch": 2.472590882862089,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043976174697912015,
      "loss": 5.7038,
      "mean_token_accuracy": 0.19323309808969497,
      "num_tokens": 54297722.0,
      "step": 21425
    },
    {
      "entropy": 6.214208793640137,
      "epoch": 2.4731679169070975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397336427861996,
      "loss": 5.9186,
      "mean_token_accuracy": 0.17974215894937515,
      "num_tokens": 54310992.0,
      "step": 21430
    },
    {
      "entropy": 6.329280424118042,
      "epoch": 2.4737449509521063,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004397055330525448,
      "loss": 5.9617,
      "mean_token_accuracy": 0.17033447325229645,
      "num_tokens": 54323540.0,
      "step": 21435
    },
    {
      "entropy": 6.164299154281617,
      "epoch": 2.4743219849971148,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004396774177791012,
      "loss": 5.8877,
      "mean_token_accuracy": 0.17505358159542084,
      "num_tokens": 54335522.0,
      "step": 21440
    },
    {
      "entropy": 6.195537281036377,
      "epoch": 2.4748990190421236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043964929696681444,
      "loss": 5.8605,
      "mean_token_accuracy": 0.1807871699333191,
      "num_tokens": 54347216.0,
      "step": 21445
    },
    {
      "entropy": 6.191735124588012,
      "epoch": 2.475476053087132,
      "grad_norm": 1.5,
      "learning_rate": 0.0004396211706166305,
      "loss": 5.8763,
      "mean_token_accuracy": 0.17860430032014846,
      "num_tokens": 54360261.0,
      "step": 21450
    },
    {
      "entropy": 6.142237758636474,
      "epoch": 2.476053087132141,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004395930387294951,
      "loss": 5.8097,
      "mean_token_accuracy": 0.18718330413103104,
      "num_tokens": 54373279.0,
      "step": 21455
    },
    {
      "entropy": 6.225560808181763,
      "epoch": 2.4766301211771493,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004395649013063546,
      "loss": 5.9007,
      "mean_token_accuracy": 0.17708432525396348,
      "num_tokens": 54386266.0,
      "step": 21460
    },
    {
      "entropy": 6.21087794303894,
      "epoch": 2.477207155222158,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004395367583481554,
      "loss": 5.8721,
      "mean_token_accuracy": 0.182944592833519,
      "num_tokens": 54398839.0,
      "step": 21465
    },
    {
      "entropy": 6.12814507484436,
      "epoch": 2.4777841892671666,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043950860985584423,
      "loss": 5.7662,
      "mean_token_accuracy": 0.18537706285715103,
      "num_tokens": 54409879.0,
      "step": 21470
    },
    {
      "entropy": 6.1023633003234865,
      "epoch": 2.4783612233121755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004394804558303675,
      "loss": 5.8117,
      "mean_token_accuracy": 0.17815116792917252,
      "num_tokens": 54421641.0,
      "step": 21475
    },
    {
      "entropy": 6.273129940032959,
      "epoch": 2.478938257357184,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004394522962726724,
      "loss": 5.9005,
      "mean_token_accuracy": 0.17826270163059235,
      "num_tokens": 54434169.0,
      "step": 21480
    },
    {
      "entropy": 6.222082042694092,
      "epoch": 2.4795152914021927,
      "grad_norm": 1.0,
      "learning_rate": 0.00043942413118370604,
      "loss": 5.8673,
      "mean_token_accuracy": 0.17851683497428894,
      "num_tokens": 54447814.0,
      "step": 21485
    },
    {
      "entropy": 6.1406800746917725,
      "epoch": 2.4800923254472016,
      "grad_norm": 2.203125,
      "learning_rate": 0.0004393959605644157,
      "loss": 5.9221,
      "mean_token_accuracy": 0.17649478018283843,
      "num_tokens": 54460589.0,
      "step": 21490
    },
    {
      "entropy": 6.239633512496948,
      "epoch": 2.48066935949221,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004393677844157488,
      "loss": 5.8195,
      "mean_token_accuracy": 0.18858450055122375,
      "num_tokens": 54473328.0,
      "step": 21495
    },
    {
      "entropy": 6.185845947265625,
      "epoch": 2.481246393537219,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004393396027386532,
      "loss": 5.8978,
      "mean_token_accuracy": 0.18077711910009384,
      "num_tokens": 54486763.0,
      "step": 21500
    },
    {
      "entropy": 6.112710046768188,
      "epoch": 2.4818234275822273,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043931141553407667,
      "loss": 5.8241,
      "mean_token_accuracy": 0.1887489527463913,
      "num_tokens": 54499081.0,
      "step": 21505
    },
    {
      "entropy": 6.161634206771851,
      "epoch": 2.482400461627236,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004392832228029672,
      "loss": 5.7967,
      "mean_token_accuracy": 0.1848980739712715,
      "num_tokens": 54511776.0,
      "step": 21510
    },
    {
      "entropy": 6.168097496032715,
      "epoch": 2.4829774956722446,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043925502454627327,
      "loss": 5.835,
      "mean_token_accuracy": 0.18515201210975646,
      "num_tokens": 54524784.0,
      "step": 21515
    },
    {
      "entropy": 6.2054120063781735,
      "epoch": 2.4835545297172534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004392268207649431,
      "loss": 5.8878,
      "mean_token_accuracy": 0.177184621989727,
      "num_tokens": 54537028.0,
      "step": 21520
    },
    {
      "entropy": 6.310422897338867,
      "epoch": 2.484131563762262,
      "grad_norm": 1.8125,
      "learning_rate": 0.00043919861145992546,
      "loss": 6.0682,
      "mean_token_accuracy": 0.16663777828216553,
      "num_tokens": 54550889.0,
      "step": 21525
    },
    {
      "entropy": 6.25194239616394,
      "epoch": 2.4847085978072707,
      "grad_norm": 1.203125,
      "learning_rate": 0.000439170396632169,
      "loss": 5.9824,
      "mean_token_accuracy": 0.17062321603298186,
      "num_tokens": 54565218.0,
      "step": 21530
    },
    {
      "entropy": 6.218824434280395,
      "epoch": 2.485285631852279,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004391421762826229,
      "loss": 5.8584,
      "mean_token_accuracy": 0.177971813082695,
      "num_tokens": 54577622.0,
      "step": 21535
    },
    {
      "entropy": 6.178410625457763,
      "epoch": 2.485862665897288,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004391139504122362,
      "loss": 5.8539,
      "mean_token_accuracy": 0.1888299837708473,
      "num_tokens": 54590213.0,
      "step": 21540
    },
    {
      "entropy": 6.145999145507813,
      "epoch": 2.4864396999422964,
      "grad_norm": 1.34375,
      "learning_rate": 0.00043908571902195827,
      "loss": 5.8629,
      "mean_token_accuracy": 0.18371896594762802,
      "num_tokens": 54603615.0,
      "step": 21545
    },
    {
      "entropy": 6.193444538116455,
      "epoch": 2.4870167339873053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043905748211273864,
      "loss": 5.8509,
      "mean_token_accuracy": 0.17773962914943695,
      "num_tokens": 54615924.0,
      "step": 21550
    },
    {
      "entropy": 6.248504543304444,
      "epoch": 2.4875937680323137,
      "grad_norm": 2.0,
      "learning_rate": 0.00043902923968552714,
      "loss": 5.8459,
      "mean_token_accuracy": 0.18016792088747025,
      "num_tokens": 54629184.0,
      "step": 21555
    },
    {
      "entropy": 6.171130847930908,
      "epoch": 2.4881708020773226,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004390009917412737,
      "loss": 5.9015,
      "mean_token_accuracy": 0.18237817287445068,
      "num_tokens": 54642669.0,
      "step": 21560
    },
    {
      "entropy": 6.189037704467774,
      "epoch": 2.4887478361223314,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043897273828092824,
      "loss": 5.8547,
      "mean_token_accuracy": 0.18548583984375,
      "num_tokens": 54655229.0,
      "step": 21565
    },
    {
      "entropy": 6.234196281433105,
      "epoch": 2.48932487016734,
      "grad_norm": 1.0,
      "learning_rate": 0.0004389444793054413,
      "loss": 5.9161,
      "mean_token_accuracy": 0.1780890628695488,
      "num_tokens": 54667516.0,
      "step": 21570
    },
    {
      "entropy": 6.133837509155273,
      "epoch": 2.4899019042123487,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00043891621481576324,
      "loss": 5.8095,
      "mean_token_accuracy": 0.18189657479524612,
      "num_tokens": 54679832.0,
      "step": 21575
    },
    {
      "entropy": 6.081688690185547,
      "epoch": 2.490478938257357,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004388879448128446,
      "loss": 5.7079,
      "mean_token_accuracy": 0.1902725487947464,
      "num_tokens": 54693943.0,
      "step": 21580
    },
    {
      "entropy": 6.176182842254638,
      "epoch": 2.491055972302366,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043885966929763635,
      "loss": 5.8102,
      "mean_token_accuracy": 0.18683638870716096,
      "num_tokens": 54706404.0,
      "step": 21585
    },
    {
      "entropy": 6.168729639053344,
      "epoch": 2.4916330063473744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043883138827108964,
      "loss": 5.8496,
      "mean_token_accuracy": 0.17861776649951935,
      "num_tokens": 54718323.0,
      "step": 21590
    },
    {
      "entropy": 6.206038904190064,
      "epoch": 2.4922100403923833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004388031017341555,
      "loss": 5.9334,
      "mean_token_accuracy": 0.1766161561012268,
      "num_tokens": 54730128.0,
      "step": 21595
    },
    {
      "entropy": 6.1698990821838375,
      "epoch": 2.4927870744373917,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043877480968778536,
      "loss": 5.8396,
      "mean_token_accuracy": 0.18942066431045532,
      "num_tokens": 54743437.0,
      "step": 21600
    },
    {
      "entropy": 6.120104265213013,
      "epoch": 2.4933641084824005,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004387465121329309,
      "loss": 5.9076,
      "mean_token_accuracy": 0.17566175311803817,
      "num_tokens": 54755573.0,
      "step": 21605
    },
    {
      "entropy": 6.2447713851928714,
      "epoch": 2.493941142527409,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00043871820907054375,
      "loss": 5.9155,
      "mean_token_accuracy": 0.17744822949171066,
      "num_tokens": 54768071.0,
      "step": 21610
    },
    {
      "entropy": 6.28808274269104,
      "epoch": 2.494518176572418,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043868990050157617,
      "loss": 5.9116,
      "mean_token_accuracy": 0.17353134751319885,
      "num_tokens": 54782245.0,
      "step": 21615
    },
    {
      "entropy": 6.108926153182983,
      "epoch": 2.4950952106174267,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004386615864269799,
      "loss": 5.7889,
      "mean_token_accuracy": 0.18992384821176528,
      "num_tokens": 54794335.0,
      "step": 21620
    },
    {
      "entropy": 6.127848196029663,
      "epoch": 2.495672244662435,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043863326684770746,
      "loss": 5.8774,
      "mean_token_accuracy": 0.18150349855422973,
      "num_tokens": 54807015.0,
      "step": 21625
    },
    {
      "entropy": 6.187694263458252,
      "epoch": 2.4962492787074435,
      "grad_norm": 1.0,
      "learning_rate": 0.0004386049417647115,
      "loss": 5.7918,
      "mean_token_accuracy": 0.17833972126245498,
      "num_tokens": 54821278.0,
      "step": 21630
    },
    {
      "entropy": 6.247255516052246,
      "epoch": 2.4968263127524524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004385766111789445,
      "loss": 5.9114,
      "mean_token_accuracy": 0.18119488060474395,
      "num_tokens": 54834905.0,
      "step": 21635
    },
    {
      "entropy": 6.1341376304626465,
      "epoch": 2.4974033467974612,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004385482750913595,
      "loss": 5.8336,
      "mean_token_accuracy": 0.18655134439468385,
      "num_tokens": 54847597.0,
      "step": 21640
    },
    {
      "entropy": 6.106540012359619,
      "epoch": 2.4979803808424696,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004385199335029094,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1906038284301758,
      "num_tokens": 54859747.0,
      "step": 21645
    },
    {
      "entropy": 6.140057563781738,
      "epoch": 2.4985574148874785,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043849158641454767,
      "loss": 5.8761,
      "mean_token_accuracy": 0.17953258007764816,
      "num_tokens": 54873025.0,
      "step": 21650
    },
    {
      "entropy": 6.187953186035156,
      "epoch": 2.499134448932487,
      "grad_norm": 1.40625,
      "learning_rate": 0.00043846323382722765,
      "loss": 5.8184,
      "mean_token_accuracy": 0.18455153256654738,
      "num_tokens": 54884718.0,
      "step": 21655
    },
    {
      "entropy": 6.1051928997039795,
      "epoch": 2.499711482977496,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043843487574190285,
      "loss": 5.7548,
      "mean_token_accuracy": 0.18575444221496581,
      "num_tokens": 54896557.0,
      "step": 21660
    },
    {
      "entropy": 6.281343841552735,
      "epoch": 2.500288517022504,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043840651215952726,
      "loss": 5.9865,
      "mean_token_accuracy": 0.17045235633850098,
      "num_tokens": 54909608.0,
      "step": 21665
    },
    {
      "entropy": 6.247969722747802,
      "epoch": 2.500865551067513,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043837814308105477,
      "loss": 5.8158,
      "mean_token_accuracy": 0.18350847959518432,
      "num_tokens": 54921794.0,
      "step": 21670
    },
    {
      "entropy": 6.228341293334961,
      "epoch": 2.5014425851125215,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043834976850743964,
      "loss": 5.9147,
      "mean_token_accuracy": 0.17788248509168625,
      "num_tokens": 54934527.0,
      "step": 21675
    },
    {
      "entropy": 6.1380973815917965,
      "epoch": 2.5020196191575304,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004383213884396361,
      "loss": 5.8231,
      "mean_token_accuracy": 0.18515776842832565,
      "num_tokens": 54947074.0,
      "step": 21680
    },
    {
      "entropy": 6.195892572402954,
      "epoch": 2.5025966532025388,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004382930028785989,
      "loss": 5.8595,
      "mean_token_accuracy": 0.17820197939872742,
      "num_tokens": 54958677.0,
      "step": 21685
    },
    {
      "entropy": 6.123137474060059,
      "epoch": 2.5031736872475476,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004382646118252826,
      "loss": 5.8025,
      "mean_token_accuracy": 0.1913789242506027,
      "num_tokens": 54972067.0,
      "step": 21690
    },
    {
      "entropy": 6.1770933151245115,
      "epoch": 2.5037507212925565,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043823621528064216,
      "loss": 5.8862,
      "mean_token_accuracy": 0.1763715460896492,
      "num_tokens": 54985010.0,
      "step": 21695
    },
    {
      "entropy": 6.2625343799591064,
      "epoch": 2.504327755337565,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043820781324563276,
      "loss": 5.9181,
      "mean_token_accuracy": 0.17570398151874542,
      "num_tokens": 54998731.0,
      "step": 21700
    },
    {
      "entropy": 6.143558740615845,
      "epoch": 2.5049047893825733,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043817940572120963,
      "loss": 5.7779,
      "mean_token_accuracy": 0.18513636738061906,
      "num_tokens": 55011288.0,
      "step": 21705
    },
    {
      "entropy": 6.144919776916504,
      "epoch": 2.505481823427582,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043815099270832814,
      "loss": 5.8486,
      "mean_token_accuracy": 0.18193407505750656,
      "num_tokens": 55023718.0,
      "step": 21710
    },
    {
      "entropy": 6.1648200988769535,
      "epoch": 2.506058857472591,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043812257420794415,
      "loss": 5.7996,
      "mean_token_accuracy": 0.18745406568050385,
      "num_tokens": 55036628.0,
      "step": 21715
    },
    {
      "entropy": 6.214187908172607,
      "epoch": 2.5066358915175995,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043809415022101335,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1827463150024414,
      "num_tokens": 55049765.0,
      "step": 21720
    },
    {
      "entropy": 6.14854588508606,
      "epoch": 2.5072129255626083,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004380657207484919,
      "loss": 5.8145,
      "mean_token_accuracy": 0.18578870445489884,
      "num_tokens": 55061852.0,
      "step": 21725
    },
    {
      "entropy": 6.269407463073731,
      "epoch": 2.5077899596076167,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043803728579133586,
      "loss": 6.0257,
      "mean_token_accuracy": 0.17544178813695907,
      "num_tokens": 55074560.0,
      "step": 21730
    },
    {
      "entropy": 6.204534530639648,
      "epoch": 2.5083669936526256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004380088453505017,
      "loss": 5.8251,
      "mean_token_accuracy": 0.18425490707159042,
      "num_tokens": 55086886.0,
      "step": 21735
    },
    {
      "entropy": 6.162876081466675,
      "epoch": 2.508944027697634,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043798039942694603,
      "loss": 5.8532,
      "mean_token_accuracy": 0.176558955013752,
      "num_tokens": 55100503.0,
      "step": 21740
    },
    {
      "entropy": 6.199307155609131,
      "epoch": 2.509521061742643,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043795194802162557,
      "loss": 5.7926,
      "mean_token_accuracy": 0.18185302764177322,
      "num_tokens": 55113212.0,
      "step": 21745
    },
    {
      "entropy": 6.3023333072662355,
      "epoch": 2.5100980957876513,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004379234911354973,
      "loss": 6.0092,
      "mean_token_accuracy": 0.16873298436403275,
      "num_tokens": 55127391.0,
      "step": 21750
    },
    {
      "entropy": 6.205758380889892,
      "epoch": 2.51067512983266,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043789502876951834,
      "loss": 5.8793,
      "mean_token_accuracy": 0.17774643003940582,
      "num_tokens": 55139990.0,
      "step": 21755
    },
    {
      "entropy": 6.143463611602783,
      "epoch": 2.5112521638776686,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004378665609246459,
      "loss": 5.7649,
      "mean_token_accuracy": 0.18519463241100312,
      "num_tokens": 55152472.0,
      "step": 21760
    },
    {
      "entropy": 6.227129077911377,
      "epoch": 2.5118291979226774,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00043783808760183764,
      "loss": 5.8776,
      "mean_token_accuracy": 0.18210879415273667,
      "num_tokens": 55166342.0,
      "step": 21765
    },
    {
      "entropy": 6.232834911346435,
      "epoch": 2.5124062319676863,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043780960880205125,
      "loss": 5.8924,
      "mean_token_accuracy": 0.17831755429506302,
      "num_tokens": 55177925.0,
      "step": 21770
    },
    {
      "entropy": 6.187525033950806,
      "epoch": 2.5129832660126947,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043778112452624445,
      "loss": 5.8784,
      "mean_token_accuracy": 0.17896425426006318,
      "num_tokens": 55189179.0,
      "step": 21775
    },
    {
      "entropy": 6.270234394073486,
      "epoch": 2.513560300057703,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043775263477537546,
      "loss": 5.8999,
      "mean_token_accuracy": 0.17543237060308456,
      "num_tokens": 55202244.0,
      "step": 21780
    },
    {
      "entropy": 6.174876689910889,
      "epoch": 2.514137334102712,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004377241395504024,
      "loss": 5.8461,
      "mean_token_accuracy": 0.17749062180519104,
      "num_tokens": 55214545.0,
      "step": 21785
    },
    {
      "entropy": 6.173722791671753,
      "epoch": 2.514714368147721,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004376956388522837,
      "loss": 5.7967,
      "mean_token_accuracy": 0.1798494577407837,
      "num_tokens": 55227766.0,
      "step": 21790
    },
    {
      "entropy": 6.2526191711425785,
      "epoch": 2.5152914021927293,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004376671326819781,
      "loss": 5.9369,
      "mean_token_accuracy": 0.1758306846022606,
      "num_tokens": 55239841.0,
      "step": 21795
    },
    {
      "entropy": 6.081368827819825,
      "epoch": 2.515868436237738,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043763862104044424,
      "loss": 5.7562,
      "mean_token_accuracy": 0.19104499667882918,
      "num_tokens": 55253440.0,
      "step": 21800
    },
    {
      "entropy": 6.237401914596558,
      "epoch": 2.5164454702827466,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043761010392864114,
      "loss": 5.8854,
      "mean_token_accuracy": 0.17310586273670198,
      "num_tokens": 55264698.0,
      "step": 21805
    },
    {
      "entropy": 6.2173669815063475,
      "epoch": 2.5170225043277554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043758158134752793,
      "loss": 5.8445,
      "mean_token_accuracy": 0.1770642429590225,
      "num_tokens": 55276923.0,
      "step": 21810
    },
    {
      "entropy": 6.206795644760132,
      "epoch": 2.517599538372764,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004375530532980642,
      "loss": 5.9123,
      "mean_token_accuracy": 0.1775376632809639,
      "num_tokens": 55289774.0,
      "step": 21815
    },
    {
      "entropy": 6.196282768249512,
      "epoch": 2.5181765724177727,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000437524519781209,
      "loss": 5.902,
      "mean_token_accuracy": 0.1796469047665596,
      "num_tokens": 55301877.0,
      "step": 21820
    },
    {
      "entropy": 6.160347890853882,
      "epoch": 2.5187536064627816,
      "grad_norm": 1.0,
      "learning_rate": 0.0004374959807979224,
      "loss": 5.859,
      "mean_token_accuracy": 0.18730694353580474,
      "num_tokens": 55314680.0,
      "step": 21825
    },
    {
      "entropy": 6.22357668876648,
      "epoch": 2.51933064050779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004374674363491642,
      "loss": 5.8695,
      "mean_token_accuracy": 0.17928709387779235,
      "num_tokens": 55326874.0,
      "step": 21830
    },
    {
      "entropy": 6.177434206008911,
      "epoch": 2.5199076745527984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043743888643589453,
      "loss": 5.8762,
      "mean_token_accuracy": 0.17795808613300323,
      "num_tokens": 55339281.0,
      "step": 21835
    },
    {
      "entropy": 6.172960376739502,
      "epoch": 2.5204847085978073,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043741033105907355,
      "loss": 5.86,
      "mean_token_accuracy": 0.18426252603530885,
      "num_tokens": 55351187.0,
      "step": 21840
    },
    {
      "entropy": 6.225103950500488,
      "epoch": 2.521061742642816,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004373817702196618,
      "loss": 5.8494,
      "mean_token_accuracy": 0.18199699521064758,
      "num_tokens": 55365122.0,
      "step": 21845
    },
    {
      "entropy": 6.123576259613037,
      "epoch": 2.5216387766878245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043735320391861986,
      "loss": 5.8673,
      "mean_token_accuracy": 0.17788007110357285,
      "num_tokens": 55376494.0,
      "step": 21850
    },
    {
      "entropy": 6.273983430862427,
      "epoch": 2.522215810732833,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004373246321569085,
      "loss": 5.903,
      "mean_token_accuracy": 0.18144610077142714,
      "num_tokens": 55388037.0,
      "step": 21855
    },
    {
      "entropy": 6.140970182418823,
      "epoch": 2.522792844777842,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004372960549354888,
      "loss": 5.813,
      "mean_token_accuracy": 0.1791406214237213,
      "num_tokens": 55401356.0,
      "step": 21860
    },
    {
      "entropy": 5.9759265899658205,
      "epoch": 2.5233698788228507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043726747225532195,
      "loss": 5.6977,
      "mean_token_accuracy": 0.19044179767370223,
      "num_tokens": 55414332.0,
      "step": 21865
    },
    {
      "entropy": 6.219502973556518,
      "epoch": 2.523946912867859,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004372388841173692,
      "loss": 5.8609,
      "mean_token_accuracy": 0.18201821446418762,
      "num_tokens": 55426886.0,
      "step": 21870
    },
    {
      "entropy": 6.246579217910766,
      "epoch": 2.524523946912868,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004372102905225922,
      "loss": 5.8581,
      "mean_token_accuracy": 0.18200962692499162,
      "num_tokens": 55439688.0,
      "step": 21875
    },
    {
      "entropy": 6.119673347473144,
      "epoch": 2.5251009809578764,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043718169147195275,
      "loss": 5.8156,
      "mean_token_accuracy": 0.18275847136974335,
      "num_tokens": 55451597.0,
      "step": 21880
    },
    {
      "entropy": 6.165056610107422,
      "epoch": 2.5256780150028852,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043715308696641255,
      "loss": 5.8293,
      "mean_token_accuracy": 0.17857772707939149,
      "num_tokens": 55463627.0,
      "step": 21885
    },
    {
      "entropy": 6.273014211654663,
      "epoch": 2.5262550490478937,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004371244770069338,
      "loss": 5.9275,
      "mean_token_accuracy": 0.17339976280927658,
      "num_tokens": 55477071.0,
      "step": 21890
    },
    {
      "entropy": 6.150373029708862,
      "epoch": 2.5268320830929025,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043709586159447884,
      "loss": 5.8142,
      "mean_token_accuracy": 0.1759146958589554,
      "num_tokens": 55489338.0,
      "step": 21895
    },
    {
      "entropy": 6.1999053955078125,
      "epoch": 2.5274091171379114,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043706724073001017,
      "loss": 5.875,
      "mean_token_accuracy": 0.1795965000987053,
      "num_tokens": 55503770.0,
      "step": 21900
    },
    {
      "entropy": 6.170782470703125,
      "epoch": 2.52798615118292,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004370386144144902,
      "loss": 5.8669,
      "mean_token_accuracy": 0.1774355113506317,
      "num_tokens": 55516832.0,
      "step": 21905
    },
    {
      "entropy": 6.276739454269409,
      "epoch": 2.5285631852279282,
      "grad_norm": 1.125,
      "learning_rate": 0.0004370099826488821,
      "loss": 5.8784,
      "mean_token_accuracy": 0.1805236369371414,
      "num_tokens": 55530280.0,
      "step": 21910
    },
    {
      "entropy": 6.0720977783203125,
      "epoch": 2.529140219272937,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004369813454341486,
      "loss": 5.8016,
      "mean_token_accuracy": 0.18719504177570342,
      "num_tokens": 55542688.0,
      "step": 21915
    },
    {
      "entropy": 6.099076890945435,
      "epoch": 2.529717253317946,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043695270277125303,
      "loss": 5.818,
      "mean_token_accuracy": 0.18461127281188966,
      "num_tokens": 55555921.0,
      "step": 21920
    },
    {
      "entropy": 6.199017477035523,
      "epoch": 2.5302942873629544,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004369240546611587,
      "loss": 5.8786,
      "mean_token_accuracy": 0.17888905256986617,
      "num_tokens": 55569667.0,
      "step": 21925
    },
    {
      "entropy": 6.230369281768799,
      "epoch": 2.5308713214079632,
      "grad_norm": 1.0,
      "learning_rate": 0.00043689540110482933,
      "loss": 5.8684,
      "mean_token_accuracy": 0.1831066742539406,
      "num_tokens": 55582170.0,
      "step": 21930
    },
    {
      "entropy": 6.195162010192871,
      "epoch": 2.5314483554529716,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004368667421032285,
      "loss": 5.906,
      "mean_token_accuracy": 0.17713675647974014,
      "num_tokens": 55595170.0,
      "step": 21935
    },
    {
      "entropy": 6.225862598419189,
      "epoch": 2.5320253894979805,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043683807765732013,
      "loss": 5.8582,
      "mean_token_accuracy": 0.18258154690265654,
      "num_tokens": 55606794.0,
      "step": 21940
    },
    {
      "entropy": 6.181500577926636,
      "epoch": 2.532602423542989,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043680940776806847,
      "loss": 5.8281,
      "mean_token_accuracy": 0.18431926667690277,
      "num_tokens": 55619423.0,
      "step": 21945
    },
    {
      "entropy": 6.199829006195069,
      "epoch": 2.533179457587998,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004367807324364378,
      "loss": 5.8363,
      "mean_token_accuracy": 0.18326984196901322,
      "num_tokens": 55630934.0,
      "step": 21950
    },
    {
      "entropy": 6.1872265338897705,
      "epoch": 2.533756491633006,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043675205166339247,
      "loss": 5.8933,
      "mean_token_accuracy": 0.17747756987810134,
      "num_tokens": 55643781.0,
      "step": 21955
    },
    {
      "entropy": 6.236071825027466,
      "epoch": 2.534333525678015,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004367233654498973,
      "loss": 5.9842,
      "mean_token_accuracy": 0.17563940733671188,
      "num_tokens": 55655564.0,
      "step": 21960
    },
    {
      "entropy": 6.198457717895508,
      "epoch": 2.5349105597230235,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004366946737969171,
      "loss": 5.8805,
      "mean_token_accuracy": 0.17978014647960663,
      "num_tokens": 55667408.0,
      "step": 21965
    },
    {
      "entropy": 6.210017681121826,
      "epoch": 2.5354875937680323,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004366659767054168,
      "loss": 5.7694,
      "mean_token_accuracy": 0.18559402078390122,
      "num_tokens": 55680876.0,
      "step": 21970
    },
    {
      "entropy": 6.153652334213257,
      "epoch": 2.536064627813041,
      "grad_norm": 1.359375,
      "learning_rate": 0.00043663727417636166,
      "loss": 5.8364,
      "mean_token_accuracy": 0.17415307611227035,
      "num_tokens": 55694639.0,
      "step": 21975
    },
    {
      "entropy": 6.135266590118408,
      "epoch": 2.5366416618580496,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004366085662107171,
      "loss": 5.8601,
      "mean_token_accuracy": 0.18313330560922622,
      "num_tokens": 55707221.0,
      "step": 21980
    },
    {
      "entropy": 6.238986444473267,
      "epoch": 2.537218695903058,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043657985280944875,
      "loss": 5.8111,
      "mean_token_accuracy": 0.1860315188765526,
      "num_tokens": 55720750.0,
      "step": 21985
    },
    {
      "entropy": 6.194635057449341,
      "epoch": 2.537795729948067,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004365511339735223,
      "loss": 5.8865,
      "mean_token_accuracy": 0.1789993092417717,
      "num_tokens": 55732550.0,
      "step": 21990
    },
    {
      "entropy": 6.16865177154541,
      "epoch": 2.5383727639930758,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004365224097039036,
      "loss": 5.8488,
      "mean_token_accuracy": 0.18228582441806793,
      "num_tokens": 55745468.0,
      "step": 21995
    },
    {
      "entropy": 6.109445858001709,
      "epoch": 2.538949798038084,
      "grad_norm": 1.03125,
      "learning_rate": 0.000436493680001559,
      "loss": 5.8489,
      "mean_token_accuracy": 0.18128757923841476,
      "num_tokens": 55757193.0,
      "step": 22000
    },
    {
      "entropy": 6.194368457794189,
      "epoch": 2.539526832083093,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043646494486745464,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18295118808746338,
      "num_tokens": 55770185.0,
      "step": 22005
    },
    {
      "entropy": 6.221388339996338,
      "epoch": 2.5401038661281015,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004364362043025571,
      "loss": 5.8603,
      "mean_token_accuracy": 0.17937735170125962,
      "num_tokens": 55782204.0,
      "step": 22010
    },
    {
      "entropy": 6.110798168182373,
      "epoch": 2.5406809001731103,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043640745830783297,
      "loss": 5.7682,
      "mean_token_accuracy": 0.1868910536170006,
      "num_tokens": 55795077.0,
      "step": 22015
    },
    {
      "entropy": 6.232817268371582,
      "epoch": 2.5412579342181187,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004363787068842491,
      "loss": 5.8717,
      "mean_token_accuracy": 0.18311762511730195,
      "num_tokens": 55808876.0,
      "step": 22020
    },
    {
      "entropy": 6.23801622390747,
      "epoch": 2.5418349682631276,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004363499500327727,
      "loss": 5.9718,
      "mean_token_accuracy": 0.17709683030843734,
      "num_tokens": 55822861.0,
      "step": 22025
    },
    {
      "entropy": 6.223509120941162,
      "epoch": 2.542412002308136,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043632118775437085,
      "loss": 5.8767,
      "mean_token_accuracy": 0.18123848587274552,
      "num_tokens": 55835523.0,
      "step": 22030
    },
    {
      "entropy": 6.218843221664429,
      "epoch": 2.542989036353145,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043629242005001096,
      "loss": 5.9086,
      "mean_token_accuracy": 0.17581363320350646,
      "num_tokens": 55848377.0,
      "step": 22035
    },
    {
      "entropy": 6.2499762058258055,
      "epoch": 2.5435660703981533,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004362636469206607,
      "loss": 5.9474,
      "mean_token_accuracy": 0.1790979251265526,
      "num_tokens": 55862074.0,
      "step": 22040
    },
    {
      "entropy": 6.1936921119689945,
      "epoch": 2.544143104443162,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004362348683672877,
      "loss": 5.8437,
      "mean_token_accuracy": 0.17577889263629914,
      "num_tokens": 55874816.0,
      "step": 22045
    },
    {
      "entropy": 6.256429624557495,
      "epoch": 2.544720138488171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004362060843908601,
      "loss": 5.9163,
      "mean_token_accuracy": 0.17813948839902877,
      "num_tokens": 55887424.0,
      "step": 22050
    },
    {
      "entropy": 6.168146419525146,
      "epoch": 2.5452971725331794,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004361772949923458,
      "loss": 5.8821,
      "mean_token_accuracy": 0.18185821771621705,
      "num_tokens": 55900532.0,
      "step": 22055
    },
    {
      "entropy": 6.167166423797608,
      "epoch": 2.545874206578188,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004361485001727132,
      "loss": 5.8475,
      "mean_token_accuracy": 0.18121427297592163,
      "num_tokens": 55912599.0,
      "step": 22060
    },
    {
      "entropy": 6.209183263778686,
      "epoch": 2.5464512406231967,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000436119699932931,
      "loss": 5.8887,
      "mean_token_accuracy": 0.1795339912176132,
      "num_tokens": 55925012.0,
      "step": 22065
    },
    {
      "entropy": 6.117090129852295,
      "epoch": 2.5470282746682056,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043609089427396763,
      "loss": 5.8296,
      "mean_token_accuracy": 0.18406159281730652,
      "num_tokens": 55938058.0,
      "step": 22070
    },
    {
      "entropy": 6.221741676330566,
      "epoch": 2.547605308713214,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043606208319679215,
      "loss": 5.9764,
      "mean_token_accuracy": 0.17183651328086852,
      "num_tokens": 55951439.0,
      "step": 22075
    },
    {
      "entropy": 6.1748284816741945,
      "epoch": 2.548182342758223,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004360332667023734,
      "loss": 5.8049,
      "mean_token_accuracy": 0.18733201473951339,
      "num_tokens": 55964583.0,
      "step": 22080
    },
    {
      "entropy": 6.099901103973389,
      "epoch": 2.5487593768032313,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043600444479168084,
      "loss": 5.7426,
      "mean_token_accuracy": 0.194771471619606,
      "num_tokens": 55978072.0,
      "step": 22085
    },
    {
      "entropy": 6.18526668548584,
      "epoch": 2.54933641084824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004359756174656836,
      "loss": 5.8195,
      "mean_token_accuracy": 0.18424388766288757,
      "num_tokens": 55990278.0,
      "step": 22090
    },
    {
      "entropy": 6.171966314315796,
      "epoch": 2.5499134448932486,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043594678472535153,
      "loss": 5.8932,
      "mean_token_accuracy": 0.17368784546852112,
      "num_tokens": 56003303.0,
      "step": 22095
    },
    {
      "entropy": 6.223716783523559,
      "epoch": 2.5504904789382574,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043591794657165425,
      "loss": 5.8501,
      "mean_token_accuracy": 0.1813538357615471,
      "num_tokens": 56015131.0,
      "step": 22100
    },
    {
      "entropy": 6.114982843399048,
      "epoch": 2.5510675129832663,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004358891030055617,
      "loss": 5.8066,
      "mean_token_accuracy": 0.1844347357749939,
      "num_tokens": 56027804.0,
      "step": 22105
    },
    {
      "entropy": 6.237184619903564,
      "epoch": 2.5516445470282747,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004358602540280443,
      "loss": 5.9629,
      "mean_token_accuracy": 0.17398676723241807,
      "num_tokens": 56040485.0,
      "step": 22110
    },
    {
      "entropy": 6.236841106414795,
      "epoch": 2.552221581073283,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043583139964007203,
      "loss": 5.86,
      "mean_token_accuracy": 0.1779933124780655,
      "num_tokens": 56052091.0,
      "step": 22115
    },
    {
      "entropy": 6.146714210510254,
      "epoch": 2.552798615118292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043580253984261554,
      "loss": 5.7983,
      "mean_token_accuracy": 0.18871628493070602,
      "num_tokens": 56063291.0,
      "step": 22120
    },
    {
      "entropy": 6.126740550994873,
      "epoch": 2.553375649163301,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043577367463664546,
      "loss": 5.8542,
      "mean_token_accuracy": 0.17438943684101105,
      "num_tokens": 56074219.0,
      "step": 22125
    },
    {
      "entropy": 6.120125579833984,
      "epoch": 2.5539526832083093,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004357448040231327,
      "loss": 5.8161,
      "mean_token_accuracy": 0.17922122031450272,
      "num_tokens": 56088126.0,
      "step": 22130
    },
    {
      "entropy": 6.262707757949829,
      "epoch": 2.5545297172533177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004357159280030483,
      "loss": 5.907,
      "mean_token_accuracy": 0.17797035127878189,
      "num_tokens": 56101686.0,
      "step": 22135
    },
    {
      "entropy": 6.170461416244507,
      "epoch": 2.5551067512983265,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004356870465773635,
      "loss": 5.9057,
      "mean_token_accuracy": 0.17853282690048217,
      "num_tokens": 56114605.0,
      "step": 22140
    },
    {
      "entropy": 6.090783596038818,
      "epoch": 2.5556837853433354,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043565815974704974,
      "loss": 5.7609,
      "mean_token_accuracy": 0.18233157098293304,
      "num_tokens": 56127955.0,
      "step": 22145
    },
    {
      "entropy": 6.26212511062622,
      "epoch": 2.556260819388344,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043562926751307857,
      "loss": 5.9817,
      "mean_token_accuracy": 0.17383843958377837,
      "num_tokens": 56140691.0,
      "step": 22150
    },
    {
      "entropy": 6.21481671333313,
      "epoch": 2.5568378534333527,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043560036987642177,
      "loss": 5.844,
      "mean_token_accuracy": 0.18300168365240096,
      "num_tokens": 56153815.0,
      "step": 22155
    },
    {
      "entropy": 6.226272869110107,
      "epoch": 2.557414887478361,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043557146683805134,
      "loss": 5.8628,
      "mean_token_accuracy": 0.18471735417842866,
      "num_tokens": 56167186.0,
      "step": 22160
    },
    {
      "entropy": 6.157123613357544,
      "epoch": 2.55799192152337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004355425583989393,
      "loss": 5.8508,
      "mean_token_accuracy": 0.18059379905462264,
      "num_tokens": 56179307.0,
      "step": 22165
    },
    {
      "entropy": 6.2211833000183105,
      "epoch": 2.5585689555683784,
      "grad_norm": 0.875,
      "learning_rate": 0.00043551364456005816,
      "loss": 5.8833,
      "mean_token_accuracy": 0.17549400478601457,
      "num_tokens": 56191398.0,
      "step": 22170
    },
    {
      "entropy": 6.131880474090576,
      "epoch": 2.5591459896133872,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043548472532238015,
      "loss": 5.7801,
      "mean_token_accuracy": 0.1897474303841591,
      "num_tokens": 56203871.0,
      "step": 22175
    },
    {
      "entropy": 6.11757001876831,
      "epoch": 2.559723023658396,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00043545580068687836,
      "loss": 5.8255,
      "mean_token_accuracy": 0.17875247895717622,
      "num_tokens": 56215262.0,
      "step": 22180
    },
    {
      "entropy": 6.214519453048706,
      "epoch": 2.5603000577034045,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004354268706545252,
      "loss": 5.8846,
      "mean_token_accuracy": 0.18168828189373015,
      "num_tokens": 56230190.0,
      "step": 22185
    },
    {
      "entropy": 6.1590415954589846,
      "epoch": 2.560877091748413,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00043539793522629405,
      "loss": 5.7857,
      "mean_token_accuracy": 0.18879222124814987,
      "num_tokens": 56243841.0,
      "step": 22190
    },
    {
      "entropy": 6.17527403831482,
      "epoch": 2.561454125793422,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043536899440315796,
      "loss": 5.858,
      "mean_token_accuracy": 0.18398841917514802,
      "num_tokens": 56257500.0,
      "step": 22195
    },
    {
      "entropy": 6.1491899490356445,
      "epoch": 2.5620311598384307,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004353400481860904,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1786172717809677,
      "num_tokens": 56269732.0,
      "step": 22200
    },
    {
      "entropy": 6.237949991226197,
      "epoch": 2.562608193883439,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043531109657606504,
      "loss": 5.8928,
      "mean_token_accuracy": 0.17821228951215745,
      "num_tokens": 56283192.0,
      "step": 22205
    },
    {
      "entropy": 6.160229682922363,
      "epoch": 2.563185227928448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004352821395740555,
      "loss": 5.8182,
      "mean_token_accuracy": 0.17565051317214966,
      "num_tokens": 56296312.0,
      "step": 22210
    },
    {
      "entropy": 6.200014877319336,
      "epoch": 2.5637622619734564,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004352531771810359,
      "loss": 5.8773,
      "mean_token_accuracy": 0.18192670196294786,
      "num_tokens": 56309964.0,
      "step": 22215
    },
    {
      "entropy": 6.1748268604278564,
      "epoch": 2.564339296018465,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004352242093979802,
      "loss": 5.8909,
      "mean_token_accuracy": 0.18217986822128296,
      "num_tokens": 56324002.0,
      "step": 22220
    },
    {
      "entropy": 6.218401384353638,
      "epoch": 2.5649163300634736,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004351952362258628,
      "loss": 5.8966,
      "mean_token_accuracy": 0.17519856989383698,
      "num_tokens": 56336339.0,
      "step": 22225
    },
    {
      "entropy": 6.157708406448364,
      "epoch": 2.5654933641084825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043516625766565825,
      "loss": 5.8763,
      "mean_token_accuracy": 0.18050243854522705,
      "num_tokens": 56347819.0,
      "step": 22230
    },
    {
      "entropy": 6.197275638580322,
      "epoch": 2.566070398153491,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004351372737183412,
      "loss": 5.8415,
      "mean_token_accuracy": 0.17831179946660997,
      "num_tokens": 56360764.0,
      "step": 22235
    },
    {
      "entropy": 6.172028303146362,
      "epoch": 2.5666474321985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004351082843848865,
      "loss": 5.8349,
      "mean_token_accuracy": 0.17641243785619737,
      "num_tokens": 56372731.0,
      "step": 22240
    },
    {
      "entropy": 6.195551347732544,
      "epoch": 2.567224466243508,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043507928966626916,
      "loss": 5.8413,
      "mean_token_accuracy": 0.18241804391145705,
      "num_tokens": 56385650.0,
      "step": 22245
    },
    {
      "entropy": 6.203982734680176,
      "epoch": 2.567801500288517,
      "grad_norm": 1.921875,
      "learning_rate": 0.0004350502895634644,
      "loss": 5.8601,
      "mean_token_accuracy": 0.184727743268013,
      "num_tokens": 56397076.0,
      "step": 22250
    },
    {
      "entropy": 6.13416485786438,
      "epoch": 2.568378534333526,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004350212840774476,
      "loss": 5.803,
      "mean_token_accuracy": 0.18511220961809158,
      "num_tokens": 56409779.0,
      "step": 22255
    },
    {
      "entropy": 6.203754043579101,
      "epoch": 2.5689555683785343,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043499227320919456,
      "loss": 5.903,
      "mean_token_accuracy": 0.1732098087668419,
      "num_tokens": 56423204.0,
      "step": 22260
    },
    {
      "entropy": 6.177084875106812,
      "epoch": 2.5695326024235428,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043496325695968076,
      "loss": 5.9169,
      "mean_token_accuracy": 0.17664153426885604,
      "num_tokens": 56436379.0,
      "step": 22265
    },
    {
      "entropy": 6.204816055297852,
      "epoch": 2.5701096364685516,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004349342353298823,
      "loss": 5.8336,
      "mean_token_accuracy": 0.18492823094129562,
      "num_tokens": 56449558.0,
      "step": 22270
    },
    {
      "entropy": 6.212637662887573,
      "epoch": 2.5706866705135605,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004349052083207753,
      "loss": 5.8662,
      "mean_token_accuracy": 0.17576610296964645,
      "num_tokens": 56463541.0,
      "step": 22275
    },
    {
      "entropy": 6.291454124450683,
      "epoch": 2.571263704558569,
      "grad_norm": 1.0,
      "learning_rate": 0.000434876175933336,
      "loss": 5.9757,
      "mean_token_accuracy": 0.1727263242006302,
      "num_tokens": 56476849.0,
      "step": 22280
    },
    {
      "entropy": 6.142795372009277,
      "epoch": 2.5718407386035778,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00043484713816854093,
      "loss": 5.8459,
      "mean_token_accuracy": 0.17979540526866913,
      "num_tokens": 56490089.0,
      "step": 22285
    },
    {
      "entropy": 6.150353002548218,
      "epoch": 2.572417772648586,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004348180950273668,
      "loss": 5.7761,
      "mean_token_accuracy": 0.1880112811923027,
      "num_tokens": 56501392.0,
      "step": 22290
    },
    {
      "entropy": 6.104298543930054,
      "epoch": 2.572994806693595,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004347890465107905,
      "loss": 5.7734,
      "mean_token_accuracy": 0.18166673332452773,
      "num_tokens": 56513573.0,
      "step": 22295
    },
    {
      "entropy": 6.171061754226685,
      "epoch": 2.5735718407386035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043475999261978894,
      "loss": 5.8748,
      "mean_token_accuracy": 0.17986850291490555,
      "num_tokens": 56526762.0,
      "step": 22300
    },
    {
      "entropy": 6.133188247680664,
      "epoch": 2.5741488747836123,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004347309333553393,
      "loss": 5.8224,
      "mean_token_accuracy": 0.1850934535264969,
      "num_tokens": 56538264.0,
      "step": 22305
    },
    {
      "entropy": 6.233757019042969,
      "epoch": 2.5747259088286207,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043470186871841915,
      "loss": 5.9173,
      "mean_token_accuracy": 0.16918342858552932,
      "num_tokens": 56550296.0,
      "step": 22310
    },
    {
      "entropy": 6.166134834289551,
      "epoch": 2.5753029428736296,
      "grad_norm": 1.0,
      "learning_rate": 0.000434672798710006,
      "loss": 5.8666,
      "mean_token_accuracy": 0.1760929062962532,
      "num_tokens": 56563542.0,
      "step": 22315
    },
    {
      "entropy": 6.139899110794067,
      "epoch": 2.575879976918638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043464372333107756,
      "loss": 5.8291,
      "mean_token_accuracy": 0.1866592586040497,
      "num_tokens": 56576992.0,
      "step": 22320
    },
    {
      "entropy": 6.206593036651611,
      "epoch": 2.576457010963647,
      "grad_norm": 1.171875,
      "learning_rate": 0.00043461464258261174,
      "loss": 5.8115,
      "mean_token_accuracy": 0.18988906890153884,
      "num_tokens": 56589271.0,
      "step": 22325
    },
    {
      "entropy": 5.979550552368164,
      "epoch": 2.5770340450086557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004345855564655867,
      "loss": 5.5526,
      "mean_token_accuracy": 0.20181859880685807,
      "num_tokens": 56600796.0,
      "step": 22330
    },
    {
      "entropy": 6.100833749771118,
      "epoch": 2.577611079053664,
      "grad_norm": 2.015625,
      "learning_rate": 0.00043455646498098075,
      "loss": 5.7927,
      "mean_token_accuracy": 0.18895177245140077,
      "num_tokens": 56613327.0,
      "step": 22335
    },
    {
      "entropy": 6.1338671207427975,
      "epoch": 2.5781881130986726,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043452736812977236,
      "loss": 5.8321,
      "mean_token_accuracy": 0.18056489527225494,
      "num_tokens": 56626554.0,
      "step": 22340
    },
    {
      "entropy": 6.23466854095459,
      "epoch": 2.5787651471436814,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004344982659129402,
      "loss": 5.859,
      "mean_token_accuracy": 0.1847481310367584,
      "num_tokens": 56639960.0,
      "step": 22345
    },
    {
      "entropy": 6.136874103546143,
      "epoch": 2.5793421811886903,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004344691583314631,
      "loss": 5.8237,
      "mean_token_accuracy": 0.18483690321445465,
      "num_tokens": 56651330.0,
      "step": 22350
    },
    {
      "entropy": 6.215026140213013,
      "epoch": 2.5799192152336987,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043444004538632005,
      "loss": 5.9473,
      "mean_token_accuracy": 0.18039728254079818,
      "num_tokens": 56664359.0,
      "step": 22355
    },
    {
      "entropy": 6.193694114685059,
      "epoch": 2.5804962492787076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043441092707849024,
      "loss": 5.8498,
      "mean_token_accuracy": 0.1780017077922821,
      "num_tokens": 56676095.0,
      "step": 22360
    },
    {
      "entropy": 6.206061744689942,
      "epoch": 2.581073283323716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043438180340895315,
      "loss": 5.9102,
      "mean_token_accuracy": 0.17930631786584855,
      "num_tokens": 56689918.0,
      "step": 22365
    },
    {
      "entropy": 6.277228593826294,
      "epoch": 2.581650317368725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043435267437868825,
      "loss": 5.9054,
      "mean_token_accuracy": 0.18055343627929688,
      "num_tokens": 56702308.0,
      "step": 22370
    },
    {
      "entropy": 6.1466456890106205,
      "epoch": 2.5822273514137333,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004343235399886753,
      "loss": 5.8509,
      "mean_token_accuracy": 0.18117641508579255,
      "num_tokens": 56713928.0,
      "step": 22375
    },
    {
      "entropy": 6.232672452926636,
      "epoch": 2.582804385458742,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004342944002398942,
      "loss": 5.9054,
      "mean_token_accuracy": 0.18088122457265854,
      "num_tokens": 56727065.0,
      "step": 22380
    },
    {
      "entropy": 6.162808561325074,
      "epoch": 2.583381419503751,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004342652551333251,
      "loss": 5.8572,
      "mean_token_accuracy": 0.18189468681812287,
      "num_tokens": 56739517.0,
      "step": 22385
    },
    {
      "entropy": 6.158969783782959,
      "epoch": 2.5839584535487594,
      "grad_norm": 1.953125,
      "learning_rate": 0.0004342361046699482,
      "loss": 5.8515,
      "mean_token_accuracy": 0.17724740207195283,
      "num_tokens": 56753044.0,
      "step": 22390
    },
    {
      "entropy": 6.177622509002686,
      "epoch": 2.584535487593768,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004342069488507441,
      "loss": 5.83,
      "mean_token_accuracy": 0.17820962965488435,
      "num_tokens": 56765500.0,
      "step": 22395
    },
    {
      "entropy": 6.2537233352661135,
      "epoch": 2.5851125216387767,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004341777876766933,
      "loss": 5.923,
      "mean_token_accuracy": 0.17215094119310378,
      "num_tokens": 56777895.0,
      "step": 22400
    },
    {
      "entropy": 6.206136703491211,
      "epoch": 2.5856895556837856,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004341486211487767,
      "loss": 5.7942,
      "mean_token_accuracy": 0.18706935048103332,
      "num_tokens": 56791523.0,
      "step": 22405
    },
    {
      "entropy": 6.158300065994263,
      "epoch": 2.586266589728794,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004341194492679753,
      "loss": 5.8933,
      "mean_token_accuracy": 0.17521030008792876,
      "num_tokens": 56803490.0,
      "step": 22410
    },
    {
      "entropy": 6.24560317993164,
      "epoch": 2.5868436237738024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043409027203527016,
      "loss": 5.8913,
      "mean_token_accuracy": 0.174959959089756,
      "num_tokens": 56814398.0,
      "step": 22415
    },
    {
      "entropy": 6.102668190002442,
      "epoch": 2.5874206578188113,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00043406108945164283,
      "loss": 5.8111,
      "mean_token_accuracy": 0.18775819838047028,
      "num_tokens": 56826591.0,
      "step": 22420
    },
    {
      "entropy": 6.122554016113281,
      "epoch": 2.58799769186382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043403190151807474,
      "loss": 5.8244,
      "mean_token_accuracy": 0.17712070196866989,
      "num_tokens": 56839208.0,
      "step": 22425
    },
    {
      "entropy": 6.1992920398712155,
      "epoch": 2.5885747259088285,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004340027082355477,
      "loss": 5.8622,
      "mean_token_accuracy": 0.17706735134124757,
      "num_tokens": 56852211.0,
      "step": 22430
    },
    {
      "entropy": 6.122121286392212,
      "epoch": 2.5891517599538374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004339735096050434,
      "loss": 5.8376,
      "mean_token_accuracy": 0.18258956223726272,
      "num_tokens": 56864691.0,
      "step": 22435
    },
    {
      "entropy": 6.20952353477478,
      "epoch": 2.589728793998846,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004339443056275443,
      "loss": 5.7929,
      "mean_token_accuracy": 0.18961059153079987,
      "num_tokens": 56877473.0,
      "step": 22440
    },
    {
      "entropy": 6.247283840179444,
      "epoch": 2.5903058280438547,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043391509630403224,
      "loss": 5.9191,
      "mean_token_accuracy": 0.18167265951633454,
      "num_tokens": 56889852.0,
      "step": 22445
    },
    {
      "entropy": 6.231794357299805,
      "epoch": 2.590882862088863,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043388588163549,
      "loss": 5.9432,
      "mean_token_accuracy": 0.17553926706314088,
      "num_tokens": 56901882.0,
      "step": 22450
    },
    {
      "entropy": 6.25377631187439,
      "epoch": 2.591459896133872,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043385666162289995,
      "loss": 5.8584,
      "mean_token_accuracy": 0.18067149221897125,
      "num_tokens": 56913212.0,
      "step": 22455
    },
    {
      "entropy": 6.176819562911987,
      "epoch": 2.592036930178881,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043382743626724506,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17613870203495025,
      "num_tokens": 56924490.0,
      "step": 22460
    },
    {
      "entropy": 6.216876792907715,
      "epoch": 2.5926139642238892,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004337982055695082,
      "loss": 5.9621,
      "mean_token_accuracy": 0.17417846173048018,
      "num_tokens": 56937558.0,
      "step": 22465
    },
    {
      "entropy": 6.149514627456665,
      "epoch": 2.5931909982688977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004337689695306726,
      "loss": 5.8374,
      "mean_token_accuracy": 0.18988426476716996,
      "num_tokens": 56950544.0,
      "step": 22470
    },
    {
      "entropy": 6.235901975631714,
      "epoch": 2.5937680323139065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004337397281517215,
      "loss": 5.87,
      "mean_token_accuracy": 0.17309301793575288,
      "num_tokens": 56962972.0,
      "step": 22475
    },
    {
      "entropy": 6.211310338973999,
      "epoch": 2.5943450663589154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004337104814336386,
      "loss": 5.847,
      "mean_token_accuracy": 0.17467414140701293,
      "num_tokens": 56975972.0,
      "step": 22480
    },
    {
      "entropy": 6.277341842651367,
      "epoch": 2.594922100403924,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004336812293774075,
      "loss": 5.9832,
      "mean_token_accuracy": 0.17419557720422746,
      "num_tokens": 56989106.0,
      "step": 22485
    },
    {
      "entropy": 6.155555486679077,
      "epoch": 2.595499134448932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043365197198401195,
      "loss": 5.856,
      "mean_token_accuracy": 0.17909400910139084,
      "num_tokens": 57002544.0,
      "step": 22490
    },
    {
      "entropy": 6.2688497543334964,
      "epoch": 2.596076168493941,
      "grad_norm": 1.0,
      "learning_rate": 0.0004336227092544362,
      "loss": 5.9188,
      "mean_token_accuracy": 0.17357267439365387,
      "num_tokens": 57016526.0,
      "step": 22495
    },
    {
      "entropy": 6.294514799118042,
      "epoch": 2.59665320253895,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004335934411896644,
      "loss": 5.9078,
      "mean_token_accuracy": 0.1818125218153,
      "num_tokens": 57029817.0,
      "step": 22500
    },
    {
      "entropy": 6.10988826751709,
      "epoch": 2.5972302365839584,
      "grad_norm": 0.9375,
      "learning_rate": 0.000433564167790681,
      "loss": 5.8322,
      "mean_token_accuracy": 0.18750934898853303,
      "num_tokens": 57042189.0,
      "step": 22505
    },
    {
      "entropy": 6.152314138412476,
      "epoch": 2.597807270628967,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004335348890584706,
      "loss": 5.8396,
      "mean_token_accuracy": 0.18023107200860977,
      "num_tokens": 57054297.0,
      "step": 22510
    },
    {
      "entropy": 6.193462610244751,
      "epoch": 2.5983843046739756,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043350560499401793,
      "loss": 5.8282,
      "mean_token_accuracy": 0.1868427127599716,
      "num_tokens": 57067530.0,
      "step": 22515
    },
    {
      "entropy": 6.210541725158691,
      "epoch": 2.5989613387189845,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004334763155983079,
      "loss": 5.9267,
      "mean_token_accuracy": 0.17492200434207916,
      "num_tokens": 57080877.0,
      "step": 22520
    },
    {
      "entropy": 6.236031532287598,
      "epoch": 2.599538372763993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043344702087232584,
      "loss": 5.8711,
      "mean_token_accuracy": 0.18254067599773408,
      "num_tokens": 57092716.0,
      "step": 22525
    },
    {
      "entropy": 6.0933469295501705,
      "epoch": 2.6001154068090018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004334177208170569,
      "loss": 5.8306,
      "mean_token_accuracy": 0.1827499106526375,
      "num_tokens": 57107335.0,
      "step": 22530
    },
    {
      "entropy": 6.215653657913208,
      "epoch": 2.6006924408540106,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043338841543348654,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18515121638774873,
      "num_tokens": 57121691.0,
      "step": 22535
    },
    {
      "entropy": 6.196254396438599,
      "epoch": 2.601269474899019,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043335910472260057,
      "loss": 5.9195,
      "mean_token_accuracy": 0.17326430529356002,
      "num_tokens": 57134536.0,
      "step": 22540
    },
    {
      "entropy": 6.200167274475097,
      "epoch": 2.6018465089440275,
      "grad_norm": 1.140625,
      "learning_rate": 0.00043332978868538483,
      "loss": 5.8288,
      "mean_token_accuracy": 0.1783520206809044,
      "num_tokens": 57147141.0,
      "step": 22545
    },
    {
      "entropy": 6.157841968536377,
      "epoch": 2.6024235429890363,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043330046732282515,
      "loss": 5.8142,
      "mean_token_accuracy": 0.19236651360988616,
      "num_tokens": 57160097.0,
      "step": 22550
    },
    {
      "entropy": 6.045707750320434,
      "epoch": 2.603000577034045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043327114063590797,
      "loss": 5.7335,
      "mean_token_accuracy": 0.2020871952176094,
      "num_tokens": 57173378.0,
      "step": 22555
    },
    {
      "entropy": 6.125089836120606,
      "epoch": 2.6035776110790536,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004332418086256196,
      "loss": 5.8373,
      "mean_token_accuracy": 0.1871074229478836,
      "num_tokens": 57187957.0,
      "step": 22560
    },
    {
      "entropy": 6.209305953979492,
      "epoch": 2.6041546451240625,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004332124712929466,
      "loss": 5.888,
      "mean_token_accuracy": 0.1780647099018097,
      "num_tokens": 57202186.0,
      "step": 22565
    },
    {
      "entropy": 6.264482593536377,
      "epoch": 2.604731679169071,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043318312863887566,
      "loss": 5.8676,
      "mean_token_accuracy": 0.17797986567020416,
      "num_tokens": 57213841.0,
      "step": 22570
    },
    {
      "entropy": 6.208065366744995,
      "epoch": 2.6053087132140798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043315378066439384,
      "loss": 5.8618,
      "mean_token_accuracy": 0.18109374940395356,
      "num_tokens": 57226445.0,
      "step": 22575
    },
    {
      "entropy": 6.134828519821167,
      "epoch": 2.605885747259088,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004331244273704881,
      "loss": 5.8116,
      "mean_token_accuracy": 0.19002193212509155,
      "num_tokens": 57238913.0,
      "step": 22580
    },
    {
      "entropy": 6.119336175918579,
      "epoch": 2.606462781304097,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004330950687581458,
      "loss": 5.8125,
      "mean_token_accuracy": 0.19199034571647644,
      "num_tokens": 57251082.0,
      "step": 22585
    },
    {
      "entropy": 6.197483110427856,
      "epoch": 2.6070398153491054,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043306570482835443,
      "loss": 5.8825,
      "mean_token_accuracy": 0.18067598044872285,
      "num_tokens": 57263267.0,
      "step": 22590
    },
    {
      "entropy": 6.235991144180298,
      "epoch": 2.6076168493941143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004330363355821016,
      "loss": 5.8547,
      "mean_token_accuracy": 0.18330008238554002,
      "num_tokens": 57277053.0,
      "step": 22595
    },
    {
      "entropy": 6.172784519195557,
      "epoch": 2.6081938834391227,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004330069610203751,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1853868395090103,
      "num_tokens": 57289962.0,
      "step": 22600
    },
    {
      "entropy": 6.190165567398071,
      "epoch": 2.6087709174841316,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004329775811441629,
      "loss": 5.8971,
      "mean_token_accuracy": 0.1774592414498329,
      "num_tokens": 57302397.0,
      "step": 22605
    },
    {
      "entropy": 6.18120002746582,
      "epoch": 2.6093479515291405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043294819595445324,
      "loss": 5.7932,
      "mean_token_accuracy": 0.1860073208808899,
      "num_tokens": 57316645.0,
      "step": 22610
    },
    {
      "entropy": 6.165102243423462,
      "epoch": 2.609924985574149,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004329188054522345,
      "loss": 5.8189,
      "mean_token_accuracy": 0.18246632665395737,
      "num_tokens": 57329714.0,
      "step": 22615
    },
    {
      "entropy": 6.2067405700683596,
      "epoch": 2.6105020196191573,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004328894096384952,
      "loss": 5.8971,
      "mean_token_accuracy": 0.17518771439790726,
      "num_tokens": 57342940.0,
      "step": 22620
    },
    {
      "entropy": 6.1361020565032955,
      "epoch": 2.611079053664166,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000432860008514224,
      "loss": 5.7378,
      "mean_token_accuracy": 0.18514444828033447,
      "num_tokens": 57354729.0,
      "step": 22625
    },
    {
      "entropy": 6.271011018753052,
      "epoch": 2.611656087709175,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004328306020804099,
      "loss": 5.8937,
      "mean_token_accuracy": 0.17303266376256943,
      "num_tokens": 57368778.0,
      "step": 22630
    },
    {
      "entropy": 6.180638217926026,
      "epoch": 2.6122331217541834,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043280119033804184,
      "loss": 5.8636,
      "mean_token_accuracy": 0.1785882219672203,
      "num_tokens": 57382266.0,
      "step": 22635
    },
    {
      "entropy": 6.188758850097656,
      "epoch": 2.6128101557991923,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043277177328810914,
      "loss": 5.8259,
      "mean_token_accuracy": 0.18019461184740065,
      "num_tokens": 57394127.0,
      "step": 22640
    },
    {
      "entropy": 6.19890513420105,
      "epoch": 2.6133871898442007,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004327423509316012,
      "loss": 5.8244,
      "mean_token_accuracy": 0.1807365223765373,
      "num_tokens": 57405711.0,
      "step": 22645
    },
    {
      "entropy": 6.173237562179565,
      "epoch": 2.6139642238892096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043271292326950764,
      "loss": 5.9256,
      "mean_token_accuracy": 0.17277692258358002,
      "num_tokens": 57418603.0,
      "step": 22650
    },
    {
      "entropy": 6.215972232818603,
      "epoch": 2.614541257934218,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004326834903028182,
      "loss": 5.7443,
      "mean_token_accuracy": 0.18458559960126877,
      "num_tokens": 57430582.0,
      "step": 22655
    },
    {
      "entropy": 6.168799257278442,
      "epoch": 2.615118291979227,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000432654052032523,
      "loss": 5.8031,
      "mean_token_accuracy": 0.1851634979248047,
      "num_tokens": 57442592.0,
      "step": 22660
    },
    {
      "entropy": 6.145985746383667,
      "epoch": 2.6156953260242353,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000432624608459612,
      "loss": 5.8616,
      "mean_token_accuracy": 0.18371669352054595,
      "num_tokens": 57455016.0,
      "step": 22665
    },
    {
      "entropy": 6.159763479232788,
      "epoch": 2.616272360069244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004325951595850756,
      "loss": 5.7587,
      "mean_token_accuracy": 0.1854664921760559,
      "num_tokens": 57467228.0,
      "step": 22670
    },
    {
      "entropy": 6.160779428482056,
      "epoch": 2.6168493941142525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043256570540990425,
      "loss": 5.7753,
      "mean_token_accuracy": 0.18470061123371123,
      "num_tokens": 57479654.0,
      "step": 22675
    },
    {
      "entropy": 6.203868961334228,
      "epoch": 2.6174264281592614,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004325362459350888,
      "loss": 5.8842,
      "mean_token_accuracy": 0.17926905751228334,
      "num_tokens": 57492695.0,
      "step": 22680
    },
    {
      "entropy": 6.205330038070679,
      "epoch": 2.6180034622042703,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043250678116161985,
      "loss": 5.9065,
      "mean_token_accuracy": 0.18414704203605653,
      "num_tokens": 57504619.0,
      "step": 22685
    },
    {
      "entropy": 6.232370328903198,
      "epoch": 2.6185804962492787,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043247731109048856,
      "loss": 5.8923,
      "mean_token_accuracy": 0.17923021018505098,
      "num_tokens": 57517141.0,
      "step": 22690
    },
    {
      "entropy": 6.173137092590332,
      "epoch": 2.619157530294287,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004324478357226861,
      "loss": 5.8417,
      "mean_token_accuracy": 0.18583366870880128,
      "num_tokens": 57529759.0,
      "step": 22695
    },
    {
      "entropy": 6.197889995574951,
      "epoch": 2.619734564339296,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043241835505920405,
      "loss": 5.8301,
      "mean_token_accuracy": 0.17693443298339845,
      "num_tokens": 57542587.0,
      "step": 22700
    },
    {
      "entropy": 6.198021316528321,
      "epoch": 2.620311598384305,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004323888691010337,
      "loss": 5.8535,
      "mean_token_accuracy": 0.18374460190534592,
      "num_tokens": 57554539.0,
      "step": 22705
    },
    {
      "entropy": 6.246301698684692,
      "epoch": 2.6208886324293132,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043235937784916694,
      "loss": 5.831,
      "mean_token_accuracy": 0.18612860143184662,
      "num_tokens": 57567436.0,
      "step": 22710
    },
    {
      "entropy": 6.098443222045899,
      "epoch": 2.621465666474322,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043232988130459567,
      "loss": 5.7574,
      "mean_token_accuracy": 0.18867213726043702,
      "num_tokens": 57579656.0,
      "step": 22715
    },
    {
      "entropy": 6.038045406341553,
      "epoch": 2.6220427005193305,
      "grad_norm": 0.953125,
      "learning_rate": 0.000432300379468312,
      "loss": 5.7335,
      "mean_token_accuracy": 0.1917673110961914,
      "num_tokens": 57591794.0,
      "step": 22720
    },
    {
      "entropy": 6.091014623641968,
      "epoch": 2.6226197345643394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004322708723413082,
      "loss": 5.787,
      "mean_token_accuracy": 0.18332008719444276,
      "num_tokens": 57603982.0,
      "step": 22725
    },
    {
      "entropy": 6.23874797821045,
      "epoch": 2.623196768609348,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043224135992457667,
      "loss": 5.9055,
      "mean_token_accuracy": 0.1817587748169899,
      "num_tokens": 57616769.0,
      "step": 22730
    },
    {
      "entropy": 6.125768566131592,
      "epoch": 2.6237738026543567,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004322118422191102,
      "loss": 5.7888,
      "mean_token_accuracy": 0.1895359516143799,
      "num_tokens": 57628995.0,
      "step": 22735
    },
    {
      "entropy": 6.1930314064025875,
      "epoch": 2.6243508366993655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043218231922590144,
      "loss": 5.8196,
      "mean_token_accuracy": 0.19127937406301498,
      "num_tokens": 57642417.0,
      "step": 22740
    },
    {
      "entropy": 6.0228982925415036,
      "epoch": 2.624927870744374,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004321527909459435,
      "loss": 5.6678,
      "mean_token_accuracy": 0.18513512015342712,
      "num_tokens": 57654395.0,
      "step": 22745
    },
    {
      "entropy": 6.201718091964722,
      "epoch": 2.6255049047893824,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043212325738022946,
      "loss": 5.8623,
      "mean_token_accuracy": 0.17867814153432846,
      "num_tokens": 57667010.0,
      "step": 22750
    },
    {
      "entropy": 6.13390851020813,
      "epoch": 2.6260819388343912,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004320937185297527,
      "loss": 5.8167,
      "mean_token_accuracy": 0.18668165057897568,
      "num_tokens": 57678726.0,
      "step": 22755
    },
    {
      "entropy": 6.169459342956543,
      "epoch": 2.6266589728794,
      "grad_norm": 1.53125,
      "learning_rate": 0.00043206417439550684,
      "loss": 5.893,
      "mean_token_accuracy": 0.18388821184635162,
      "num_tokens": 57692386.0,
      "step": 22760
    },
    {
      "entropy": 6.155322551727295,
      "epoch": 2.6272360069244085,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043203462497848534,
      "loss": 5.8152,
      "mean_token_accuracy": 0.18205110281705855,
      "num_tokens": 57704647.0,
      "step": 22765
    },
    {
      "entropy": 6.081495046615601,
      "epoch": 2.627813040969417,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043200507027968236,
      "loss": 5.7704,
      "mean_token_accuracy": 0.19250834733247757,
      "num_tokens": 57717652.0,
      "step": 22770
    },
    {
      "entropy": 6.16862530708313,
      "epoch": 2.628390075014426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004319755103000918,
      "loss": 5.8647,
      "mean_token_accuracy": 0.1885334476828575,
      "num_tokens": 57730531.0,
      "step": 22775
    },
    {
      "entropy": 6.079220962524414,
      "epoch": 2.6289671090594346,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004319459450407079,
      "loss": 5.6871,
      "mean_token_accuracy": 0.1937173992395401,
      "num_tokens": 57742627.0,
      "step": 22780
    },
    {
      "entropy": 6.147434091567993,
      "epoch": 2.629544143104443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043191637450252514,
      "loss": 5.8313,
      "mean_token_accuracy": 0.17829314172267913,
      "num_tokens": 57755894.0,
      "step": 22785
    },
    {
      "entropy": 6.232344913482666,
      "epoch": 2.630121177149452,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000431886798686538,
      "loss": 5.8878,
      "mean_token_accuracy": 0.1800585061311722,
      "num_tokens": 57769003.0,
      "step": 22790
    },
    {
      "entropy": 6.158716249465942,
      "epoch": 2.6306982111944603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004318572175937414,
      "loss": 5.8669,
      "mean_token_accuracy": 0.18777997195720672,
      "num_tokens": 57781230.0,
      "step": 22795
    },
    {
      "entropy": 6.16593108177185,
      "epoch": 2.631275245239469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004318276312251301,
      "loss": 5.7983,
      "mean_token_accuracy": 0.18250776678323746,
      "num_tokens": 57793167.0,
      "step": 22800
    },
    {
      "entropy": 6.145012617111206,
      "epoch": 2.6318522792844776,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004317980395816994,
      "loss": 5.789,
      "mean_token_accuracy": 0.1874377980828285,
      "num_tokens": 57805306.0,
      "step": 22805
    },
    {
      "entropy": 6.126709365844727,
      "epoch": 2.6324293133294865,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004317684426644445,
      "loss": 5.7608,
      "mean_token_accuracy": 0.18417045325040818,
      "num_tokens": 57818844.0,
      "step": 22810
    },
    {
      "entropy": 6.186257839202881,
      "epoch": 2.6330063473744953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043173884047436093,
      "loss": 5.9246,
      "mean_token_accuracy": 0.17446522414684296,
      "num_tokens": 57832017.0,
      "step": 22815
    },
    {
      "entropy": 6.1840986728668215,
      "epoch": 2.6335833814195038,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004317092330124443,
      "loss": 5.8519,
      "mean_token_accuracy": 0.18455812633037566,
      "num_tokens": 57844900.0,
      "step": 22820
    },
    {
      "entropy": 6.213629817962646,
      "epoch": 2.634160415464512,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043167962027969043,
      "loss": 5.8457,
      "mean_token_accuracy": 0.17783587872982026,
      "num_tokens": 57858297.0,
      "step": 22825
    },
    {
      "entropy": 6.163678932189941,
      "epoch": 2.634737449509521,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004316500022770954,
      "loss": 5.8192,
      "mean_token_accuracy": 0.18586844950914383,
      "num_tokens": 57870613.0,
      "step": 22830
    },
    {
      "entropy": 6.214581489562988,
      "epoch": 2.63531448355453,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00043162037900565533,
      "loss": 5.8916,
      "mean_token_accuracy": 0.1730441316962242,
      "num_tokens": 57883991.0,
      "step": 22835
    },
    {
      "entropy": 6.2626752853393555,
      "epoch": 2.6358915175995383,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043159075046636666,
      "loss": 5.9093,
      "mean_token_accuracy": 0.17782390117645264,
      "num_tokens": 57896400.0,
      "step": 22840
    },
    {
      "entropy": 6.103964328765869,
      "epoch": 2.636468551644547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004315611166602258,
      "loss": 5.7805,
      "mean_token_accuracy": 0.1902986317873001,
      "num_tokens": 57908558.0,
      "step": 22845
    },
    {
      "entropy": 6.089307260513306,
      "epoch": 2.6370455856895556,
      "grad_norm": 1.0,
      "learning_rate": 0.00043153147758822957,
      "loss": 5.7618,
      "mean_token_accuracy": 0.18054557591676712,
      "num_tokens": 57922312.0,
      "step": 22850
    },
    {
      "entropy": 6.195387697219848,
      "epoch": 2.6376226197345645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004315018332513749,
      "loss": 5.7757,
      "mean_token_accuracy": 0.18656957298517227,
      "num_tokens": 57934406.0,
      "step": 22855
    },
    {
      "entropy": 6.155811071395874,
      "epoch": 2.638199653779573,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004314721836506587,
      "loss": 5.8794,
      "mean_token_accuracy": 0.1820259690284729,
      "num_tokens": 57946854.0,
      "step": 22860
    },
    {
      "entropy": 6.215509033203125,
      "epoch": 2.6387766878245817,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043144252878707845,
      "loss": 5.9281,
      "mean_token_accuracy": 0.1770135998725891,
      "num_tokens": 57959075.0,
      "step": 22865
    },
    {
      "entropy": 6.120176601409912,
      "epoch": 2.63935372186959,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004314128686616314,
      "loss": 5.7203,
      "mean_token_accuracy": 0.19129203855991364,
      "num_tokens": 57970501.0,
      "step": 22870
    },
    {
      "entropy": 6.048280096054077,
      "epoch": 2.639930755914599,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004313832032753152,
      "loss": 5.7591,
      "mean_token_accuracy": 0.1910104587674141,
      "num_tokens": 57983683.0,
      "step": 22875
    },
    {
      "entropy": 6.211901140213013,
      "epoch": 2.6405077899596074,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043135353262912764,
      "loss": 5.8558,
      "mean_token_accuracy": 0.18604885488748552,
      "num_tokens": 57997603.0,
      "step": 22880
    },
    {
      "entropy": 6.283529663085938,
      "epoch": 2.6410848240046163,
      "grad_norm": 0.90625,
      "learning_rate": 0.00043132385672406664,
      "loss": 5.9466,
      "mean_token_accuracy": 0.1737075001001358,
      "num_tokens": 58010225.0,
      "step": 22885
    },
    {
      "entropy": 6.132202672958374,
      "epoch": 2.641661858049625,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004312941755611305,
      "loss": 5.7351,
      "mean_token_accuracy": 0.19030017256736756,
      "num_tokens": 58023581.0,
      "step": 22890
    },
    {
      "entropy": 6.167155075073242,
      "epoch": 2.6422388920946336,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043126448914131724,
      "loss": 5.8681,
      "mean_token_accuracy": 0.1788954794406891,
      "num_tokens": 58035508.0,
      "step": 22895
    },
    {
      "entropy": 6.248766279220581,
      "epoch": 2.642815926139642,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004312347974656256,
      "loss": 5.8354,
      "mean_token_accuracy": 0.17860854268074036,
      "num_tokens": 58046762.0,
      "step": 22900
    },
    {
      "entropy": 6.228690481185913,
      "epoch": 2.643392960184651,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004312051005350541,
      "loss": 5.9067,
      "mean_token_accuracy": 0.17975085377693176,
      "num_tokens": 58059626.0,
      "step": 22905
    },
    {
      "entropy": 6.136456918716431,
      "epoch": 2.6439699942296597,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043117539835060164,
      "loss": 5.7981,
      "mean_token_accuracy": 0.19189803451299667,
      "num_tokens": 58073482.0,
      "step": 22910
    },
    {
      "entropy": 6.153267812728882,
      "epoch": 2.644547028274668,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004311456909132673,
      "loss": 5.8394,
      "mean_token_accuracy": 0.184129199385643,
      "num_tokens": 58086375.0,
      "step": 22915
    },
    {
      "entropy": 6.2739767074584964,
      "epoch": 2.645124062319677,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004311159782240502,
      "loss": 5.8521,
      "mean_token_accuracy": 0.17692941278219224,
      "num_tokens": 58097953.0,
      "step": 22920
    },
    {
      "entropy": 6.16820387840271,
      "epoch": 2.6457010963646854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004310862602839496,
      "loss": 5.8726,
      "mean_token_accuracy": 0.18403665572404862,
      "num_tokens": 58109602.0,
      "step": 22925
    },
    {
      "entropy": 6.049224853515625,
      "epoch": 2.6462781304096943,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004310565370939653,
      "loss": 5.7647,
      "mean_token_accuracy": 0.1887705072760582,
      "num_tokens": 58123033.0,
      "step": 22930
    },
    {
      "entropy": 6.20648980140686,
      "epoch": 2.6468551644547027,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043102680865509687,
      "loss": 5.9095,
      "mean_token_accuracy": 0.1757308855652809,
      "num_tokens": 58135629.0,
      "step": 22935
    },
    {
      "entropy": 6.24998369216919,
      "epoch": 2.6474321984997116,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004309970749683442,
      "loss": 5.8592,
      "mean_token_accuracy": 0.18160168528556825,
      "num_tokens": 58149178.0,
      "step": 22940
    },
    {
      "entropy": 6.153095293045044,
      "epoch": 2.64800923254472,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043096733603470737,
      "loss": 5.7384,
      "mean_token_accuracy": 0.18905709087848663,
      "num_tokens": 58160752.0,
      "step": 22945
    },
    {
      "entropy": 6.163722038269043,
      "epoch": 2.648586266589729,
      "grad_norm": 1.28125,
      "learning_rate": 0.00043093759185518677,
      "loss": 5.9308,
      "mean_token_accuracy": 0.17847282886505128,
      "num_tokens": 58174680.0,
      "step": 22950
    },
    {
      "entropy": 6.252099990844727,
      "epoch": 2.6491633006347373,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043090784243078264,
      "loss": 5.8979,
      "mean_token_accuracy": 0.17616451978683473,
      "num_tokens": 58186945.0,
      "step": 22955
    },
    {
      "entropy": 6.223245477676391,
      "epoch": 2.649740334679746,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004308780877624957,
      "loss": 5.8383,
      "mean_token_accuracy": 0.18151946514844894,
      "num_tokens": 58198940.0,
      "step": 22960
    },
    {
      "entropy": 6.184252595901489,
      "epoch": 2.650317368724755,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004308483278513267,
      "loss": 5.8718,
      "mean_token_accuracy": 0.18212129324674606,
      "num_tokens": 58212778.0,
      "step": 22965
    },
    {
      "entropy": 6.1689540386199955,
      "epoch": 2.6508944027697634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043081856269827656,
      "loss": 5.8303,
      "mean_token_accuracy": 0.18636085242033004,
      "num_tokens": 58225378.0,
      "step": 22970
    },
    {
      "entropy": 6.157000350952148,
      "epoch": 2.651471436814772,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004307887923043465,
      "loss": 5.792,
      "mean_token_accuracy": 0.1931401327252388,
      "num_tokens": 58238012.0,
      "step": 22975
    },
    {
      "entropy": 6.238514184951782,
      "epoch": 2.6520484708597807,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004307590166705379,
      "loss": 5.9452,
      "mean_token_accuracy": 0.1742991492152214,
      "num_tokens": 58250527.0,
      "step": 22980
    },
    {
      "entropy": 6.142331838607788,
      "epoch": 2.6526255049047895,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043072923579785203,
      "loss": 5.7686,
      "mean_token_accuracy": 0.1900982975959778,
      "num_tokens": 58262805.0,
      "step": 22985
    },
    {
      "entropy": 6.192170715332031,
      "epoch": 2.653202538949798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004306994496872907,
      "loss": 5.9087,
      "mean_token_accuracy": 0.17857062667608262,
      "num_tokens": 58275622.0,
      "step": 22990
    },
    {
      "entropy": 6.0951831340789795,
      "epoch": 2.653779572994807,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043066965833985567,
      "loss": 5.9036,
      "mean_token_accuracy": 0.17930659353733064,
      "num_tokens": 58288505.0,
      "step": 22995
    },
    {
      "entropy": 6.184918308258057,
      "epoch": 2.6543566070398152,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043063986175654906,
      "loss": 5.8034,
      "mean_token_accuracy": 0.18367572724819184,
      "num_tokens": 58301083.0,
      "step": 23000
    },
    {
      "entropy": 6.157433605194091,
      "epoch": 2.654933641084824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043061005993837306,
      "loss": 5.7581,
      "mean_token_accuracy": 0.18601444363594055,
      "num_tokens": 58313494.0,
      "step": 23005
    },
    {
      "entropy": 6.1464615821838375,
      "epoch": 2.6555106751298325,
      "grad_norm": 1.328125,
      "learning_rate": 0.00043058025288632995,
      "loss": 5.826,
      "mean_token_accuracy": 0.189591246843338,
      "num_tokens": 58327350.0,
      "step": 23010
    },
    {
      "entropy": 6.205576419830322,
      "epoch": 2.6560877091748414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043055044060142224,
      "loss": 5.8823,
      "mean_token_accuracy": 0.18251786530017852,
      "num_tokens": 58340664.0,
      "step": 23015
    },
    {
      "entropy": 6.193451690673828,
      "epoch": 2.6566647432198502,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00043052062308465277,
      "loss": 5.8058,
      "mean_token_accuracy": 0.17998293340206145,
      "num_tokens": 58353461.0,
      "step": 23020
    },
    {
      "entropy": 6.192967462539673,
      "epoch": 2.6572417772648587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004304908003370244,
      "loss": 5.8518,
      "mean_token_accuracy": 0.18132518231868744,
      "num_tokens": 58366872.0,
      "step": 23025
    },
    {
      "entropy": 6.190752363204956,
      "epoch": 2.657818811309867,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043046097235954024,
      "loss": 5.7997,
      "mean_token_accuracy": 0.19072414934635162,
      "num_tokens": 58378587.0,
      "step": 23030
    },
    {
      "entropy": 6.135037183761597,
      "epoch": 2.658395845354876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043043113915320345,
      "loss": 5.7715,
      "mean_token_accuracy": 0.1846110001206398,
      "num_tokens": 58390368.0,
      "step": 23035
    },
    {
      "entropy": 6.167164945602417,
      "epoch": 2.658972879399885,
      "grad_norm": 0.921875,
      "learning_rate": 0.00043040130071901747,
      "loss": 5.8829,
      "mean_token_accuracy": 0.18048466444015504,
      "num_tokens": 58403803.0,
      "step": 23040
    },
    {
      "entropy": 6.209956645965576,
      "epoch": 2.659549913444893,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043037145705798605,
      "loss": 5.86,
      "mean_token_accuracy": 0.18303288370370865,
      "num_tokens": 58416819.0,
      "step": 23045
    },
    {
      "entropy": 6.196471643447876,
      "epoch": 2.6601269474899016,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043034160817111275,
      "loss": 5.8312,
      "mean_token_accuracy": 0.18922780454158783,
      "num_tokens": 58428681.0,
      "step": 23050
    },
    {
      "entropy": 6.160715961456299,
      "epoch": 2.6607039815349105,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043031175405940157,
      "loss": 5.8466,
      "mean_token_accuracy": 0.18477230817079543,
      "num_tokens": 58440235.0,
      "step": 23055
    },
    {
      "entropy": 6.191605281829834,
      "epoch": 2.6612810155799194,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004302818947238568,
      "loss": 5.8602,
      "mean_token_accuracy": 0.18682175427675246,
      "num_tokens": 58452547.0,
      "step": 23060
    },
    {
      "entropy": 6.091343355178833,
      "epoch": 2.661858049624928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043025203016548263,
      "loss": 5.7705,
      "mean_token_accuracy": 0.1843006744980812,
      "num_tokens": 58466683.0,
      "step": 23065
    },
    {
      "entropy": 6.142626857757568,
      "epoch": 2.6624350836699366,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00043022216038528347,
      "loss": 5.8357,
      "mean_token_accuracy": 0.18257568180561065,
      "num_tokens": 58480353.0,
      "step": 23070
    },
    {
      "entropy": 6.223905229568482,
      "epoch": 2.663012117714945,
      "grad_norm": 1.125,
      "learning_rate": 0.00043019228538426416,
      "loss": 5.8335,
      "mean_token_accuracy": 0.18898356705904007,
      "num_tokens": 58493213.0,
      "step": 23075
    },
    {
      "entropy": 6.1856465339660645,
      "epoch": 2.663589151759954,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043016240516342933,
      "loss": 5.8817,
      "mean_token_accuracy": 0.1808542713522911,
      "num_tokens": 58505386.0,
      "step": 23080
    },
    {
      "entropy": 6.160248756408691,
      "epoch": 2.6641661858049623,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00043013251972378404,
      "loss": 5.8299,
      "mean_token_accuracy": 0.18217852115631103,
      "num_tokens": 58518067.0,
      "step": 23085
    },
    {
      "entropy": 6.190211963653565,
      "epoch": 2.664743219849971,
      "grad_norm": 0.875,
      "learning_rate": 0.00043010262906633355,
      "loss": 5.8739,
      "mean_token_accuracy": 0.17838504314422607,
      "num_tokens": 58532138.0,
      "step": 23090
    },
    {
      "entropy": 6.195870304107666,
      "epoch": 2.66532025389498,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043007273319208326,
      "loss": 5.89,
      "mean_token_accuracy": 0.18403186947107314,
      "num_tokens": 58546464.0,
      "step": 23095
    },
    {
      "entropy": 6.2662766456604,
      "epoch": 2.6658972879399885,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004300428321020385,
      "loss": 5.8947,
      "mean_token_accuracy": 0.17792199552059174,
      "num_tokens": 58559463.0,
      "step": 23100
    },
    {
      "entropy": 6.2804466724395756,
      "epoch": 2.666474321984997,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00043001292579720514,
      "loss": 5.8947,
      "mean_token_accuracy": 0.18157547563314438,
      "num_tokens": 58572125.0,
      "step": 23105
    },
    {
      "entropy": 6.106372451782226,
      "epoch": 2.6670513560300058,
      "grad_norm": 2.296875,
      "learning_rate": 0.000429983014278589,
      "loss": 5.7776,
      "mean_token_accuracy": 0.18735108077526091,
      "num_tokens": 58585312.0,
      "step": 23110
    },
    {
      "entropy": 6.1410932540893555,
      "epoch": 2.6676283900750146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004299530975471962,
      "loss": 5.7834,
      "mean_token_accuracy": 0.1863879904150963,
      "num_tokens": 58598006.0,
      "step": 23115
    },
    {
      "entropy": 6.153367948532105,
      "epoch": 2.668205424120023,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004299231756040329,
      "loss": 5.7685,
      "mean_token_accuracy": 0.18700756430625914,
      "num_tokens": 58609741.0,
      "step": 23120
    },
    {
      "entropy": 6.168306684494018,
      "epoch": 2.668782458165032,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004298932484501055,
      "loss": 5.8204,
      "mean_token_accuracy": 0.18301159292459487,
      "num_tokens": 58623076.0,
      "step": 23125
    },
    {
      "entropy": 6.227231359481811,
      "epoch": 2.6693594922100403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042986331608642076,
      "loss": 5.8543,
      "mean_token_accuracy": 0.1850983902812004,
      "num_tokens": 58634558.0,
      "step": 23130
    },
    {
      "entropy": 6.062982988357544,
      "epoch": 2.669936526255049,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042983337851398525,
      "loss": 5.7634,
      "mean_token_accuracy": 0.1951274588704109,
      "num_tokens": 58647305.0,
      "step": 23135
    },
    {
      "entropy": 6.174078559875488,
      "epoch": 2.6705135603000576,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042980343573380595,
      "loss": 5.8479,
      "mean_token_accuracy": 0.18469816595315933,
      "num_tokens": 58659723.0,
      "step": 23140
    },
    {
      "entropy": 6.139213705062867,
      "epoch": 2.6710905943450665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042977348774688996,
      "loss": 5.8854,
      "mean_token_accuracy": 0.18793033212423324,
      "num_tokens": 58670932.0,
      "step": 23145
    },
    {
      "entropy": 6.221418571472168,
      "epoch": 2.671667628390075,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004297435345542445,
      "loss": 5.852,
      "mean_token_accuracy": 0.17453705370426179,
      "num_tokens": 58682337.0,
      "step": 23150
    },
    {
      "entropy": 6.268413162231445,
      "epoch": 2.6722446624350837,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004297135761568773,
      "loss": 5.8087,
      "mean_token_accuracy": 0.18703240305185317,
      "num_tokens": 58694667.0,
      "step": 23155
    },
    {
      "entropy": 6.147421216964721,
      "epoch": 2.672821696480092,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042968361255579573,
      "loss": 5.8385,
      "mean_token_accuracy": 0.18575343489646912,
      "num_tokens": 58707478.0,
      "step": 23160
    },
    {
      "entropy": 6.187603569030761,
      "epoch": 2.673398730525101,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042965364375200765,
      "loss": 5.8071,
      "mean_token_accuracy": 0.18542922735214235,
      "num_tokens": 58720660.0,
      "step": 23165
    },
    {
      "entropy": 6.148350715637207,
      "epoch": 2.67397576457011,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004296236697465212,
      "loss": 5.7836,
      "mean_token_accuracy": 0.18650804907083512,
      "num_tokens": 58732835.0,
      "step": 23170
    },
    {
      "entropy": 6.130046653747558,
      "epoch": 2.6745527986151183,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004295936905403444,
      "loss": 5.8191,
      "mean_token_accuracy": 0.1780240699648857,
      "num_tokens": 58745432.0,
      "step": 23175
    },
    {
      "entropy": 6.156208848953247,
      "epoch": 2.6751298326601267,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004295637061344855,
      "loss": 5.7747,
      "mean_token_accuracy": 0.18617688864469528,
      "num_tokens": 58758837.0,
      "step": 23180
    },
    {
      "entropy": 6.129476070404053,
      "epoch": 2.6757068667051356,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004295337165299532,
      "loss": 5.7591,
      "mean_token_accuracy": 0.1869099661707878,
      "num_tokens": 58772347.0,
      "step": 23185
    },
    {
      "entropy": 6.170158958435058,
      "epoch": 2.6762839007501444,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042950372172775614,
      "loss": 5.8263,
      "mean_token_accuracy": 0.1768513023853302,
      "num_tokens": 58784278.0,
      "step": 23190
    },
    {
      "entropy": 6.151766395568847,
      "epoch": 2.676860934795153,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000429473721728903,
      "loss": 5.7627,
      "mean_token_accuracy": 0.18710385710000993,
      "num_tokens": 58795788.0,
      "step": 23195
    },
    {
      "entropy": 6.208158493041992,
      "epoch": 2.6774379688401617,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004294437165344031,
      "loss": 5.8631,
      "mean_token_accuracy": 0.17992291152477263,
      "num_tokens": 58808394.0,
      "step": 23200
    },
    {
      "entropy": 6.161670732498169,
      "epoch": 2.67801500288517,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042941370614526554,
      "loss": 5.7899,
      "mean_token_accuracy": 0.19086406975984574,
      "num_tokens": 58819570.0,
      "step": 23205
    },
    {
      "entropy": 6.124743986129761,
      "epoch": 2.678592036930179,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004293836905624996,
      "loss": 5.843,
      "mean_token_accuracy": 0.1866921991109848,
      "num_tokens": 58831049.0,
      "step": 23210
    },
    {
      "entropy": 6.289630317687989,
      "epoch": 2.6791690709751874,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042935366978711487,
      "loss": 5.951,
      "mean_token_accuracy": 0.1732814535498619,
      "num_tokens": 58844753.0,
      "step": 23215
    },
    {
      "entropy": 6.1924584865570065,
      "epoch": 2.6797461050201963,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042932364382012125,
      "loss": 5.828,
      "mean_token_accuracy": 0.1831527054309845,
      "num_tokens": 58856796.0,
      "step": 23220
    },
    {
      "entropy": 6.21537504196167,
      "epoch": 2.6803231390652047,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042929361266252843,
      "loss": 5.8528,
      "mean_token_accuracy": 0.1887661799788475,
      "num_tokens": 58869046.0,
      "step": 23225
    },
    {
      "entropy": 6.167904663085937,
      "epoch": 2.6809001731102136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004292635763153466,
      "loss": 5.8335,
      "mean_token_accuracy": 0.18496952652931214,
      "num_tokens": 58880993.0,
      "step": 23230
    },
    {
      "entropy": 6.216453933715821,
      "epoch": 2.681477207155222,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042923353477958607,
      "loss": 5.814,
      "mean_token_accuracy": 0.1836557060480118,
      "num_tokens": 58893519.0,
      "step": 23235
    },
    {
      "entropy": 6.17032527923584,
      "epoch": 2.682054241200231,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042920348805625716,
      "loss": 5.8219,
      "mean_token_accuracy": 0.18259599655866623,
      "num_tokens": 58905904.0,
      "step": 23240
    },
    {
      "entropy": 6.185873937606812,
      "epoch": 2.6826312752452397,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042917343614637057,
      "loss": 5.85,
      "mean_token_accuracy": 0.1853163793683052,
      "num_tokens": 58918271.0,
      "step": 23245
    },
    {
      "entropy": 6.196112489700317,
      "epoch": 2.683208309290248,
      "grad_norm": 1.265625,
      "learning_rate": 0.00042914337905093695,
      "loss": 5.8383,
      "mean_token_accuracy": 0.1808432936668396,
      "num_tokens": 58930381.0,
      "step": 23250
    },
    {
      "entropy": 6.197277450561524,
      "epoch": 2.6837853433352565,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004291133167709674,
      "loss": 5.7465,
      "mean_token_accuracy": 0.18375710546970367,
      "num_tokens": 58942506.0,
      "step": 23255
    },
    {
      "entropy": 6.129807233810425,
      "epoch": 2.6843623773802654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042908324930747297,
      "loss": 5.8696,
      "mean_token_accuracy": 0.18315427452325822,
      "num_tokens": 58954383.0,
      "step": 23260
    },
    {
      "entropy": 6.197905158996582,
      "epoch": 2.6849394114252743,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042905317666146496,
      "loss": 5.948,
      "mean_token_accuracy": 0.17622523307800292,
      "num_tokens": 58967110.0,
      "step": 23265
    },
    {
      "entropy": 6.260257434844971,
      "epoch": 2.6855164454702827,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00042902309883395497,
      "loss": 5.8235,
      "mean_token_accuracy": 0.18633091747760772,
      "num_tokens": 58980314.0,
      "step": 23270
    },
    {
      "entropy": 6.18588695526123,
      "epoch": 2.6860934795152915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004289930158259545,
      "loss": 5.9316,
      "mean_token_accuracy": 0.17162706702947617,
      "num_tokens": 58993544.0,
      "step": 23275
    },
    {
      "entropy": 6.2314550399780275,
      "epoch": 2.6866705135603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00042896292763847544,
      "loss": 5.8778,
      "mean_token_accuracy": 0.19867659658193587,
      "num_tokens": 59006043.0,
      "step": 23280
    },
    {
      "entropy": 6.270898771286011,
      "epoch": 2.687247547605309,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042893283427252984,
      "loss": 5.8935,
      "mean_token_accuracy": 0.17977669984102249,
      "num_tokens": 59018691.0,
      "step": 23285
    },
    {
      "entropy": 6.188410377502441,
      "epoch": 2.6878245816503172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004289027357291297,
      "loss": 5.8397,
      "mean_token_accuracy": 0.1844502329826355,
      "num_tokens": 59031493.0,
      "step": 23290
    },
    {
      "entropy": 6.071956205368042,
      "epoch": 2.688401615695326,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004288726320092876,
      "loss": 5.722,
      "mean_token_accuracy": 0.1941444382071495,
      "num_tokens": 59045464.0,
      "step": 23295
    },
    {
      "entropy": 6.251468181610107,
      "epoch": 2.6889786497403345,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004288425231140159,
      "loss": 5.823,
      "mean_token_accuracy": 0.18416701704263688,
      "num_tokens": 59057391.0,
      "step": 23300
    },
    {
      "entropy": 6.070117807388305,
      "epoch": 2.6895556837853434,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042881240904432737,
      "loss": 5.6905,
      "mean_token_accuracy": 0.19433396309614182,
      "num_tokens": 59070175.0,
      "step": 23305
    },
    {
      "entropy": 6.149970483779907,
      "epoch": 2.690132717830352,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004287822898012348,
      "loss": 5.8407,
      "mean_token_accuracy": 0.185240176320076,
      "num_tokens": 59083605.0,
      "step": 23310
    },
    {
      "entropy": 6.134791517257691,
      "epoch": 2.6907097518753607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004287521653857514,
      "loss": 5.7687,
      "mean_token_accuracy": 0.19423387795686722,
      "num_tokens": 59095799.0,
      "step": 23315
    },
    {
      "entropy": 6.160860347747803,
      "epoch": 2.6912867859203695,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042872203579889025,
      "loss": 5.8445,
      "mean_token_accuracy": 0.185671566426754,
      "num_tokens": 59109982.0,
      "step": 23320
    },
    {
      "entropy": 6.205086088180542,
      "epoch": 2.691863819965378,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042869190104166476,
      "loss": 5.8742,
      "mean_token_accuracy": 0.1778896704316139,
      "num_tokens": 59122478.0,
      "step": 23325
    },
    {
      "entropy": 6.201913833618164,
      "epoch": 2.6924408540103864,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004286617611150886,
      "loss": 5.8548,
      "mean_token_accuracy": 0.18144448697566987,
      "num_tokens": 59134342.0,
      "step": 23330
    },
    {
      "entropy": 6.270397281646728,
      "epoch": 2.693017888055395,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042863161602017536,
      "loss": 5.9195,
      "mean_token_accuracy": 0.17555266469717026,
      "num_tokens": 59146628.0,
      "step": 23335
    },
    {
      "entropy": 6.141443252563477,
      "epoch": 2.693594922100404,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004286014657579391,
      "loss": 5.7835,
      "mean_token_accuracy": 0.1902693510055542,
      "num_tokens": 59160043.0,
      "step": 23340
    },
    {
      "entropy": 6.2056488513946535,
      "epoch": 2.6941719561454125,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004285713103293938,
      "loss": 5.8536,
      "mean_token_accuracy": 0.17837955802679062,
      "num_tokens": 59173710.0,
      "step": 23345
    },
    {
      "entropy": 6.128909587860107,
      "epoch": 2.6947489901904214,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004285411497355537,
      "loss": 5.7606,
      "mean_token_accuracy": 0.18160916566848756,
      "num_tokens": 59186069.0,
      "step": 23350
    },
    {
      "entropy": 6.173884010314941,
      "epoch": 2.6953260242354298,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042851098397743334,
      "loss": 5.8683,
      "mean_token_accuracy": 0.18132909536361694,
      "num_tokens": 59199305.0,
      "step": 23355
    },
    {
      "entropy": 6.2042299747467045,
      "epoch": 2.6959030582804386,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004284808130560473,
      "loss": 5.8095,
      "mean_token_accuracy": 0.18014252334833145,
      "num_tokens": 59212135.0,
      "step": 23360
    },
    {
      "entropy": 6.250422477722168,
      "epoch": 2.696480092325447,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042845063697241034,
      "loss": 5.8693,
      "mean_token_accuracy": 0.17942632585763932,
      "num_tokens": 59226651.0,
      "step": 23365
    },
    {
      "entropy": 6.111573362350464,
      "epoch": 2.697057126370456,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004284204557275374,
      "loss": 5.8208,
      "mean_token_accuracy": 0.17859306782484055,
      "num_tokens": 59238630.0,
      "step": 23370
    },
    {
      "entropy": 6.157173442840576,
      "epoch": 2.6976341604154648,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004283902693224437,
      "loss": 5.825,
      "mean_token_accuracy": 0.18013526797294616,
      "num_tokens": 59251465.0,
      "step": 23375
    },
    {
      "entropy": 6.197487735748291,
      "epoch": 2.698211194460473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042836007775814446,
      "loss": 5.7956,
      "mean_token_accuracy": 0.19154615551233292,
      "num_tokens": 59264058.0,
      "step": 23380
    },
    {
      "entropy": 6.117607116699219,
      "epoch": 2.6987882285054816,
      "grad_norm": 1.125,
      "learning_rate": 0.0004283298810356551,
      "loss": 5.7133,
      "mean_token_accuracy": 0.19167850017547608,
      "num_tokens": 59276937.0,
      "step": 23385
    },
    {
      "entropy": 6.117186164855957,
      "epoch": 2.6993652625504905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004282996791559914,
      "loss": 5.8164,
      "mean_token_accuracy": 0.18485463559627532,
      "num_tokens": 59288747.0,
      "step": 23390
    },
    {
      "entropy": 6.153876781463623,
      "epoch": 2.6999422965954993,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042826947212016924,
      "loss": 5.8205,
      "mean_token_accuracy": 0.18580360263586043,
      "num_tokens": 59301690.0,
      "step": 23395
    },
    {
      "entropy": 6.133680391311645,
      "epoch": 2.7005193306405078,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004282392599292045,
      "loss": 5.7532,
      "mean_token_accuracy": 0.18659163266420364,
      "num_tokens": 59312964.0,
      "step": 23400
    },
    {
      "entropy": 6.234754753112793,
      "epoch": 2.701096364685516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004282090425841133,
      "loss": 5.9417,
      "mean_token_accuracy": 0.17786754071712493,
      "num_tokens": 59325242.0,
      "step": 23405
    },
    {
      "entropy": 6.179014682769775,
      "epoch": 2.701673398730525,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004281788200859121,
      "loss": 5.8338,
      "mean_token_accuracy": 0.1878078043460846,
      "num_tokens": 59337877.0,
      "step": 23410
    },
    {
      "entropy": 6.206895065307617,
      "epoch": 2.702250432775534,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042814859243561727,
      "loss": 5.8435,
      "mean_token_accuracy": 0.18348279148340224,
      "num_tokens": 59349757.0,
      "step": 23415
    },
    {
      "entropy": 6.163533639907837,
      "epoch": 2.7028274668205423,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042811835963424573,
      "loss": 5.7973,
      "mean_token_accuracy": 0.186604243516922,
      "num_tokens": 59364268.0,
      "step": 23420
    },
    {
      "entropy": 6.209364223480224,
      "epoch": 2.703404500865551,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042808812168281413,
      "loss": 5.8853,
      "mean_token_accuracy": 0.1777723625302315,
      "num_tokens": 59376896.0,
      "step": 23425
    },
    {
      "entropy": 6.226865530014038,
      "epoch": 2.7039815349105596,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042805787858233966,
      "loss": 5.8765,
      "mean_token_accuracy": 0.17968610525131226,
      "num_tokens": 59390905.0,
      "step": 23430
    },
    {
      "entropy": 6.150434446334839,
      "epoch": 2.7045585689555685,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004280276303338395,
      "loss": 5.7907,
      "mean_token_accuracy": 0.19126541763544083,
      "num_tokens": 59403896.0,
      "step": 23435
    },
    {
      "entropy": 6.1717979431152346,
      "epoch": 2.705135603000577,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000427997376938331,
      "loss": 5.7485,
      "mean_token_accuracy": 0.18864676505327224,
      "num_tokens": 59416050.0,
      "step": 23440
    },
    {
      "entropy": 6.150090980529785,
      "epoch": 2.7057126370455857,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004279671183968317,
      "loss": 5.8254,
      "mean_token_accuracy": 0.1847640573978424,
      "num_tokens": 59428127.0,
      "step": 23445
    },
    {
      "entropy": 6.165274810791016,
      "epoch": 2.7062896710905946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004279368547103593,
      "loss": 5.8112,
      "mean_token_accuracy": 0.18552787899971007,
      "num_tokens": 59442708.0,
      "step": 23450
    },
    {
      "entropy": 6.168972063064575,
      "epoch": 2.706866705135603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004279065858799318,
      "loss": 5.8875,
      "mean_token_accuracy": 0.18229120522737502,
      "num_tokens": 59456126.0,
      "step": 23455
    },
    {
      "entropy": 6.168537998199463,
      "epoch": 2.7074437391806114,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042787631190656725,
      "loss": 5.83,
      "mean_token_accuracy": 0.1861049562692642,
      "num_tokens": 59469331.0,
      "step": 23460
    },
    {
      "entropy": 6.075778579711914,
      "epoch": 2.7080207732256203,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00042784603279128386,
      "loss": 5.7633,
      "mean_token_accuracy": 0.19445625245571135,
      "num_tokens": 59483084.0,
      "step": 23465
    },
    {
      "entropy": 6.192923355102539,
      "epoch": 2.708597807270629,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004278157485351001,
      "loss": 5.8616,
      "mean_token_accuracy": 0.17950499057769775,
      "num_tokens": 59494746.0,
      "step": 23470
    },
    {
      "entropy": 6.203206491470337,
      "epoch": 2.7091748413156376,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042778545913903454,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18603405952453614,
      "num_tokens": 59507266.0,
      "step": 23475
    },
    {
      "entropy": 6.229127359390259,
      "epoch": 2.7097518753606464,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042775516460410594,
      "loss": 5.8741,
      "mean_token_accuracy": 0.1794537678360939,
      "num_tokens": 59521876.0,
      "step": 23480
    },
    {
      "entropy": 6.201447010040283,
      "epoch": 2.710328909405655,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004277248649313333,
      "loss": 5.8491,
      "mean_token_accuracy": 0.18010489642620087,
      "num_tokens": 59535512.0,
      "step": 23485
    },
    {
      "entropy": 6.195682144165039,
      "epoch": 2.7109059434506637,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004276945601217357,
      "loss": 5.9564,
      "mean_token_accuracy": 0.17316356003284455,
      "num_tokens": 59549596.0,
      "step": 23490
    },
    {
      "entropy": 6.20859842300415,
      "epoch": 2.711482977495672,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042766425017633236,
      "loss": 5.8888,
      "mean_token_accuracy": 0.18614009022712708,
      "num_tokens": 59562455.0,
      "step": 23495
    },
    {
      "entropy": 6.136179733276367,
      "epoch": 2.712060011540681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042763393509614284,
      "loss": 5.773,
      "mean_token_accuracy": 0.18201442062854767,
      "num_tokens": 59574897.0,
      "step": 23500
    },
    {
      "entropy": 6.198714685440064,
      "epoch": 2.7126370455856894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004276036148821867,
      "loss": 5.7977,
      "mean_token_accuracy": 0.18542479425668718,
      "num_tokens": 59586472.0,
      "step": 23505
    },
    {
      "entropy": 6.178706121444702,
      "epoch": 2.7132140796306983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042757328953548385,
      "loss": 5.7969,
      "mean_token_accuracy": 0.1838708609342575,
      "num_tokens": 59597788.0,
      "step": 23510
    },
    {
      "entropy": 6.116839790344239,
      "epoch": 2.7137911136757067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004275429590570541,
      "loss": 5.819,
      "mean_token_accuracy": 0.17778913378715516,
      "num_tokens": 59611045.0,
      "step": 23515
    },
    {
      "entropy": 6.062789058685302,
      "epoch": 2.7143681477207156,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00042751262344791776,
      "loss": 5.6316,
      "mean_token_accuracy": 0.1953550472855568,
      "num_tokens": 59624251.0,
      "step": 23520
    },
    {
      "entropy": 6.168734741210938,
      "epoch": 2.7149451817657244,
      "grad_norm": 0.96875,
      "learning_rate": 0.000427482282709095,
      "loss": 5.769,
      "mean_token_accuracy": 0.18882142454385759,
      "num_tokens": 59635604.0,
      "step": 23525
    },
    {
      "entropy": 6.176315593719482,
      "epoch": 2.715522215810733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004274519368416065,
      "loss": 5.764,
      "mean_token_accuracy": 0.18330251425504684,
      "num_tokens": 59647591.0,
      "step": 23530
    },
    {
      "entropy": 6.176680040359497,
      "epoch": 2.7160992498557412,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004274215858464727,
      "loss": 5.8774,
      "mean_token_accuracy": 0.17968627959489822,
      "num_tokens": 59660693.0,
      "step": 23535
    },
    {
      "entropy": 6.156770038604736,
      "epoch": 2.71667628390075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042739122972471473,
      "loss": 5.7895,
      "mean_token_accuracy": 0.1843475267291069,
      "num_tokens": 59673792.0,
      "step": 23540
    },
    {
      "entropy": 6.001822853088379,
      "epoch": 2.717253317945759,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004273608684773534,
      "loss": 5.6485,
      "mean_token_accuracy": 0.20021015107631684,
      "num_tokens": 59687415.0,
      "step": 23545
    },
    {
      "entropy": 6.165789175033569,
      "epoch": 2.7178303519907674,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042733050210540984,
      "loss": 5.7931,
      "mean_token_accuracy": 0.1891214907169342,
      "num_tokens": 59699721.0,
      "step": 23550
    },
    {
      "entropy": 6.175017166137695,
      "epoch": 2.7184073860357763,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004273001306099056,
      "loss": 5.814,
      "mean_token_accuracy": 0.19010271430015563,
      "num_tokens": 59712517.0,
      "step": 23555
    },
    {
      "entropy": 6.080675554275513,
      "epoch": 2.7189844200807847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004272697539918621,
      "loss": 5.7821,
      "mean_token_accuracy": 0.19143579602241517,
      "num_tokens": 59725132.0,
      "step": 23560
    },
    {
      "entropy": 6.092270803451538,
      "epoch": 2.7195614541257935,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004272393722523011,
      "loss": 5.8172,
      "mean_token_accuracy": 0.18737081438302994,
      "num_tokens": 59738515.0,
      "step": 23565
    },
    {
      "entropy": 6.146021604537964,
      "epoch": 2.720138488170802,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042720898539224436,
      "loss": 5.7586,
      "mean_token_accuracy": 0.1918032720685005,
      "num_tokens": 59750141.0,
      "step": 23570
    },
    {
      "entropy": 6.23243556022644,
      "epoch": 2.720715522215811,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004271785934127141,
      "loss": 6.0125,
      "mean_token_accuracy": 0.17802267223596574,
      "num_tokens": 59762580.0,
      "step": 23575
    },
    {
      "entropy": 6.210177707672119,
      "epoch": 2.7212925562608192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042714819631473234,
      "loss": 5.8767,
      "mean_token_accuracy": 0.17955715209245682,
      "num_tokens": 59776039.0,
      "step": 23580
    },
    {
      "entropy": 6.204678726196289,
      "epoch": 2.721869590305828,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042711779409932164,
      "loss": 5.8523,
      "mean_token_accuracy": 0.18440286666154862,
      "num_tokens": 59789305.0,
      "step": 23585
    },
    {
      "entropy": 6.169595003128052,
      "epoch": 2.7224466243508365,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042708738676750446,
      "loss": 5.7659,
      "mean_token_accuracy": 0.19036545902490615,
      "num_tokens": 59801771.0,
      "step": 23590
    },
    {
      "entropy": 6.230128431320191,
      "epoch": 2.7230236583958454,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004270569743203036,
      "loss": 5.8152,
      "mean_token_accuracy": 0.18158550709486007,
      "num_tokens": 59815281.0,
      "step": 23595
    },
    {
      "entropy": 6.185557126998901,
      "epoch": 2.7236006924408542,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004270265567587419,
      "loss": 5.8143,
      "mean_token_accuracy": 0.18081941306591034,
      "num_tokens": 59827292.0,
      "step": 23600
    },
    {
      "entropy": 6.1336852550506595,
      "epoch": 2.7241777264858626,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042699613408384246,
      "loss": 5.7825,
      "mean_token_accuracy": 0.18778499513864516,
      "num_tokens": 59840024.0,
      "step": 23605
    },
    {
      "entropy": 6.224751758575439,
      "epoch": 2.724754760530871,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004269657062966286,
      "loss": 5.8772,
      "mean_token_accuracy": 0.17981754541397094,
      "num_tokens": 59851499.0,
      "step": 23610
    },
    {
      "entropy": 6.124369859695435,
      "epoch": 2.72533179457588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004269352733981238,
      "loss": 5.7821,
      "mean_token_accuracy": 0.18569391816854477,
      "num_tokens": 59863057.0,
      "step": 23615
    },
    {
      "entropy": 6.123807907104492,
      "epoch": 2.725908828620889,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042690483538935137,
      "loss": 5.7824,
      "mean_token_accuracy": 0.18660981357097625,
      "num_tokens": 59876132.0,
      "step": 23620
    },
    {
      "entropy": 6.167468118667602,
      "epoch": 2.726485862665897,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004268743922713354,
      "loss": 5.8265,
      "mean_token_accuracy": 0.1824280649423599,
      "num_tokens": 59888543.0,
      "step": 23625
    },
    {
      "entropy": 6.211517953872681,
      "epoch": 2.727062896710906,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042684394404509957,
      "loss": 5.7933,
      "mean_token_accuracy": 0.18391778320074081,
      "num_tokens": 59900508.0,
      "step": 23630
    },
    {
      "entropy": 6.127938079833984,
      "epoch": 2.7276399307559145,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042681349071166814,
      "loss": 5.8312,
      "mean_token_accuracy": 0.18280233144760133,
      "num_tokens": 59913065.0,
      "step": 23635
    },
    {
      "entropy": 6.210356664657593,
      "epoch": 2.7282169648009233,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042678303227206536,
      "loss": 5.8787,
      "mean_token_accuracy": 0.1786747008562088,
      "num_tokens": 59926173.0,
      "step": 23640
    },
    {
      "entropy": 6.241761875152588,
      "epoch": 2.7287939988459318,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004267525687273157,
      "loss": 5.8811,
      "mean_token_accuracy": 0.1799023285508156,
      "num_tokens": 59939206.0,
      "step": 23645
    },
    {
      "entropy": 6.265921354293823,
      "epoch": 2.7293710328909406,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042672210007844383,
      "loss": 5.8631,
      "mean_token_accuracy": 0.18714073896408082,
      "num_tokens": 59951093.0,
      "step": 23650
    },
    {
      "entropy": 6.129597568511963,
      "epoch": 2.7299480669359495,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00042669162632647434,
      "loss": 5.8481,
      "mean_token_accuracy": 0.17962536364793777,
      "num_tokens": 59962804.0,
      "step": 23655
    },
    {
      "entropy": 6.179275226593018,
      "epoch": 2.730525100980958,
      "grad_norm": 1.28125,
      "learning_rate": 0.00042666114747243243,
      "loss": 5.8492,
      "mean_token_accuracy": 0.17971125096082688,
      "num_tokens": 59975649.0,
      "step": 23660
    },
    {
      "entropy": 6.15836820602417,
      "epoch": 2.7311021350259663,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004266306635173432,
      "loss": 5.8096,
      "mean_token_accuracy": 0.18173106014728546,
      "num_tokens": 59989496.0,
      "step": 23665
    },
    {
      "entropy": 6.243234539031983,
      "epoch": 2.731679169070975,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004266001744622319,
      "loss": 5.9029,
      "mean_token_accuracy": 0.18467528074979783,
      "num_tokens": 60002125.0,
      "step": 23670
    },
    {
      "entropy": 6.198999881744385,
      "epoch": 2.732256203115984,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000426569680308124,
      "loss": 5.8536,
      "mean_token_accuracy": 0.1782796248793602,
      "num_tokens": 60013395.0,
      "step": 23675
    },
    {
      "entropy": 6.21286907196045,
      "epoch": 2.7328332371609925,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042653918105604524,
      "loss": 5.8317,
      "mean_token_accuracy": 0.1833585724234581,
      "num_tokens": 60026527.0,
      "step": 23680
    },
    {
      "entropy": 6.197463417053223,
      "epoch": 2.733410271206001,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042650867670702123,
      "loss": 5.905,
      "mean_token_accuracy": 0.18060424476861953,
      "num_tokens": 60039671.0,
      "step": 23685
    },
    {
      "entropy": 6.147560119628906,
      "epoch": 2.7339873052510097,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004264781672620783,
      "loss": 5.8248,
      "mean_token_accuracy": 0.18963334411382676,
      "num_tokens": 60051908.0,
      "step": 23690
    },
    {
      "entropy": 6.218137741088867,
      "epoch": 2.7345643392960186,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004264476527222425,
      "loss": 5.8712,
      "mean_token_accuracy": 0.18271903246641158,
      "num_tokens": 60065295.0,
      "step": 23695
    },
    {
      "entropy": 6.164173746109009,
      "epoch": 2.735141373341027,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004264171330885401,
      "loss": 5.7658,
      "mean_token_accuracy": 0.18747562766075135,
      "num_tokens": 60076903.0,
      "step": 23700
    },
    {
      "entropy": 6.170087385177612,
      "epoch": 2.735718407386036,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042638660836199756,
      "loss": 5.8227,
      "mean_token_accuracy": 0.175394731760025,
      "num_tokens": 60089091.0,
      "step": 23705
    },
    {
      "entropy": 6.17940125465393,
      "epoch": 2.7362954414310443,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00042635607854364166,
      "loss": 5.7276,
      "mean_token_accuracy": 0.18890163749456407,
      "num_tokens": 60100986.0,
      "step": 23710
    },
    {
      "entropy": 6.069400596618652,
      "epoch": 2.736872475476053,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042632554363449926,
      "loss": 5.7463,
      "mean_token_accuracy": 0.1896691083908081,
      "num_tokens": 60113806.0,
      "step": 23715
    },
    {
      "entropy": 6.206710290908814,
      "epoch": 2.7374495095210616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042629500363559734,
      "loss": 5.8174,
      "mean_token_accuracy": 0.1911667987704277,
      "num_tokens": 60126076.0,
      "step": 23720
    },
    {
      "entropy": 6.235867977142334,
      "epoch": 2.7380265435660704,
      "grad_norm": 0.90625,
      "learning_rate": 0.00042626445854796326,
      "loss": 5.896,
      "mean_token_accuracy": 0.18047737032175065,
      "num_tokens": 60139435.0,
      "step": 23725
    },
    {
      "entropy": 6.174714612960815,
      "epoch": 2.7386035776110793,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004262339083726241,
      "loss": 5.8573,
      "mean_token_accuracy": 0.17811036556959153,
      "num_tokens": 60152252.0,
      "step": 23730
    },
    {
      "entropy": 6.1112288475036625,
      "epoch": 2.7391806116560877,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004262033531106076,
      "loss": 5.7212,
      "mean_token_accuracy": 0.18555221408605577,
      "num_tokens": 60164575.0,
      "step": 23735
    },
    {
      "entropy": 6.204973840713501,
      "epoch": 2.739757645701096,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042617279276294145,
      "loss": 5.8614,
      "mean_token_accuracy": 0.17902258932590484,
      "num_tokens": 60177085.0,
      "step": 23740
    },
    {
      "entropy": 6.185987663269043,
      "epoch": 2.740334679746105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261422273306535,
      "loss": 5.787,
      "mean_token_accuracy": 0.18618640303611755,
      "num_tokens": 60188923.0,
      "step": 23745
    },
    {
      "entropy": 6.16752028465271,
      "epoch": 2.740911713791114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004261116568147718,
      "loss": 5.8181,
      "mean_token_accuracy": 0.18826280981302262,
      "num_tokens": 60200439.0,
      "step": 23750
    },
    {
      "entropy": 6.118758583068848,
      "epoch": 2.7414887478361223,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004260810812163246,
      "loss": 5.7694,
      "mean_token_accuracy": 0.19570343494415282,
      "num_tokens": 60213618.0,
      "step": 23755
    },
    {
      "entropy": 6.246321058273315,
      "epoch": 2.742065781881131,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042605050053634027,
      "loss": 5.851,
      "mean_token_accuracy": 0.18420463651418686,
      "num_tokens": 60225617.0,
      "step": 23760
    },
    {
      "entropy": 6.147588872909546,
      "epoch": 2.7426428159261396,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004260199147758474,
      "loss": 5.7904,
      "mean_token_accuracy": 0.18074362277984618,
      "num_tokens": 60238406.0,
      "step": 23765
    },
    {
      "entropy": 6.154901933670044,
      "epoch": 2.7432198499711484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004259893239358747,
      "loss": 5.8856,
      "mean_token_accuracy": 0.17359059005975724,
      "num_tokens": 60250733.0,
      "step": 23770
    },
    {
      "entropy": 6.17216010093689,
      "epoch": 2.743796884016157,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042595872801745106,
      "loss": 5.7994,
      "mean_token_accuracy": 0.18428043723106385,
      "num_tokens": 60262625.0,
      "step": 23775
    },
    {
      "entropy": 6.229062128067016,
      "epoch": 2.7443739180611657,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004259281270216056,
      "loss": 5.9112,
      "mean_token_accuracy": 0.18263706117868422,
      "num_tokens": 60274428.0,
      "step": 23780
    },
    {
      "entropy": 6.202558994293213,
      "epoch": 2.744950952106174,
      "grad_norm": 1.1875,
      "learning_rate": 0.00042589752094936763,
      "loss": 5.9094,
      "mean_token_accuracy": 0.17718964219093322,
      "num_tokens": 60286131.0,
      "step": 23785
    },
    {
      "entropy": 6.151939344406128,
      "epoch": 2.745527986151183,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004258669098017664,
      "loss": 5.7916,
      "mean_token_accuracy": 0.18867217451334,
      "num_tokens": 60299151.0,
      "step": 23790
    },
    {
      "entropy": 6.088631534576416,
      "epoch": 2.7461050201961914,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042583629357983164,
      "loss": 5.7308,
      "mean_token_accuracy": 0.1825684502720833,
      "num_tokens": 60311557.0,
      "step": 23795
    },
    {
      "entropy": 6.234161376953125,
      "epoch": 2.7466820542412003,
      "grad_norm": 1.359375,
      "learning_rate": 0.00042580567228459303,
      "loss": 5.8852,
      "mean_token_accuracy": 0.17799129337072372,
      "num_tokens": 60324364.0,
      "step": 23800
    },
    {
      "entropy": 6.242687606811524,
      "epoch": 2.747259088286209,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004257750459170806,
      "loss": 5.8751,
      "mean_token_accuracy": 0.18489039391279222,
      "num_tokens": 60336141.0,
      "step": 23805
    },
    {
      "entropy": 6.207920408248901,
      "epoch": 2.7478361223312175,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042574441447832436,
      "loss": 5.866,
      "mean_token_accuracy": 0.18113742619752884,
      "num_tokens": 60349006.0,
      "step": 23810
    },
    {
      "entropy": 6.169639921188354,
      "epoch": 2.748413156376226,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004257137779693546,
      "loss": 5.8399,
      "mean_token_accuracy": 0.18505763113498688,
      "num_tokens": 60361361.0,
      "step": 23815
    },
    {
      "entropy": 6.1327532768249515,
      "epoch": 2.748990190421235,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004256831363912018,
      "loss": 5.8371,
      "mean_token_accuracy": 0.1820719614624977,
      "num_tokens": 60375261.0,
      "step": 23820
    },
    {
      "entropy": 6.17534327507019,
      "epoch": 2.7495672244662437,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042565248974489654,
      "loss": 5.738,
      "mean_token_accuracy": 0.18637765496969222,
      "num_tokens": 60387129.0,
      "step": 23825
    },
    {
      "entropy": 6.1087689876556395,
      "epoch": 2.750144258511252,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004256218380314697,
      "loss": 5.7644,
      "mean_token_accuracy": 0.19168006926774978,
      "num_tokens": 60398581.0,
      "step": 23830
    },
    {
      "entropy": 6.216979026794434,
      "epoch": 2.750721292556261,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004255911812519521,
      "loss": 5.867,
      "mean_token_accuracy": 0.18237381279468537,
      "num_tokens": 60412105.0,
      "step": 23835
    },
    {
      "entropy": 6.239885520935059,
      "epoch": 2.7512983266012694,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004255605194073749,
      "loss": 5.7899,
      "mean_token_accuracy": 0.18544149100780488,
      "num_tokens": 60423880.0,
      "step": 23840
    },
    {
      "entropy": 6.188281583786011,
      "epoch": 2.7518753606462782,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004255298524987695,
      "loss": 5.8774,
      "mean_token_accuracy": 0.17875241339206696,
      "num_tokens": 60435968.0,
      "step": 23845
    },
    {
      "entropy": 6.1402302265167235,
      "epoch": 2.7524523946912867,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004254991805271672,
      "loss": 5.8443,
      "mean_token_accuracy": 0.17654843181371688,
      "num_tokens": 60449003.0,
      "step": 23850
    },
    {
      "entropy": 6.2346738338470455,
      "epoch": 2.7530294287362955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042546850349359976,
      "loss": 5.8464,
      "mean_token_accuracy": 0.17728771269321442,
      "num_tokens": 60462083.0,
      "step": 23855
    },
    {
      "entropy": 6.215177822113037,
      "epoch": 2.753606462781304,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042543782139909894,
      "loss": 5.8151,
      "mean_token_accuracy": 0.1816414251923561,
      "num_tokens": 60473804.0,
      "step": 23860
    },
    {
      "entropy": 6.173917818069458,
      "epoch": 2.754183496826313,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042540713424469667,
      "loss": 5.8632,
      "mean_token_accuracy": 0.18780239522457123,
      "num_tokens": 60486361.0,
      "step": 23865
    },
    {
      "entropy": 6.1345775604248045,
      "epoch": 2.7547605308713212,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042537644203142523,
      "loss": 5.7347,
      "mean_token_accuracy": 0.19797743409872054,
      "num_tokens": 60498522.0,
      "step": 23870
    },
    {
      "entropy": 6.249479818344116,
      "epoch": 2.75533756491633,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042534574476031683,
      "loss": 5.9229,
      "mean_token_accuracy": 0.17691561430692673,
      "num_tokens": 60510974.0,
      "step": 23875
    },
    {
      "entropy": 6.248273038864136,
      "epoch": 2.755914598961339,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000425315042432404,
      "loss": 5.8542,
      "mean_token_accuracy": 0.18343111127614975,
      "num_tokens": 60524756.0,
      "step": 23880
    },
    {
      "entropy": 6.040663337707519,
      "epoch": 2.7564916330063474,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042528433504871935,
      "loss": 5.7302,
      "mean_token_accuracy": 0.19474226385354995,
      "num_tokens": 60536286.0,
      "step": 23885
    },
    {
      "entropy": 6.215526485443116,
      "epoch": 2.757068667051356,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004252536226102958,
      "loss": 5.8284,
      "mean_token_accuracy": 0.1841632381081581,
      "num_tokens": 60548971.0,
      "step": 23890
    },
    {
      "entropy": 6.172806930541992,
      "epoch": 2.7576457010963646,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004252229051181662,
      "loss": 5.8706,
      "mean_token_accuracy": 0.1718157559633255,
      "num_tokens": 60562078.0,
      "step": 23895
    },
    {
      "entropy": 6.1992229461669925,
      "epoch": 2.7582227351413735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042519218257336383,
      "loss": 5.8082,
      "mean_token_accuracy": 0.18301571756601334,
      "num_tokens": 60575378.0,
      "step": 23900
    },
    {
      "entropy": 6.12810206413269,
      "epoch": 2.758799769186382,
      "grad_norm": 0.875,
      "learning_rate": 0.00042516145497692204,
      "loss": 5.8073,
      "mean_token_accuracy": 0.18741518259048462,
      "num_tokens": 60588958.0,
      "step": 23905
    },
    {
      "entropy": 6.090682315826416,
      "epoch": 2.759376803231391,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042513072232987426,
      "loss": 5.7329,
      "mean_token_accuracy": 0.19598086625337602,
      "num_tokens": 60600904.0,
      "step": 23910
    },
    {
      "entropy": 6.212270641326905,
      "epoch": 2.759953837276399,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004250999846332543,
      "loss": 5.785,
      "mean_token_accuracy": 0.18600962162017823,
      "num_tokens": 60612655.0,
      "step": 23915
    },
    {
      "entropy": 6.168171691894531,
      "epoch": 2.760530871321408,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042506924188809574,
      "loss": 5.8476,
      "mean_token_accuracy": 0.1838272586464882,
      "num_tokens": 60625491.0,
      "step": 23920
    },
    {
      "entropy": 6.154070663452148,
      "epoch": 2.7611079053664165,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000425038494095433,
      "loss": 5.8238,
      "mean_token_accuracy": 0.18354423195123673,
      "num_tokens": 60637814.0,
      "step": 23925
    },
    {
      "entropy": 6.20363221168518,
      "epoch": 2.7616849394114253,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042500774125629986,
      "loss": 5.8165,
      "mean_token_accuracy": 0.18765195161104203,
      "num_tokens": 60650031.0,
      "step": 23930
    },
    {
      "entropy": 6.206214237213135,
      "epoch": 2.762261973456434,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004249769833717309,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18461532145738602,
      "num_tokens": 60662257.0,
      "step": 23935
    },
    {
      "entropy": 6.219572067260742,
      "epoch": 2.7628390075014426,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042494622044276066,
      "loss": 5.8635,
      "mean_token_accuracy": 0.18422316014766693,
      "num_tokens": 60675557.0,
      "step": 23940
    },
    {
      "entropy": 6.136008644104004,
      "epoch": 2.763416041546451,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004249154524704237,
      "loss": 5.8008,
      "mean_token_accuracy": 0.18340324759483337,
      "num_tokens": 60686984.0,
      "step": 23945
    },
    {
      "entropy": 6.2107688903808596,
      "epoch": 2.76399307559146,
      "grad_norm": 0.984375,
      "learning_rate": 0.000424884679455755,
      "loss": 5.8884,
      "mean_token_accuracy": 0.17726579159498215,
      "num_tokens": 60698904.0,
      "step": 23950
    },
    {
      "entropy": 6.175592947006225,
      "epoch": 2.7645701096364688,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042485390139978955,
      "loss": 5.798,
      "mean_token_accuracy": 0.18251847475767136,
      "num_tokens": 60710972.0,
      "step": 23955
    },
    {
      "entropy": 6.187010955810547,
      "epoch": 2.765147143681477,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004248231183035626,
      "loss": 5.8954,
      "mean_token_accuracy": 0.18239185959100723,
      "num_tokens": 60723780.0,
      "step": 23960
    },
    {
      "entropy": 6.154711198806763,
      "epoch": 2.7657241777264856,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004247923301681095,
      "loss": 5.7858,
      "mean_token_accuracy": 0.1891516774892807,
      "num_tokens": 60735381.0,
      "step": 23965
    },
    {
      "entropy": 6.0915274143219,
      "epoch": 2.7663012117714945,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042476153699446567,
      "loss": 5.7645,
      "mean_token_accuracy": 0.19561954885721206,
      "num_tokens": 60748495.0,
      "step": 23970
    },
    {
      "entropy": 6.231597948074341,
      "epoch": 2.7668782458165033,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042473073878366695,
      "loss": 5.8777,
      "mean_token_accuracy": 0.17465647608041762,
      "num_tokens": 60761984.0,
      "step": 23975
    },
    {
      "entropy": 6.224951648712159,
      "epoch": 2.7674552798615117,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004246999355367493,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1791698306798935,
      "num_tokens": 60774590.0,
      "step": 23980
    },
    {
      "entropy": 6.245056867599487,
      "epoch": 2.7680323139065206,
      "grad_norm": 0.859375,
      "learning_rate": 0.00042466912725474865,
      "loss": 5.8721,
      "mean_token_accuracy": 0.18126921802759172,
      "num_tokens": 60788647.0,
      "step": 23985
    },
    {
      "entropy": 6.220222806930542,
      "epoch": 2.768609347951529,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042463831393870123,
      "loss": 5.8197,
      "mean_token_accuracy": 0.18331341594457626,
      "num_tokens": 60799913.0,
      "step": 23990
    },
    {
      "entropy": 6.18059401512146,
      "epoch": 2.769186381996538,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042460749558964346,
      "loss": 5.9081,
      "mean_token_accuracy": 0.17576878666877746,
      "num_tokens": 60812343.0,
      "step": 23995
    },
    {
      "entropy": 6.242327404022217,
      "epoch": 2.7697634160415463,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004245766722086118,
      "loss": 5.8221,
      "mean_token_accuracy": 0.1815970480442047,
      "num_tokens": 60824677.0,
      "step": 24000
    },
    {
      "epoch": 2.7697634160415463,
      "eval_entropy": 6.063597902168358,
      "eval_loss": 5.926702976226807,
      "eval_mean_token_accuracy": 0.1853055248070346,
      "eval_num_tokens": 60824677.0,
      "eval_runtime": 22.7196,
      "eval_samples_per_second": 1238.402,
      "eval_steps_per_second": 154.8,
      "step": 24000
    },
    {
      "entropy": 6.19868278503418,
      "epoch": 2.770340450086555,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042454584379664326,
      "loss": 5.8272,
      "mean_token_accuracy": 0.18617044091224672,
      "num_tokens": 60837453.0,
      "step": 24005
    },
    {
      "entropy": 6.142568969726563,
      "epoch": 2.770917484131564,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245150103547744,
      "loss": 5.7947,
      "mean_token_accuracy": 0.18765759468078613,
      "num_tokens": 60850229.0,
      "step": 24010
    },
    {
      "entropy": 6.224448204040527,
      "epoch": 2.7714945181765724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042448417188404246,
      "loss": 5.809,
      "mean_token_accuracy": 0.1853519007563591,
      "num_tokens": 60862108.0,
      "step": 24015
    },
    {
      "entropy": 6.148353958129883,
      "epoch": 2.772071552221581,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004244533283854847,
      "loss": 5.8472,
      "mean_token_accuracy": 0.18446369469165802,
      "num_tokens": 60875339.0,
      "step": 24020
    },
    {
      "entropy": 6.144625186920166,
      "epoch": 2.7726485862665897,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004244224798601385,
      "loss": 5.8111,
      "mean_token_accuracy": 0.1819521129131317,
      "num_tokens": 60888869.0,
      "step": 24025
    },
    {
      "entropy": 6.181143283843994,
      "epoch": 2.7732256203115986,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042439162630904134,
      "loss": 5.8931,
      "mean_token_accuracy": 0.17629420906305313,
      "num_tokens": 60902344.0,
      "step": 24030
    },
    {
      "entropy": 6.243525695800781,
      "epoch": 2.773802654356607,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004243607677332311,
      "loss": 5.8645,
      "mean_token_accuracy": 0.17917369455099105,
      "num_tokens": 60915114.0,
      "step": 24035
    },
    {
      "entropy": 6.192512559890747,
      "epoch": 2.7743796884016154,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042432990413374556,
      "loss": 5.7962,
      "mean_token_accuracy": 0.18864602446556092,
      "num_tokens": 60928228.0,
      "step": 24040
    },
    {
      "entropy": 6.135471296310425,
      "epoch": 2.7749567224466243,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042429903551162283,
      "loss": 5.7809,
      "mean_token_accuracy": 0.1851405009627342,
      "num_tokens": 60941923.0,
      "step": 24045
    },
    {
      "entropy": 6.176653289794922,
      "epoch": 2.775533756491633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004242681618679013,
      "loss": 5.7971,
      "mean_token_accuracy": 0.1788968861103058,
      "num_tokens": 60955537.0,
      "step": 24050
    },
    {
      "entropy": 6.1557084083557125,
      "epoch": 2.7761107905366416,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00042423728320361924,
      "loss": 5.8469,
      "mean_token_accuracy": 0.18565410226583481,
      "num_tokens": 60968121.0,
      "step": 24055
    },
    {
      "entropy": 6.171410846710205,
      "epoch": 2.7766878245816504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042420639951981534,
      "loss": 5.7975,
      "mean_token_accuracy": 0.186527718603611,
      "num_tokens": 60980256.0,
      "step": 24060
    },
    {
      "entropy": 6.258060264587402,
      "epoch": 2.777264858626659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042417551081752825,
      "loss": 5.8906,
      "mean_token_accuracy": 0.176424802839756,
      "num_tokens": 60992989.0,
      "step": 24065
    },
    {
      "entropy": 6.216572904586792,
      "epoch": 2.7778418926716677,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042414461709779696,
      "loss": 5.8576,
      "mean_token_accuracy": 0.17997607290744783,
      "num_tokens": 61005792.0,
      "step": 24070
    },
    {
      "entropy": 6.156715631484985,
      "epoch": 2.778418926716676,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004241137183616606,
      "loss": 5.8729,
      "mean_token_accuracy": 0.18681135177612304,
      "num_tokens": 61019448.0,
      "step": 24075
    },
    {
      "entropy": 6.177115440368652,
      "epoch": 2.778995960761685,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004240828146101584,
      "loss": 5.7828,
      "mean_token_accuracy": 0.18995569050312042,
      "num_tokens": 61033449.0,
      "step": 24080
    },
    {
      "entropy": 6.263371229171753,
      "epoch": 2.779572994806694,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042405190584432966,
      "loss": 5.8635,
      "mean_token_accuracy": 0.18435943126678467,
      "num_tokens": 61044912.0,
      "step": 24085
    },
    {
      "entropy": 6.180558919906616,
      "epoch": 2.7801500288517023,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004240209920652142,
      "loss": 5.7965,
      "mean_token_accuracy": 0.1852031782269478,
      "num_tokens": 61057003.0,
      "step": 24090
    },
    {
      "entropy": 6.15169939994812,
      "epoch": 2.7807270628967107,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042399007327385173,
      "loss": 5.7984,
      "mean_token_accuracy": 0.18453529328107834,
      "num_tokens": 61068721.0,
      "step": 24095
    },
    {
      "entropy": 6.172250032424927,
      "epoch": 2.7813040969417195,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004239591494712821,
      "loss": 5.7722,
      "mean_token_accuracy": 0.18817009627819062,
      "num_tokens": 61081938.0,
      "step": 24100
    },
    {
      "entropy": 6.201770067214966,
      "epoch": 2.7818811309867284,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004239282206585455,
      "loss": 5.8643,
      "mean_token_accuracy": 0.18166793286800384,
      "num_tokens": 61094918.0,
      "step": 24105
    },
    {
      "entropy": 6.125176095962525,
      "epoch": 2.782458165031737,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004238972868366822,
      "loss": 5.7933,
      "mean_token_accuracy": 0.18579421043395997,
      "num_tokens": 61107201.0,
      "step": 24110
    },
    {
      "entropy": 6.129706287384034,
      "epoch": 2.7830351990767457,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042386634800673255,
      "loss": 5.6957,
      "mean_token_accuracy": 0.20030567944049835,
      "num_tokens": 61121092.0,
      "step": 24115
    },
    {
      "entropy": 6.168320417404175,
      "epoch": 2.783612233121754,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004238354041697372,
      "loss": 5.8612,
      "mean_token_accuracy": 0.1814897760748863,
      "num_tokens": 61133819.0,
      "step": 24120
    },
    {
      "entropy": 6.208771133422852,
      "epoch": 2.784189267166763,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042380445532673705,
      "loss": 5.9235,
      "mean_token_accuracy": 0.17213354408740997,
      "num_tokens": 61147182.0,
      "step": 24125
    },
    {
      "entropy": 6.264519929885864,
      "epoch": 2.7847663012117714,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004237735014787729,
      "loss": 5.9022,
      "mean_token_accuracy": 0.1765430524945259,
      "num_tokens": 61159729.0,
      "step": 24130
    },
    {
      "entropy": 6.2641078472137455,
      "epoch": 2.7853433352567802,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423742542626886,
      "loss": 5.8449,
      "mean_token_accuracy": 0.1809821233153343,
      "num_tokens": 61172528.0,
      "step": 24135
    },
    {
      "entropy": 6.265717029571533,
      "epoch": 2.7859203693017887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004237115787721176,
      "loss": 5.9428,
      "mean_token_accuracy": 0.175088506937027,
      "num_tokens": 61186874.0,
      "step": 24140
    },
    {
      "entropy": 6.226970577239991,
      "epoch": 2.7864974033467975,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042368060991550895,
      "loss": 5.7829,
      "mean_token_accuracy": 0.1872684195637703,
      "num_tokens": 61199353.0,
      "step": 24145
    },
    {
      "entropy": 6.1586408615112305,
      "epoch": 2.787074437391806,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004236496360581019,
      "loss": 5.8003,
      "mean_token_accuracy": 0.18614619225263596,
      "num_tokens": 61211799.0,
      "step": 24150
    },
    {
      "entropy": 6.219261741638183,
      "epoch": 2.787651471436815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042361865720093826,
      "loss": 5.837,
      "mean_token_accuracy": 0.18358803391456605,
      "num_tokens": 61225029.0,
      "step": 24155
    },
    {
      "entropy": 6.140805244445801,
      "epoch": 2.7882285054818237,
      "grad_norm": 1.328125,
      "learning_rate": 0.00042358767334505984,
      "loss": 5.7485,
      "mean_token_accuracy": 0.18711104840040207,
      "num_tokens": 61238415.0,
      "step": 24160
    },
    {
      "entropy": 6.15421233177185,
      "epoch": 2.788805539526832,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042355668449150884,
      "loss": 5.8419,
      "mean_token_accuracy": 0.1833636298775673,
      "num_tokens": 61251193.0,
      "step": 24165
    },
    {
      "entropy": 6.163024520874023,
      "epoch": 2.7893825735718405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042352569064132756,
      "loss": 5.8197,
      "mean_token_accuracy": 0.18791975378990172,
      "num_tokens": 61264752.0,
      "step": 24170
    },
    {
      "entropy": 6.17081298828125,
      "epoch": 2.7899596076168494,
      "grad_norm": 1.203125,
      "learning_rate": 0.00042349469179555845,
      "loss": 5.8006,
      "mean_token_accuracy": 0.19208011627197266,
      "num_tokens": 61277409.0,
      "step": 24175
    },
    {
      "entropy": 6.171252822875976,
      "epoch": 2.790536641661858,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042346368795524416,
      "loss": 5.7636,
      "mean_token_accuracy": 0.18603039532899857,
      "num_tokens": 61288810.0,
      "step": 24180
    },
    {
      "entropy": 6.219677972793579,
      "epoch": 2.7911136757068666,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004234326791214274,
      "loss": 5.7651,
      "mean_token_accuracy": 0.18796777874231338,
      "num_tokens": 61301080.0,
      "step": 24185
    },
    {
      "entropy": 6.190580749511719,
      "epoch": 2.7916907097518755,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004234016652951513,
      "loss": 5.8611,
      "mean_token_accuracy": 0.1816395029425621,
      "num_tokens": 61313202.0,
      "step": 24190
    },
    {
      "entropy": 6.104137182235718,
      "epoch": 2.792267743796884,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042337064647745887,
      "loss": 5.7755,
      "mean_token_accuracy": 0.18244900554418564,
      "num_tokens": 61326090.0,
      "step": 24195
    },
    {
      "entropy": 6.272701168060303,
      "epoch": 2.792844777841893,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042333962266939343,
      "loss": 5.8703,
      "mean_token_accuracy": 0.1829444631934166,
      "num_tokens": 61338640.0,
      "step": 24200
    },
    {
      "entropy": 6.183998870849609,
      "epoch": 2.793421811886901,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00042330859387199846,
      "loss": 5.7816,
      "mean_token_accuracy": 0.19163548201322556,
      "num_tokens": 61351324.0,
      "step": 24205
    },
    {
      "entropy": 6.169540452957153,
      "epoch": 2.79399884593191,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004232775600863176,
      "loss": 5.7587,
      "mean_token_accuracy": 0.1945983037352562,
      "num_tokens": 61362547.0,
      "step": 24210
    },
    {
      "entropy": 6.0055286407470705,
      "epoch": 2.7945758799769185,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00042324652131339475,
      "loss": 5.6191,
      "mean_token_accuracy": 0.2038264036178589,
      "num_tokens": 61376084.0,
      "step": 24215
    },
    {
      "entropy": 6.18643159866333,
      "epoch": 2.7951529140219273,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004232154775542737,
      "loss": 5.8277,
      "mean_token_accuracy": 0.18713755309581756,
      "num_tokens": 61388440.0,
      "step": 24220
    },
    {
      "entropy": 6.238617372512818,
      "epoch": 2.7957299480669358,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004231844288099987,
      "loss": 5.8876,
      "mean_token_accuracy": 0.17568867206573485,
      "num_tokens": 61402255.0,
      "step": 24225
    },
    {
      "entropy": 6.165315437316894,
      "epoch": 2.7963069821119446,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00042315337508161405,
      "loss": 5.7447,
      "mean_token_accuracy": 0.18500784784555435,
      "num_tokens": 61414871.0,
      "step": 24230
    },
    {
      "entropy": 6.181436586380005,
      "epoch": 2.7968840161569535,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042312231637016423,
      "loss": 5.8117,
      "mean_token_accuracy": 0.18590328246355056,
      "num_tokens": 61427127.0,
      "step": 24235
    },
    {
      "entropy": 6.183506155014038,
      "epoch": 2.797461050201962,
      "grad_norm": 1.0,
      "learning_rate": 0.00042309125267669386,
      "loss": 5.7866,
      "mean_token_accuracy": 0.1843393936753273,
      "num_tokens": 61439191.0,
      "step": 24240
    },
    {
      "entropy": 6.134817886352539,
      "epoch": 2.7980380842469703,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004230601840022478,
      "loss": 5.7783,
      "mean_token_accuracy": 0.18546461462974548,
      "num_tokens": 61452508.0,
      "step": 24245
    },
    {
      "entropy": 6.21579909324646,
      "epoch": 2.798615118291979,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000423029110347871,
      "loss": 5.8466,
      "mean_token_accuracy": 0.17796468436717988,
      "num_tokens": 61466475.0,
      "step": 24250
    },
    {
      "entropy": 6.159425258636475,
      "epoch": 2.799192152336988,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042299803171460854,
      "loss": 5.7605,
      "mean_token_accuracy": 0.1877009317278862,
      "num_tokens": 61478523.0,
      "step": 24255
    },
    {
      "entropy": 6.210866689682007,
      "epoch": 2.7997691863819965,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00042296694810350585,
      "loss": 5.7951,
      "mean_token_accuracy": 0.18575677275657654,
      "num_tokens": 61490591.0,
      "step": 24260
    },
    {
      "entropy": 6.19249119758606,
      "epoch": 2.8003462204270053,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004229358595156083,
      "loss": 5.8166,
      "mean_token_accuracy": 0.18493679761886597,
      "num_tokens": 61503414.0,
      "step": 24265
    },
    {
      "entropy": 6.172161340713501,
      "epoch": 2.8009232544720137,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004229047659519617,
      "loss": 5.7625,
      "mean_token_accuracy": 0.19090480357408524,
      "num_tokens": 61516655.0,
      "step": 24270
    },
    {
      "entropy": 6.3251043319702145,
      "epoch": 2.8015002885170226,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004228736674136117,
      "loss": 5.9208,
      "mean_token_accuracy": 0.18344197422266006,
      "num_tokens": 61529589.0,
      "step": 24275
    },
    {
      "entropy": 6.168008756637573,
      "epoch": 2.802077322562031,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004228425639016044,
      "loss": 5.7674,
      "mean_token_accuracy": 0.18482994735240937,
      "num_tokens": 61541648.0,
      "step": 24280
    },
    {
      "entropy": 6.171618986129761,
      "epoch": 2.80265435660704,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004228114554169858,
      "loss": 5.8482,
      "mean_token_accuracy": 0.18270285427570343,
      "num_tokens": 61552864.0,
      "step": 24285
    },
    {
      "entropy": 6.185155296325684,
      "epoch": 2.8032313906520487,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004227803419608024,
      "loss": 5.9095,
      "mean_token_accuracy": 0.17858732789754866,
      "num_tokens": 61565179.0,
      "step": 24290
    },
    {
      "entropy": 6.277147340774536,
      "epoch": 2.803808424697057,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042274922353410056,
      "loss": 5.8606,
      "mean_token_accuracy": 0.18767596781253815,
      "num_tokens": 61577386.0,
      "step": 24295
    },
    {
      "entropy": 6.189168405532837,
      "epoch": 2.8043854587420656,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042271810013792696,
      "loss": 5.8136,
      "mean_token_accuracy": 0.1885599747300148,
      "num_tokens": 61590519.0,
      "step": 24300
    },
    {
      "entropy": 6.151488494873047,
      "epoch": 2.8049624927870744,
      "grad_norm": 0.875,
      "learning_rate": 0.0004226869717733285,
      "loss": 5.8437,
      "mean_token_accuracy": 0.18423136025667192,
      "num_tokens": 61603323.0,
      "step": 24305
    },
    {
      "entropy": 6.205613613128662,
      "epoch": 2.8055395268320833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042265583844135207,
      "loss": 5.7631,
      "mean_token_accuracy": 0.18863353431224822,
      "num_tokens": 61615637.0,
      "step": 24310
    },
    {
      "entropy": 6.155217981338501,
      "epoch": 2.8061165608770917,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042262470014304486,
      "loss": 5.7638,
      "mean_token_accuracy": 0.19537921100854874,
      "num_tokens": 61628004.0,
      "step": 24315
    },
    {
      "entropy": 6.1550205707550045,
      "epoch": 2.8066935949221,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004225935568794542,
      "loss": 5.7275,
      "mean_token_accuracy": 0.19332590252161025,
      "num_tokens": 61641274.0,
      "step": 24320
    },
    {
      "entropy": 6.209946870803833,
      "epoch": 2.807270628967109,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042256240865162764,
      "loss": 5.9141,
      "mean_token_accuracy": 0.1799610212445259,
      "num_tokens": 61654398.0,
      "step": 24325
    },
    {
      "entropy": 6.233845233917236,
      "epoch": 2.807847663012118,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042253125546061265,
      "loss": 5.873,
      "mean_token_accuracy": 0.17753645926713943,
      "num_tokens": 61667261.0,
      "step": 24330
    },
    {
      "entropy": 6.28053126335144,
      "epoch": 2.8084246970571263,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004225000973074572,
      "loss": 5.9042,
      "mean_token_accuracy": 0.18290190845727922,
      "num_tokens": 61678871.0,
      "step": 24335
    },
    {
      "entropy": 6.1552042961120605,
      "epoch": 2.809001731102135,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042246893419320923,
      "loss": 5.8203,
      "mean_token_accuracy": 0.19056282788515091,
      "num_tokens": 61692124.0,
      "step": 24340
    },
    {
      "entropy": 6.1690881729125975,
      "epoch": 2.8095787651471436,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000422437766118917,
      "loss": 6.0049,
      "mean_token_accuracy": 0.18519874513149262,
      "num_tokens": 61704164.0,
      "step": 24345
    },
    {
      "entropy": 6.116635322570801,
      "epoch": 2.8101557991921524,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004224065930856286,
      "loss": 5.7702,
      "mean_token_accuracy": 0.18663939386606215,
      "num_tokens": 61716031.0,
      "step": 24350
    },
    {
      "entropy": 6.252239322662353,
      "epoch": 2.810732833237161,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004223754150943927,
      "loss": 5.8029,
      "mean_token_accuracy": 0.1906353548169136,
      "num_tokens": 61728663.0,
      "step": 24355
    },
    {
      "entropy": 6.196207809448242,
      "epoch": 2.8113098672821697,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000422344232146258,
      "loss": 5.7841,
      "mean_token_accuracy": 0.18982133716344834,
      "num_tokens": 61740998.0,
      "step": 24360
    },
    {
      "entropy": 6.106575345993042,
      "epoch": 2.8118869013271786,
      "grad_norm": 0.859375,
      "learning_rate": 0.00042231304424227315,
      "loss": 5.8081,
      "mean_token_accuracy": 0.18961165547370912,
      "num_tokens": 61753231.0,
      "step": 24365
    },
    {
      "entropy": 6.193637418746948,
      "epoch": 2.812463935372187,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042228185138348736,
      "loss": 5.8824,
      "mean_token_accuracy": 0.17971400022506714,
      "num_tokens": 61765046.0,
      "step": 24370
    },
    {
      "entropy": 6.177429723739624,
      "epoch": 2.8130409694171954,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004222506535709496,
      "loss": 5.7369,
      "mean_token_accuracy": 0.1918184667825699,
      "num_tokens": 61776705.0,
      "step": 24375
    },
    {
      "entropy": 6.229557085037231,
      "epoch": 2.8136180034622043,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004222194508057092,
      "loss": 5.7933,
      "mean_token_accuracy": 0.18529320508241653,
      "num_tokens": 61790682.0,
      "step": 24380
    },
    {
      "entropy": 6.120730209350586,
      "epoch": 2.814195037507213,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004221882430888157,
      "loss": 5.7524,
      "mean_token_accuracy": 0.1902666687965393,
      "num_tokens": 61803329.0,
      "step": 24385
    },
    {
      "entropy": 6.138819074630737,
      "epoch": 2.8147720715522215,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042215703042131883,
      "loss": 5.7971,
      "mean_token_accuracy": 0.18944026827812194,
      "num_tokens": 61815996.0,
      "step": 24390
    },
    {
      "entropy": 6.155755615234375,
      "epoch": 2.8153491055972304,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004221258128042682,
      "loss": 5.7808,
      "mean_token_accuracy": 0.19141545295715331,
      "num_tokens": 61828281.0,
      "step": 24395
    },
    {
      "entropy": 6.221925210952759,
      "epoch": 2.815926139642239,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042209459023871405,
      "loss": 5.8719,
      "mean_token_accuracy": 0.1785180911421776,
      "num_tokens": 61840894.0,
      "step": 24400
    },
    {
      "entropy": 6.228180551528931,
      "epoch": 2.8165031736872477,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042206336272570643,
      "loss": 5.8429,
      "mean_token_accuracy": 0.18124893456697463,
      "num_tokens": 61852767.0,
      "step": 24405
    },
    {
      "entropy": 6.226462078094483,
      "epoch": 2.817080207732256,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042203213026629566,
      "loss": 5.7607,
      "mean_token_accuracy": 0.18627342879772185,
      "num_tokens": 61866497.0,
      "step": 24410
    },
    {
      "entropy": 6.182011890411377,
      "epoch": 2.817657241777265,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042200089286153217,
      "loss": 5.8813,
      "mean_token_accuracy": 0.18050605803728104,
      "num_tokens": 61879024.0,
      "step": 24415
    },
    {
      "entropy": 6.139136266708374,
      "epoch": 2.8182342758222734,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004219696505124667,
      "loss": 5.7468,
      "mean_token_accuracy": 0.18669730871915818,
      "num_tokens": 61890582.0,
      "step": 24420
    },
    {
      "entropy": 6.165212678909302,
      "epoch": 2.8188113098672822,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042193840322015,
      "loss": 5.7605,
      "mean_token_accuracy": 0.18675775378942489,
      "num_tokens": 61903140.0,
      "step": 24425
    },
    {
      "entropy": 6.1244893074035645,
      "epoch": 2.8193883439122907,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042190715098563313,
      "loss": 5.7172,
      "mean_token_accuracy": 0.1883280023932457,
      "num_tokens": 61915953.0,
      "step": 24430
    },
    {
      "entropy": 6.139819765090943,
      "epoch": 2.8199653779572995,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004218758938099672,
      "loss": 5.8658,
      "mean_token_accuracy": 0.18507246673107147,
      "num_tokens": 61928120.0,
      "step": 24435
    },
    {
      "entropy": 6.2485874652862545,
      "epoch": 2.8205424120023084,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004218446316942035,
      "loss": 5.8294,
      "mean_token_accuracy": 0.18644487261772155,
      "num_tokens": 61941164.0,
      "step": 24440
    },
    {
      "entropy": 6.145779275894165,
      "epoch": 2.821119446047317,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004218133646393937,
      "loss": 5.7889,
      "mean_token_accuracy": 0.19110679924488067,
      "num_tokens": 61954310.0,
      "step": 24445
    },
    {
      "entropy": 6.198514032363891,
      "epoch": 2.821696480092325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004217820926465892,
      "loss": 5.8485,
      "mean_token_accuracy": 0.18392210304737092,
      "num_tokens": 61966952.0,
      "step": 24450
    },
    {
      "entropy": 6.200435400009155,
      "epoch": 2.822273514137334,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004217508157168418,
      "loss": 5.8006,
      "mean_token_accuracy": 0.1839941754937172,
      "num_tokens": 61979547.0,
      "step": 24455
    },
    {
      "entropy": 6.1755468368530275,
      "epoch": 2.822850548182343,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042171953385120384,
      "loss": 5.8424,
      "mean_token_accuracy": 0.18718830198049546,
      "num_tokens": 61992188.0,
      "step": 24460
    },
    {
      "entropy": 6.224154090881347,
      "epoch": 2.8234275822273514,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042168824705072713,
      "loss": 5.8042,
      "mean_token_accuracy": 0.19423746019601823,
      "num_tokens": 62005426.0,
      "step": 24465
    },
    {
      "entropy": 6.062999820709228,
      "epoch": 2.82400461627236,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004216569553164641,
      "loss": 5.7434,
      "mean_token_accuracy": 0.1871136888861656,
      "num_tokens": 62018247.0,
      "step": 24470
    },
    {
      "entropy": 6.092436838150024,
      "epoch": 2.8245816503173686,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042162565864946723,
      "loss": 5.8167,
      "mean_token_accuracy": 0.18545281887054443,
      "num_tokens": 62030503.0,
      "step": 24475
    },
    {
      "entropy": 6.31914496421814,
      "epoch": 2.8251586843623775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004215943570507891,
      "loss": 5.8503,
      "mean_token_accuracy": 0.18152353614568711,
      "num_tokens": 62043453.0,
      "step": 24480
    },
    {
      "entropy": 6.148798513412475,
      "epoch": 2.825735718407386,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042156305052148263,
      "loss": 5.7275,
      "mean_token_accuracy": 0.19739930331707,
      "num_tokens": 62057317.0,
      "step": 24485
    },
    {
      "entropy": 6.10639181137085,
      "epoch": 2.8263127524523948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042153173906260083,
      "loss": 5.8201,
      "mean_token_accuracy": 0.18516720235347747,
      "num_tokens": 62069360.0,
      "step": 24490
    },
    {
      "entropy": 6.182285165786743,
      "epoch": 2.826889786497403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042150042267519665,
      "loss": 5.8326,
      "mean_token_accuracy": 0.18462236523628234,
      "num_tokens": 62082157.0,
      "step": 24495
    },
    {
      "entropy": 6.143859577178955,
      "epoch": 2.827466820542412,
      "grad_norm": 1.0,
      "learning_rate": 0.00042146910136032367,
      "loss": 5.7447,
      "mean_token_accuracy": 0.19464441984891892,
      "num_tokens": 62094268.0,
      "step": 24500
    },
    {
      "entropy": 6.169033098220825,
      "epoch": 2.8280438545874205,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042143777511903513,
      "loss": 5.7834,
      "mean_token_accuracy": 0.18847947865724562,
      "num_tokens": 62107610.0,
      "step": 24505
    },
    {
      "entropy": 6.150124740600586,
      "epoch": 2.8286208886324293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042140644395238477,
      "loss": 5.7807,
      "mean_token_accuracy": 0.19214766174554826,
      "num_tokens": 62120279.0,
      "step": 24510
    },
    {
      "entropy": 6.096494579315186,
      "epoch": 2.829197922677438,
      "grad_norm": 1.0,
      "learning_rate": 0.0004213751078614264,
      "loss": 5.703,
      "mean_token_accuracy": 0.1975172147154808,
      "num_tokens": 62134204.0,
      "step": 24515
    },
    {
      "entropy": 6.1175798892974855,
      "epoch": 2.8297749567224466,
      "grad_norm": 0.953125,
      "learning_rate": 0.000421343766847214,
      "loss": 5.7332,
      "mean_token_accuracy": 0.19724698662757872,
      "num_tokens": 62147241.0,
      "step": 24520
    },
    {
      "entropy": 6.1617988586425785,
      "epoch": 2.830351990767455,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004213124209108016,
      "loss": 5.761,
      "mean_token_accuracy": 0.18851809501647948,
      "num_tokens": 62159935.0,
      "step": 24525
    },
    {
      "entropy": 6.100178480148315,
      "epoch": 2.830929024812464,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004212810700532437,
      "loss": 5.7487,
      "mean_token_accuracy": 0.19553690850734712,
      "num_tokens": 62172498.0,
      "step": 24530
    },
    {
      "entropy": 6.1133757591247555,
      "epoch": 2.8315060588574728,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004212497142755947,
      "loss": 5.7746,
      "mean_token_accuracy": 0.183367919921875,
      "num_tokens": 62186150.0,
      "step": 24535
    },
    {
      "entropy": 6.262248611450195,
      "epoch": 2.832083092902481,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042121835357890916,
      "loss": 5.8965,
      "mean_token_accuracy": 0.18269695043563844,
      "num_tokens": 62199455.0,
      "step": 24540
    },
    {
      "entropy": 6.212676286697388,
      "epoch": 2.83266012694749,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004211869879642419,
      "loss": 5.8295,
      "mean_token_accuracy": 0.1768409490585327,
      "num_tokens": 62213459.0,
      "step": 24545
    },
    {
      "entropy": 6.166826057434082,
      "epoch": 2.8332371609924984,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042115561743264797,
      "loss": 5.7797,
      "mean_token_accuracy": 0.18591320812702178,
      "num_tokens": 62225250.0,
      "step": 24550
    },
    {
      "entropy": 6.147188901901245,
      "epoch": 2.8338141950375073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004211242419851824,
      "loss": 5.7759,
      "mean_token_accuracy": 0.19254399687051774,
      "num_tokens": 62238334.0,
      "step": 24555
    },
    {
      "entropy": 6.216676187515259,
      "epoch": 2.8343912290825157,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00042109286162290055,
      "loss": 5.8274,
      "mean_token_accuracy": 0.18938308507204055,
      "num_tokens": 62251498.0,
      "step": 24560
    },
    {
      "entropy": 6.222916889190674,
      "epoch": 2.8349682631275246,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004210614763468579,
      "loss": 5.8604,
      "mean_token_accuracy": 0.18268202245235443,
      "num_tokens": 62264423.0,
      "step": 24565
    },
    {
      "entropy": 6.155679655075073,
      "epoch": 2.8355452971725335,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042103008615811,
      "loss": 5.7626,
      "mean_token_accuracy": 0.18701843172311783,
      "num_tokens": 62276889.0,
      "step": 24570
    },
    {
      "entropy": 6.207239389419556,
      "epoch": 2.836122331217542,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004209986910577127,
      "loss": 5.8413,
      "mean_token_accuracy": 0.18645787090063096,
      "num_tokens": 62289043.0,
      "step": 24575
    },
    {
      "entropy": 6.19259238243103,
      "epoch": 2.8366993652625503,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042096729104672194,
      "loss": 5.8514,
      "mean_token_accuracy": 0.18569687902927398,
      "num_tokens": 62301445.0,
      "step": 24580
    },
    {
      "entropy": 6.210863542556763,
      "epoch": 2.837276399307559,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042093588612619385,
      "loss": 5.8636,
      "mean_token_accuracy": 0.17950042635202407,
      "num_tokens": 62314263.0,
      "step": 24585
    },
    {
      "entropy": 6.2886528968811035,
      "epoch": 2.837853433352568,
      "grad_norm": 1.0,
      "learning_rate": 0.0004209044762971847,
      "loss": 5.8165,
      "mean_token_accuracy": 0.19024156779050827,
      "num_tokens": 62327873.0,
      "step": 24590
    },
    {
      "entropy": 6.238538980484009,
      "epoch": 2.8384304673975764,
      "grad_norm": 1.109375,
      "learning_rate": 0.000420873061560751,
      "loss": 5.8346,
      "mean_token_accuracy": 0.17738230377435685,
      "num_tokens": 62340827.0,
      "step": 24595
    },
    {
      "entropy": 6.246775531768799,
      "epoch": 2.839007501442585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004208416419179494,
      "loss": 5.929,
      "mean_token_accuracy": 0.1809936672449112,
      "num_tokens": 62352995.0,
      "step": 24600
    },
    {
      "entropy": 6.232808542251587,
      "epoch": 2.8395845354875937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042081021736983657,
      "loss": 5.8607,
      "mean_token_accuracy": 0.1842508167028427,
      "num_tokens": 62365647.0,
      "step": 24605
    },
    {
      "entropy": 6.190806531906128,
      "epoch": 2.8401615695326026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004207787879174695,
      "loss": 5.7846,
      "mean_token_accuracy": 0.18463329076766968,
      "num_tokens": 62377277.0,
      "step": 24610
    },
    {
      "entropy": 6.135664463043213,
      "epoch": 2.840738603577611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042074735356190525,
      "loss": 5.8261,
      "mean_token_accuracy": 0.18321821540594102,
      "num_tokens": 62389338.0,
      "step": 24615
    },
    {
      "entropy": 6.20417971611023,
      "epoch": 2.84131563762262,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004207159143042012,
      "loss": 5.8274,
      "mean_token_accuracy": 0.18021536618471146,
      "num_tokens": 62401933.0,
      "step": 24620
    },
    {
      "entropy": 6.188966274261475,
      "epoch": 2.8418926716676283,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004206844701454148,
      "loss": 5.7641,
      "mean_token_accuracy": 0.1921963319182396,
      "num_tokens": 62414738.0,
      "step": 24625
    },
    {
      "entropy": 6.237261819839477,
      "epoch": 2.842469705712637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042065302108660355,
      "loss": 5.8905,
      "mean_token_accuracy": 0.17836265861988068,
      "num_tokens": 62426565.0,
      "step": 24630
    },
    {
      "entropy": 6.100368928909302,
      "epoch": 2.8430467397576455,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004206215671288253,
      "loss": 5.7612,
      "mean_token_accuracy": 0.19400888532400132,
      "num_tokens": 62438992.0,
      "step": 24635
    },
    {
      "entropy": 6.1758119583129885,
      "epoch": 2.8436237738026544,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042059010827313794,
      "loss": 5.8212,
      "mean_token_accuracy": 0.18566527515649794,
      "num_tokens": 62453004.0,
      "step": 24640
    },
    {
      "entropy": 6.297257661819458,
      "epoch": 2.8442008078476633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042055864452059967,
      "loss": 5.8788,
      "mean_token_accuracy": 0.17976783514022826,
      "num_tokens": 62466367.0,
      "step": 24645
    },
    {
      "entropy": 6.137363243103027,
      "epoch": 2.8447778418926717,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004205271758722687,
      "loss": 5.7024,
      "mean_token_accuracy": 0.19127903282642364,
      "num_tokens": 62479881.0,
      "step": 24650
    },
    {
      "entropy": 6.236714792251587,
      "epoch": 2.84535487593768,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004204957023292034,
      "loss": 5.9681,
      "mean_token_accuracy": 0.17598249912261962,
      "num_tokens": 62493634.0,
      "step": 24655
    },
    {
      "entropy": 6.183877468109131,
      "epoch": 2.845931909982689,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042046422389246246,
      "loss": 5.7305,
      "mean_token_accuracy": 0.1946361169219017,
      "num_tokens": 62505045.0,
      "step": 24660
    },
    {
      "entropy": 6.135823774337768,
      "epoch": 2.846508944027698,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042043274056310454,
      "loss": 5.8104,
      "mean_token_accuracy": 0.1842285603284836,
      "num_tokens": 62517403.0,
      "step": 24665
    },
    {
      "entropy": 6.123856830596924,
      "epoch": 2.8470859780727062,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004204012523421888,
      "loss": 5.749,
      "mean_token_accuracy": 0.1858529895544052,
      "num_tokens": 62529276.0,
      "step": 24670
    },
    {
      "entropy": 6.1114264011383055,
      "epoch": 2.847663012117715,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000420369759230774,
      "loss": 5.661,
      "mean_token_accuracy": 0.19927244931459426,
      "num_tokens": 62542705.0,
      "step": 24675
    },
    {
      "entropy": 6.153840398788452,
      "epoch": 2.8482400461627235,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00042033826122991956,
      "loss": 5.7966,
      "mean_token_accuracy": 0.1818278804421425,
      "num_tokens": 62555547.0,
      "step": 24680
    },
    {
      "entropy": 6.046644973754883,
      "epoch": 2.8488170802077324,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000420306758340685,
      "loss": 5.7347,
      "mean_token_accuracy": 0.18865351378917694,
      "num_tokens": 62569118.0,
      "step": 24685
    },
    {
      "entropy": 6.122761535644531,
      "epoch": 2.849394114252741,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042027525056412973,
      "loss": 5.7241,
      "mean_token_accuracy": 0.18449884206056594,
      "num_tokens": 62580632.0,
      "step": 24690
    },
    {
      "entropy": 6.106462812423706,
      "epoch": 2.8499711482977497,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004202437379013136,
      "loss": 5.6609,
      "mean_token_accuracy": 0.19353797286748886,
      "num_tokens": 62593554.0,
      "step": 24695
    },
    {
      "entropy": 6.127063751220703,
      "epoch": 2.850548182342758,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004202122203532965,
      "loss": 5.8283,
      "mean_token_accuracy": 0.18047186881303787,
      "num_tokens": 62606454.0,
      "step": 24700
    },
    {
      "entropy": 6.195889329910278,
      "epoch": 2.851125216387767,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00042018069792113845,
      "loss": 5.8153,
      "mean_token_accuracy": 0.1847582593560219,
      "num_tokens": 62620665.0,
      "step": 24705
    },
    {
      "entropy": 6.117110347747802,
      "epoch": 2.8517022504327754,
      "grad_norm": 0.875,
      "learning_rate": 0.0004201491706058998,
      "loss": 5.7621,
      "mean_token_accuracy": 0.19461952298879623,
      "num_tokens": 62632849.0,
      "step": 24710
    },
    {
      "entropy": 6.181552171707153,
      "epoch": 2.8522792844777842,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004201176384086408,
      "loss": 5.7987,
      "mean_token_accuracy": 0.18936560451984405,
      "num_tokens": 62644355.0,
      "step": 24715
    },
    {
      "entropy": 6.179149341583252,
      "epoch": 2.852856318522793,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004200861013304222,
      "loss": 5.7981,
      "mean_token_accuracy": 0.19061631411314012,
      "num_tokens": 62656642.0,
      "step": 24720
    },
    {
      "entropy": 6.137886047363281,
      "epoch": 2.8534333525678015,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042005455937230466,
      "loss": 5.7535,
      "mean_token_accuracy": 0.18561313301324844,
      "num_tokens": 62669993.0,
      "step": 24725
    },
    {
      "entropy": 6.233130073547363,
      "epoch": 2.85401038661281,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000420023012535349,
      "loss": 5.841,
      "mean_token_accuracy": 0.18337019979953767,
      "num_tokens": 62682336.0,
      "step": 24730
    },
    {
      "entropy": 6.261335992813111,
      "epoch": 2.854587420657819,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004199914608206164,
      "loss": 5.8806,
      "mean_token_accuracy": 0.18259891420602797,
      "num_tokens": 62694893.0,
      "step": 24735
    },
    {
      "entropy": 6.158029985427857,
      "epoch": 2.8551644547028276,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000419959904229168,
      "loss": 5.7876,
      "mean_token_accuracy": 0.1849174126982689,
      "num_tokens": 62707486.0,
      "step": 24740
    },
    {
      "entropy": 6.280964326858521,
      "epoch": 2.855741488747836,
      "grad_norm": 0.875,
      "learning_rate": 0.0004199283427620653,
      "loss": 5.8971,
      "mean_token_accuracy": 0.18116282671689987,
      "num_tokens": 62721350.0,
      "step": 24745
    },
    {
      "entropy": 6.071496391296387,
      "epoch": 2.856318522792845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041989677642036973,
      "loss": 5.6762,
      "mean_token_accuracy": 0.1947805941104889,
      "num_tokens": 62734133.0,
      "step": 24750
    },
    {
      "entropy": 6.117844104766846,
      "epoch": 2.8568955568378533,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041986520520514303,
      "loss": 5.7009,
      "mean_token_accuracy": 0.19389776438474654,
      "num_tokens": 62746218.0,
      "step": 24755
    },
    {
      "entropy": 6.211564588546753,
      "epoch": 2.857472590882862,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041983362911744715,
      "loss": 5.859,
      "mean_token_accuracy": 0.18451776057481767,
      "num_tokens": 62759137.0,
      "step": 24760
    },
    {
      "entropy": 6.204450511932373,
      "epoch": 2.8580496249278706,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004198020481583441,
      "loss": 5.826,
      "mean_token_accuracy": 0.18509777784347534,
      "num_tokens": 62772287.0,
      "step": 24765
    },
    {
      "entropy": 6.2304027557373045,
      "epoch": 2.8586266589728795,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041977046232889606,
      "loss": 5.8035,
      "mean_token_accuracy": 0.18370132446289061,
      "num_tokens": 62785354.0,
      "step": 24770
    },
    {
      "entropy": 6.21972804069519,
      "epoch": 2.859203693017888,
      "grad_norm": 1.703125,
      "learning_rate": 0.00041973887163016546,
      "loss": 5.8589,
      "mean_token_accuracy": 0.1828300714492798,
      "num_tokens": 62798356.0,
      "step": 24775
    },
    {
      "entropy": 6.135252046585083,
      "epoch": 2.8597807270628968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004197072760632147,
      "loss": 5.7418,
      "mean_token_accuracy": 0.19170485138893129,
      "num_tokens": 62809966.0,
      "step": 24780
    },
    {
      "entropy": 6.168164825439453,
      "epoch": 2.860357761107905,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041967567562910675,
      "loss": 5.7514,
      "mean_token_accuracy": 0.18814047425985336,
      "num_tokens": 62821833.0,
      "step": 24785
    },
    {
      "entropy": 6.176520299911499,
      "epoch": 2.860934795152914,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041964407032890426,
      "loss": 5.753,
      "mean_token_accuracy": 0.193022421002388,
      "num_tokens": 62833619.0,
      "step": 24790
    },
    {
      "entropy": 6.154019165039062,
      "epoch": 2.861511829197923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00041961246016367033,
      "loss": 5.7934,
      "mean_token_accuracy": 0.1848568081855774,
      "num_tokens": 62846373.0,
      "step": 24795
    },
    {
      "entropy": 6.189772129058838,
      "epoch": 2.8620888632429313,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041958084513446807,
      "loss": 5.8802,
      "mean_token_accuracy": 0.18780339807271956,
      "num_tokens": 62859311.0,
      "step": 24800
    },
    {
      "entropy": 6.194692850112915,
      "epoch": 2.8626658972879397,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000419549225242361,
      "loss": 5.8434,
      "mean_token_accuracy": 0.18864173144102098,
      "num_tokens": 62871824.0,
      "step": 24805
    },
    {
      "entropy": 6.156630945205689,
      "epoch": 2.8632429313329486,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004195176004884124,
      "loss": 5.7789,
      "mean_token_accuracy": 0.1891074389219284,
      "num_tokens": 62883795.0,
      "step": 24810
    },
    {
      "entropy": 6.197795820236206,
      "epoch": 2.8638199653779575,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004194859708736863,
      "loss": 5.8544,
      "mean_token_accuracy": 0.18445490747690202,
      "num_tokens": 62896331.0,
      "step": 24815
    },
    {
      "entropy": 6.237935161590576,
      "epoch": 2.864396999422966,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041945433639924617,
      "loss": 5.8694,
      "mean_token_accuracy": 0.18260905593633653,
      "num_tokens": 62907656.0,
      "step": 24820
    },
    {
      "entropy": 6.060537433624267,
      "epoch": 2.8649740334679747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041942269706615623,
      "loss": 5.7009,
      "mean_token_accuracy": 0.19327386766672133,
      "num_tokens": 62921132.0,
      "step": 24825
    },
    {
      "entropy": 6.110379600524903,
      "epoch": 2.865551067512983,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041939105287548055,
      "loss": 5.7741,
      "mean_token_accuracy": 0.19144950956106185,
      "num_tokens": 62933989.0,
      "step": 24830
    },
    {
      "entropy": 6.133320426940918,
      "epoch": 2.866128101557992,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041935940382828365,
      "loss": 5.8507,
      "mean_token_accuracy": 0.1785710334777832,
      "num_tokens": 62946284.0,
      "step": 24835
    },
    {
      "entropy": 6.146940135955811,
      "epoch": 2.8667051356030004,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004193277499256299,
      "loss": 5.8296,
      "mean_token_accuracy": 0.19064295440912246,
      "num_tokens": 62958690.0,
      "step": 24840
    },
    {
      "entropy": 6.231950187683106,
      "epoch": 2.8672821696480093,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041929609116858385,
      "loss": 5.8362,
      "mean_token_accuracy": 0.18415248095989228,
      "num_tokens": 62970374.0,
      "step": 24845
    },
    {
      "entropy": 6.094867515563965,
      "epoch": 2.8678592036930177,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041926442755821055,
      "loss": 5.7411,
      "mean_token_accuracy": 0.19062764048576356,
      "num_tokens": 62982245.0,
      "step": 24850
    },
    {
      "entropy": 6.138445854187012,
      "epoch": 2.8684362377380266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004192327590955748,
      "loss": 5.7563,
      "mean_token_accuracy": 0.18891474306583406,
      "num_tokens": 62995546.0,
      "step": 24855
    },
    {
      "entropy": 6.205418252944947,
      "epoch": 2.869013271783035,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041920108578174197,
      "loss": 5.8411,
      "mean_token_accuracy": 0.18665518164634703,
      "num_tokens": 63006937.0,
      "step": 24860
    },
    {
      "entropy": 6.2241579532623295,
      "epoch": 2.869590305828044,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004191694076177772,
      "loss": 5.747,
      "mean_token_accuracy": 0.18665951490402222,
      "num_tokens": 63019722.0,
      "step": 24865
    },
    {
      "entropy": 6.169949579238891,
      "epoch": 2.8701673398730527,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041913772460474603,
      "loss": 5.8733,
      "mean_token_accuracy": 0.18358385413885117,
      "num_tokens": 63031627.0,
      "step": 24870
    },
    {
      "entropy": 6.195762872695923,
      "epoch": 2.870744373918061,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041910603674371406,
      "loss": 5.8848,
      "mean_token_accuracy": 0.180596986413002,
      "num_tokens": 63043781.0,
      "step": 24875
    },
    {
      "entropy": 6.222542381286621,
      "epoch": 2.8713214079630696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004190743440357471,
      "loss": 5.7805,
      "mean_token_accuracy": 0.18950323909521102,
      "num_tokens": 63057196.0,
      "step": 24880
    },
    {
      "entropy": 6.131774806976319,
      "epoch": 2.8718984420080784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004190426464819112,
      "loss": 5.748,
      "mean_token_accuracy": 0.1905028373003006,
      "num_tokens": 63070013.0,
      "step": 24885
    },
    {
      "entropy": 6.189122009277344,
      "epoch": 2.8724754760530873,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004190109440832724,
      "loss": 5.8275,
      "mean_token_accuracy": 0.18342624455690384,
      "num_tokens": 63082733.0,
      "step": 24890
    },
    {
      "entropy": 6.245053148269653,
      "epoch": 2.8730525100980957,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041897923684089714,
      "loss": 5.8664,
      "mean_token_accuracy": 0.18298491835594177,
      "num_tokens": 63095275.0,
      "step": 24895
    },
    {
      "entropy": 6.156535053253174,
      "epoch": 2.8736295441431046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004189475247558516,
      "loss": 5.8121,
      "mean_token_accuracy": 0.18865811079740524,
      "num_tokens": 63108184.0,
      "step": 24900
    },
    {
      "entropy": 6.182178401947022,
      "epoch": 2.874206578188113,
      "grad_norm": 1.671875,
      "learning_rate": 0.00041891580782920263,
      "loss": 5.7241,
      "mean_token_accuracy": 0.18523198664188384,
      "num_tokens": 63120157.0,
      "step": 24905
    },
    {
      "entropy": 6.139427661895752,
      "epoch": 2.874783612233122,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041888408606201695,
      "loss": 5.7997,
      "mean_token_accuracy": 0.18744218498468398,
      "num_tokens": 63132390.0,
      "step": 24910
    },
    {
      "entropy": 6.125626707077027,
      "epoch": 2.8753606462781303,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004188523594553615,
      "loss": 5.8164,
      "mean_token_accuracy": 0.1867370069026947,
      "num_tokens": 63145567.0,
      "step": 24915
    },
    {
      "entropy": 6.212437915802002,
      "epoch": 2.875937680323139,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004188206280103034,
      "loss": 5.8159,
      "mean_token_accuracy": 0.18868955969810486,
      "num_tokens": 63157921.0,
      "step": 24920
    },
    {
      "entropy": 6.220640325546265,
      "epoch": 2.876514714368148,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00041878889172790995,
      "loss": 5.7672,
      "mean_token_accuracy": 0.19261593669652938,
      "num_tokens": 63169848.0,
      "step": 24925
    },
    {
      "entropy": 6.192274141311645,
      "epoch": 2.8770917484131564,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004187571506092485,
      "loss": 5.9087,
      "mean_token_accuracy": 0.18260124921798707,
      "num_tokens": 63182827.0,
      "step": 24930
    },
    {
      "entropy": 6.171605968475342,
      "epoch": 2.877668782458165,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004187254046553867,
      "loss": 5.8115,
      "mean_token_accuracy": 0.18643954247236252,
      "num_tokens": 63195636.0,
      "step": 24935
    },
    {
      "entropy": 6.2117541313171385,
      "epoch": 2.8782458165031737,
      "grad_norm": 1.1875,
      "learning_rate": 0.00041869365386739234,
      "loss": 5.8603,
      "mean_token_accuracy": 0.18119559735059737,
      "num_tokens": 63210045.0,
      "step": 24940
    },
    {
      "entropy": 6.231222820281983,
      "epoch": 2.8788228505481825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041866189824633335,
      "loss": 5.7923,
      "mean_token_accuracy": 0.1802116572856903,
      "num_tokens": 63223696.0,
      "step": 24945
    },
    {
      "entropy": 6.154783868789673,
      "epoch": 2.879399884593191,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004186301377932777,
      "loss": 5.7872,
      "mean_token_accuracy": 0.18940824270248413,
      "num_tokens": 63236325.0,
      "step": 24950
    },
    {
      "entropy": 6.149294233322143,
      "epoch": 2.8799769186381994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004185983725092937,
      "loss": 5.8537,
      "mean_token_accuracy": 0.18002549558877945,
      "num_tokens": 63249099.0,
      "step": 24955
    },
    {
      "entropy": 6.297961044311523,
      "epoch": 2.8805539526832082,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004185666023954498,
      "loss": 5.9022,
      "mean_token_accuracy": 0.18504872769117356,
      "num_tokens": 63261346.0,
      "step": 24960
    },
    {
      "entropy": 6.129104566574097,
      "epoch": 2.881130986728217,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004185348274528146,
      "loss": 5.7287,
      "mean_token_accuracy": 0.19116574078798293,
      "num_tokens": 63274574.0,
      "step": 24965
    },
    {
      "entropy": 6.100711250305176,
      "epoch": 2.8817080207732255,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041850304768245675,
      "loss": 5.7079,
      "mean_token_accuracy": 0.19610214680433274,
      "num_tokens": 63286708.0,
      "step": 24970
    },
    {
      "entropy": 6.199342584609985,
      "epoch": 2.8822850548182344,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004184712630854451,
      "loss": 5.7402,
      "mean_token_accuracy": 0.18573962301015853,
      "num_tokens": 63298769.0,
      "step": 24975
    },
    {
      "entropy": 6.1939269542694095,
      "epoch": 2.882862088863243,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004184394736628488,
      "loss": 5.8171,
      "mean_token_accuracy": 0.18373859226703643,
      "num_tokens": 63311228.0,
      "step": 24980
    },
    {
      "entropy": 6.223619508743286,
      "epoch": 2.8834391229082517,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004184076794157371,
      "loss": 5.8734,
      "mean_token_accuracy": 0.1806168407201767,
      "num_tokens": 63323450.0,
      "step": 24985
    },
    {
      "entropy": 6.152713680267334,
      "epoch": 2.88401615695326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004183758803451793,
      "loss": 5.7335,
      "mean_token_accuracy": 0.19684639424085618,
      "num_tokens": 63336948.0,
      "step": 24990
    },
    {
      "entropy": 6.186556482315064,
      "epoch": 2.884593190998269,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041834407645224505,
      "loss": 5.8128,
      "mean_token_accuracy": 0.18363571912050247,
      "num_tokens": 63348328.0,
      "step": 24995
    },
    {
      "entropy": 6.155422973632812,
      "epoch": 2.885170225043278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004183122677380039,
      "loss": 5.7555,
      "mean_token_accuracy": 0.18719779700040817,
      "num_tokens": 63359761.0,
      "step": 25000
    },
    {
      "entropy": 6.171518087387085,
      "epoch": 2.885747259088286,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041828045420352585,
      "loss": 5.789,
      "mean_token_accuracy": 0.1850059986114502,
      "num_tokens": 63371949.0,
      "step": 25005
    },
    {
      "entropy": 6.162282085418701,
      "epoch": 2.8863242931332946,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041824863584988086,
      "loss": 5.8306,
      "mean_token_accuracy": 0.1803960993885994,
      "num_tokens": 63385021.0,
      "step": 25010
    },
    {
      "entropy": 6.129031467437744,
      "epoch": 2.8869013271783035,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004182168126781392,
      "loss": 5.7323,
      "mean_token_accuracy": 0.19003615975379945,
      "num_tokens": 63398391.0,
      "step": 25015
    },
    {
      "entropy": 6.286478471755982,
      "epoch": 2.8874783612233124,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004181849846893712,
      "loss": 5.8697,
      "mean_token_accuracy": 0.17848527878522874,
      "num_tokens": 63411161.0,
      "step": 25020
    },
    {
      "entropy": 6.103060293197632,
      "epoch": 2.888055395268321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004181531518846474,
      "loss": 5.7708,
      "mean_token_accuracy": 0.1908055678009987,
      "num_tokens": 63424686.0,
      "step": 25025
    },
    {
      "entropy": 6.163610553741455,
      "epoch": 2.8886324293133296,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041812131426503834,
      "loss": 5.8309,
      "mean_token_accuracy": 0.18692435473203659,
      "num_tokens": 63437424.0,
      "step": 25030
    },
    {
      "entropy": 6.226094055175781,
      "epoch": 2.889209463358338,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004180894718316151,
      "loss": 5.8824,
      "mean_token_accuracy": 0.1801483854651451,
      "num_tokens": 63452194.0,
      "step": 25035
    },
    {
      "entropy": 6.224396705627441,
      "epoch": 2.889786497403347,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041805762458544846,
      "loss": 5.7795,
      "mean_token_accuracy": 0.1879183515906334,
      "num_tokens": 63465113.0,
      "step": 25040
    },
    {
      "entropy": 6.2037876605987545,
      "epoch": 2.8903635314483553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041802577252760963,
      "loss": 5.858,
      "mean_token_accuracy": 0.18454415649175643,
      "num_tokens": 63477791.0,
      "step": 25045
    },
    {
      "entropy": 6.176325798034668,
      "epoch": 2.890940565493364,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041799391565917,
      "loss": 5.8111,
      "mean_token_accuracy": 0.18566223084926606,
      "num_tokens": 63490280.0,
      "step": 25050
    },
    {
      "entropy": 6.242665338516235,
      "epoch": 2.8915175995383726,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041796205398120115,
      "loss": 5.8352,
      "mean_token_accuracy": 0.18668083399534224,
      "num_tokens": 63502617.0,
      "step": 25055
    },
    {
      "entropy": 6.117324352264404,
      "epoch": 2.8920946335833815,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041793018749477454,
      "loss": 5.64,
      "mean_token_accuracy": 0.19256674945354463,
      "num_tokens": 63515209.0,
      "step": 25060
    },
    {
      "entropy": 6.201242065429687,
      "epoch": 2.89267166762839,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004178983162009621,
      "loss": 5.8782,
      "mean_token_accuracy": 0.18686506897211075,
      "num_tokens": 63527870.0,
      "step": 25065
    },
    {
      "entropy": 6.1623327255249025,
      "epoch": 2.8932487016733988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004178664401008358,
      "loss": 5.7887,
      "mean_token_accuracy": 0.18549591600894927,
      "num_tokens": 63541187.0,
      "step": 25070
    },
    {
      "entropy": 6.197692775726319,
      "epoch": 2.8938257357184076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041783455919546767,
      "loss": 5.7883,
      "mean_token_accuracy": 0.1858491361141205,
      "num_tokens": 63553892.0,
      "step": 25075
    },
    {
      "entropy": 6.131242895126343,
      "epoch": 2.894402769763416,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041780267348593016,
      "loss": 5.7234,
      "mean_token_accuracy": 0.19532606154680252,
      "num_tokens": 63565149.0,
      "step": 25080
    },
    {
      "entropy": 6.114775514602661,
      "epoch": 2.8949798038084245,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004177707829732956,
      "loss": 5.7759,
      "mean_token_accuracy": 0.1934100180864334,
      "num_tokens": 63577114.0,
      "step": 25085
    },
    {
      "entropy": 6.142133188247681,
      "epoch": 2.8955568378534333,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041773888765863677,
      "loss": 5.7239,
      "mean_token_accuracy": 0.19358953088521957,
      "num_tokens": 63589689.0,
      "step": 25090
    },
    {
      "entropy": 6.262591075897217,
      "epoch": 2.896133871898442,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004177069875430263,
      "loss": 5.8313,
      "mean_token_accuracy": 0.1825080171227455,
      "num_tokens": 63601888.0,
      "step": 25095
    },
    {
      "entropy": 6.180368328094483,
      "epoch": 2.8967109059434506,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004176750826275372,
      "loss": 5.8646,
      "mean_token_accuracy": 0.17671932876110077,
      "num_tokens": 63614663.0,
      "step": 25100
    },
    {
      "entropy": 6.135539102554321,
      "epoch": 2.8972879399884595,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004176431729132426,
      "loss": 5.7626,
      "mean_token_accuracy": 0.18939124047756195,
      "num_tokens": 63626191.0,
      "step": 25105
    },
    {
      "entropy": 6.164480209350586,
      "epoch": 2.897864974033468,
      "grad_norm": 1.0,
      "learning_rate": 0.00041761125840121565,
      "loss": 5.7858,
      "mean_token_accuracy": 0.18773820847272873,
      "num_tokens": 63638778.0,
      "step": 25110
    },
    {
      "entropy": 6.133819675445556,
      "epoch": 2.8984420080784767,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004175793390925299,
      "loss": 5.6957,
      "mean_token_accuracy": 0.1860012397170067,
      "num_tokens": 63653369.0,
      "step": 25115
    },
    {
      "entropy": 6.226192331314087,
      "epoch": 2.899019042123485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004175474149882589,
      "loss": 5.7866,
      "mean_token_accuracy": 0.18243784159421922,
      "num_tokens": 63665442.0,
      "step": 25120
    },
    {
      "entropy": 6.103700160980225,
      "epoch": 2.899596076168494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004175154860894765,
      "loss": 5.7158,
      "mean_token_accuracy": 0.2028720647096634,
      "num_tokens": 63678221.0,
      "step": 25125
    },
    {
      "entropy": 6.091815423965454,
      "epoch": 2.9001731102135024,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004174835523972563,
      "loss": 5.7016,
      "mean_token_accuracy": 0.19763491600751876,
      "num_tokens": 63690619.0,
      "step": 25130
    },
    {
      "entropy": 6.127394723892212,
      "epoch": 2.9007501442585113,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00041745161391267276,
      "loss": 5.7686,
      "mean_token_accuracy": 0.190999336540699,
      "num_tokens": 63703742.0,
      "step": 25135
    },
    {
      "entropy": 6.158084869384766,
      "epoch": 2.9013271783035197,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041741967063679993,
      "loss": 5.7971,
      "mean_token_accuracy": 0.182604843378067,
      "num_tokens": 63716929.0,
      "step": 25140
    },
    {
      "entropy": 6.11389274597168,
      "epoch": 2.9019042123485286,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041738772257071216,
      "loss": 5.7366,
      "mean_token_accuracy": 0.18962296694517136,
      "num_tokens": 63730301.0,
      "step": 25145
    },
    {
      "entropy": 6.1299182891845705,
      "epoch": 2.9024812463935374,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004173557697154841,
      "loss": 5.6606,
      "mean_token_accuracy": 0.1935422122478485,
      "num_tokens": 63744294.0,
      "step": 25150
    },
    {
      "entropy": 6.182842779159546,
      "epoch": 2.903058280438546,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041732381207219046,
      "loss": 5.8583,
      "mean_token_accuracy": 0.17903222292661666,
      "num_tokens": 63756634.0,
      "step": 25155
    },
    {
      "entropy": 6.194612741470337,
      "epoch": 2.9036353144835543,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041729184964190607,
      "loss": 5.7642,
      "mean_token_accuracy": 0.18766499906778336,
      "num_tokens": 63768920.0,
      "step": 25160
    },
    {
      "entropy": 6.1540539264678955,
      "epoch": 2.904212348528563,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041725988242570603,
      "loss": 5.7271,
      "mean_token_accuracy": 0.19254759401082994,
      "num_tokens": 63781315.0,
      "step": 25165
    },
    {
      "entropy": 6.171711206436157,
      "epoch": 2.904789382573572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041722791042466545,
      "loss": 5.8349,
      "mean_token_accuracy": 0.18272597938776017,
      "num_tokens": 63793453.0,
      "step": 25170
    },
    {
      "entropy": 6.258182001113892,
      "epoch": 2.9053664166185804,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004171959336398598,
      "loss": 5.8302,
      "mean_token_accuracy": 0.18577373325824736,
      "num_tokens": 63806929.0,
      "step": 25175
    },
    {
      "entropy": 6.104452896118164,
      "epoch": 2.9059434506635893,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004171639520723646,
      "loss": 5.7474,
      "mean_token_accuracy": 0.19049512594938278,
      "num_tokens": 63819794.0,
      "step": 25180
    },
    {
      "entropy": 6.116134595870972,
      "epoch": 2.9065204847085977,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004171319657232554,
      "loss": 5.6947,
      "mean_token_accuracy": 0.1970692068338394,
      "num_tokens": 63831494.0,
      "step": 25185
    },
    {
      "entropy": 6.144589519500732,
      "epoch": 2.9070975187536066,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004170999745936082,
      "loss": 5.7827,
      "mean_token_accuracy": 0.18278268724679947,
      "num_tokens": 63844472.0,
      "step": 25190
    },
    {
      "entropy": 6.1718353748321535,
      "epoch": 2.907674552798615,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004170679786844989,
      "loss": 5.7711,
      "mean_token_accuracy": 0.18623567074537278,
      "num_tokens": 63856028.0,
      "step": 25195
    },
    {
      "entropy": 6.118298292160034,
      "epoch": 2.908251586843624,
      "grad_norm": 1.125,
      "learning_rate": 0.0004170359779970037,
      "loss": 5.7784,
      "mean_token_accuracy": 0.17982750236988068,
      "num_tokens": 63870137.0,
      "step": 25200
    },
    {
      "entropy": 6.2560032367706295,
      "epoch": 2.9088286208886327,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004170039725321989,
      "loss": 5.8316,
      "mean_token_accuracy": 0.1833975613117218,
      "num_tokens": 63882152.0,
      "step": 25205
    },
    {
      "entropy": 6.146197700500489,
      "epoch": 2.909405654933641,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041697196229116106,
      "loss": 5.7697,
      "mean_token_accuracy": 0.18322059661149978,
      "num_tokens": 63894624.0,
      "step": 25210
    },
    {
      "entropy": 6.017905426025391,
      "epoch": 2.9099826889786495,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004169399472749668,
      "loss": 5.5523,
      "mean_token_accuracy": 0.20819265246391297,
      "num_tokens": 63907439.0,
      "step": 25215
    },
    {
      "entropy": 6.194882583618164,
      "epoch": 2.9105597230236584,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000416907927484693,
      "loss": 5.8419,
      "mean_token_accuracy": 0.18219007104635238,
      "num_tokens": 63920432.0,
      "step": 25220
    },
    {
      "entropy": 6.209870100021362,
      "epoch": 2.9111367570686673,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041687590292141644,
      "loss": 5.764,
      "mean_token_accuracy": 0.1942257508635521,
      "num_tokens": 63935682.0,
      "step": 25225
    },
    {
      "entropy": 6.073653507232666,
      "epoch": 2.9117137911136757,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00041684387358621435,
      "loss": 5.7607,
      "mean_token_accuracy": 0.20043881237506866,
      "num_tokens": 63949109.0,
      "step": 25230
    },
    {
      "entropy": 6.208702230453492,
      "epoch": 2.912290825158684,
      "grad_norm": 1.140625,
      "learning_rate": 0.00041681183948016404,
      "loss": 5.7593,
      "mean_token_accuracy": 0.18617864996194838,
      "num_tokens": 63962859.0,
      "step": 25235
    },
    {
      "entropy": 6.241482782363891,
      "epoch": 2.912867859203693,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000416779800604343,
      "loss": 5.889,
      "mean_token_accuracy": 0.17649127393960953,
      "num_tokens": 63975641.0,
      "step": 25240
    },
    {
      "entropy": 6.192310428619384,
      "epoch": 2.913444893248702,
      "grad_norm": 0.859375,
      "learning_rate": 0.00041674775695982883,
      "loss": 5.8007,
      "mean_token_accuracy": 0.18588081002235413,
      "num_tokens": 63988419.0,
      "step": 25245
    },
    {
      "entropy": 6.142047166824341,
      "epoch": 2.9140219272937102,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004167157085476992,
      "loss": 5.7904,
      "mean_token_accuracy": 0.18138567954301835,
      "num_tokens": 64000217.0,
      "step": 25250
    },
    {
      "entropy": 6.193149423599243,
      "epoch": 2.914598961338719,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004166836553690321,
      "loss": 5.8271,
      "mean_token_accuracy": 0.1848229244351387,
      "num_tokens": 64013083.0,
      "step": 25255
    },
    {
      "entropy": 6.105286693572998,
      "epoch": 2.9151759953837275,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004166515974249056,
      "loss": 5.7222,
      "mean_token_accuracy": 0.1918478712439537,
      "num_tokens": 64025769.0,
      "step": 25260
    },
    {
      "entropy": 6.225525331497193,
      "epoch": 2.9157530294287364,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041661953471639797,
      "loss": 5.7891,
      "mean_token_accuracy": 0.1853489100933075,
      "num_tokens": 64037526.0,
      "step": 25265
    },
    {
      "entropy": 6.127283811569214,
      "epoch": 2.916330063473745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041658746724458775,
      "loss": 5.7521,
      "mean_token_accuracy": 0.19490110874176025,
      "num_tokens": 64050542.0,
      "step": 25270
    },
    {
      "entropy": 6.165182209014892,
      "epoch": 2.9169070975187537,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041655539501055333,
      "loss": 5.6944,
      "mean_token_accuracy": 0.1959969386458397,
      "num_tokens": 64063097.0,
      "step": 25275
    },
    {
      "entropy": 6.180014944076538,
      "epoch": 2.9174841315637625,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004165233180153735,
      "loss": 5.8078,
      "mean_token_accuracy": 0.18777673840522766,
      "num_tokens": 64076104.0,
      "step": 25280
    },
    {
      "entropy": 6.1819206237792965,
      "epoch": 2.918061165608771,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004164912362601272,
      "loss": 5.791,
      "mean_token_accuracy": 0.18828150480985642,
      "num_tokens": 64090351.0,
      "step": 25285
    },
    {
      "entropy": 6.089034175872802,
      "epoch": 2.9186381996537794,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041645914974589347,
      "loss": 5.6435,
      "mean_token_accuracy": 0.19078212827444077,
      "num_tokens": 64103408.0,
      "step": 25290
    },
    {
      "entropy": 6.064705944061279,
      "epoch": 2.919215233698788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041642705847375137,
      "loss": 5.6788,
      "mean_token_accuracy": 0.19690516591072083,
      "num_tokens": 64116733.0,
      "step": 25295
    },
    {
      "entropy": 6.123806381225586,
      "epoch": 2.919792267743797,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004163949624447806,
      "loss": 5.7584,
      "mean_token_accuracy": 0.19039979279041291,
      "num_tokens": 64129040.0,
      "step": 25300
    },
    {
      "entropy": 6.1485144138336185,
      "epoch": 2.9203693017888055,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041636286166006045,
      "loss": 5.7789,
      "mean_token_accuracy": 0.1803423896431923,
      "num_tokens": 64140437.0,
      "step": 25305
    },
    {
      "entropy": 6.081819486618042,
      "epoch": 2.9209463358338144,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004163307561206706,
      "loss": 5.6848,
      "mean_token_accuracy": 0.19848719239234924,
      "num_tokens": 64153771.0,
      "step": 25310
    },
    {
      "entropy": 6.139738512039185,
      "epoch": 2.9215233698788228,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000416298645827691,
      "loss": 5.7579,
      "mean_token_accuracy": 0.18604142516851424,
      "num_tokens": 64165492.0,
      "step": 25315
    },
    {
      "entropy": 6.2124255180358885,
      "epoch": 2.9221004039238316,
      "grad_norm": 1.0,
      "learning_rate": 0.0004162665307822016,
      "loss": 5.7815,
      "mean_token_accuracy": 0.1850111499428749,
      "num_tokens": 64178898.0,
      "step": 25320
    },
    {
      "entropy": 6.2393402576446535,
      "epoch": 2.92267743796884,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00041623441098528267,
      "loss": 5.8291,
      "mean_token_accuracy": 0.18239922523498536,
      "num_tokens": 64190529.0,
      "step": 25325
    },
    {
      "entropy": 6.213286304473877,
      "epoch": 2.923254472013849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041620228643801443,
      "loss": 5.7182,
      "mean_token_accuracy": 0.1871095210313797,
      "num_tokens": 64203073.0,
      "step": 25330
    },
    {
      "entropy": 6.192132139205933,
      "epoch": 2.9238315060588573,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004161701571414775,
      "loss": 5.8446,
      "mean_token_accuracy": 0.1829073280096054,
      "num_tokens": 64216103.0,
      "step": 25335
    },
    {
      "entropy": 6.195889139175415,
      "epoch": 2.924408540103866,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041613802309675235,
      "loss": 5.8485,
      "mean_token_accuracy": 0.18504202216863633,
      "num_tokens": 64228451.0,
      "step": 25340
    },
    {
      "entropy": 6.136723518371582,
      "epoch": 2.9249855741488746,
      "grad_norm": 1.4375,
      "learning_rate": 0.00041610588430492,
      "loss": 5.7099,
      "mean_token_accuracy": 0.19221784323453903,
      "num_tokens": 64239858.0,
      "step": 25345
    },
    {
      "entropy": 6.118024206161499,
      "epoch": 2.9255626081938835,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041607374076706123,
      "loss": 5.7102,
      "mean_token_accuracy": 0.19761201739311218,
      "num_tokens": 64252150.0,
      "step": 25350
    },
    {
      "entropy": 6.271979570388794,
      "epoch": 2.9261396422388923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041604159248425737,
      "loss": 5.9261,
      "mean_token_accuracy": 0.1815897911787033,
      "num_tokens": 64264600.0,
      "step": 25355
    },
    {
      "entropy": 6.208084678649902,
      "epoch": 2.9267166762839008,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004160094394575896,
      "loss": 5.7605,
      "mean_token_accuracy": 0.18767553120851516,
      "num_tokens": 64276533.0,
      "step": 25360
    },
    {
      "entropy": 6.174940776824951,
      "epoch": 2.927293710328909,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004159772816881393,
      "loss": 5.7561,
      "mean_token_accuracy": 0.19796690642833709,
      "num_tokens": 64288628.0,
      "step": 25365
    },
    {
      "entropy": 6.180430889129639,
      "epoch": 2.927870744373918,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00041594511917698813,
      "loss": 5.7819,
      "mean_token_accuracy": 0.19161266684532166,
      "num_tokens": 64301584.0,
      "step": 25370
    },
    {
      "entropy": 6.164112091064453,
      "epoch": 2.928447778418927,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041591295192521796,
      "loss": 5.8281,
      "mean_token_accuracy": 0.18357709646224976,
      "num_tokens": 64313363.0,
      "step": 25375
    },
    {
      "entropy": 6.125766372680664,
      "epoch": 2.9290248124639353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004158807799339107,
      "loss": 5.7293,
      "mean_token_accuracy": 0.19762576073408128,
      "num_tokens": 64325284.0,
      "step": 25380
    },
    {
      "entropy": 6.22836012840271,
      "epoch": 2.929601846508944,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041584860320414827,
      "loss": 5.8249,
      "mean_token_accuracy": 0.18303842544555665,
      "num_tokens": 64336886.0,
      "step": 25385
    },
    {
      "entropy": 6.172658538818359,
      "epoch": 2.9301788805539526,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041581642173701305,
      "loss": 5.7617,
      "mean_token_accuracy": 0.18735713511705399,
      "num_tokens": 64349299.0,
      "step": 25390
    },
    {
      "entropy": 6.0893817901611325,
      "epoch": 2.9307559145989615,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041578423553358747,
      "loss": 5.7768,
      "mean_token_accuracy": 0.18666261583566665,
      "num_tokens": 64360997.0,
      "step": 25395
    },
    {
      "entropy": 6.207407903671265,
      "epoch": 2.93133294864397,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00041575204459495406,
      "loss": 5.82,
      "mean_token_accuracy": 0.1894814893603325,
      "num_tokens": 64375083.0,
      "step": 25400
    },
    {
      "entropy": 6.165468978881836,
      "epoch": 2.9319099826889787,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041571984892219556,
      "loss": 5.8625,
      "mean_token_accuracy": 0.1828040137887001,
      "num_tokens": 64387882.0,
      "step": 25405
    },
    {
      "entropy": 6.068056869506836,
      "epoch": 2.932487016733987,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004156876485163948,
      "loss": 5.6644,
      "mean_token_accuracy": 0.19635545760393142,
      "num_tokens": 64399876.0,
      "step": 25410
    },
    {
      "entropy": 6.147058820724487,
      "epoch": 2.933064050778996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041565544337863493,
      "loss": 5.7543,
      "mean_token_accuracy": 0.19321614056825637,
      "num_tokens": 64411479.0,
      "step": 25415
    },
    {
      "entropy": 6.110343503952026,
      "epoch": 2.9336410848240044,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000415623233509999,
      "loss": 5.7264,
      "mean_token_accuracy": 0.19031729251146318,
      "num_tokens": 64425009.0,
      "step": 25420
    },
    {
      "entropy": 6.072913074493409,
      "epoch": 2.9342181188690133,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004155910189115705,
      "loss": 5.7416,
      "mean_token_accuracy": 0.19084058105945587,
      "num_tokens": 64436635.0,
      "step": 25425
    },
    {
      "entropy": 6.064367866516113,
      "epoch": 2.934795152914022,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041555879958443296,
      "loss": 5.6888,
      "mean_token_accuracy": 0.19550212770700454,
      "num_tokens": 64449338.0,
      "step": 25430
    },
    {
      "entropy": 6.207567739486694,
      "epoch": 2.9353721869590306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041552657552967,
      "loss": 5.8814,
      "mean_token_accuracy": 0.1787579044699669,
      "num_tokens": 64461124.0,
      "step": 25435
    },
    {
      "entropy": 6.271834039688111,
      "epoch": 2.935949221004039,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004154943467483654,
      "loss": 5.8426,
      "mean_token_accuracy": 0.19018956869840623,
      "num_tokens": 64474093.0,
      "step": 25440
    },
    {
      "entropy": 6.188777160644531,
      "epoch": 2.936526255049048,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004154621132416033,
      "loss": 5.7508,
      "mean_token_accuracy": 0.18046221733093262,
      "num_tokens": 64486408.0,
      "step": 25445
    },
    {
      "entropy": 6.0724719047546385,
      "epoch": 2.9371032890940567,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041542987501046777,
      "loss": 5.7547,
      "mean_token_accuracy": 0.19533833414316176,
      "num_tokens": 64499390.0,
      "step": 25450
    },
    {
      "entropy": 6.1832510471344,
      "epoch": 2.937680323139065,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041539763205604316,
      "loss": 5.746,
      "mean_token_accuracy": 0.1905984714627266,
      "num_tokens": 64511307.0,
      "step": 25455
    },
    {
      "entropy": 6.153995132446289,
      "epoch": 2.938257357184074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004153653843794139,
      "loss": 5.7321,
      "mean_token_accuracy": 0.19005557894706726,
      "num_tokens": 64523027.0,
      "step": 25460
    },
    {
      "entropy": 6.180675983428955,
      "epoch": 2.9388343912290824,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004153331319816647,
      "loss": 5.9144,
      "mean_token_accuracy": 0.17679677456617354,
      "num_tokens": 64535472.0,
      "step": 25465
    },
    {
      "entropy": 6.186362266540527,
      "epoch": 2.9394114252740913,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004153008748638803,
      "loss": 5.7573,
      "mean_token_accuracy": 0.18966278731822966,
      "num_tokens": 64547935.0,
      "step": 25470
    },
    {
      "entropy": 6.203519296646118,
      "epoch": 2.9399884593190997,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004152686130271457,
      "loss": 5.8733,
      "mean_token_accuracy": 0.18067746311426164,
      "num_tokens": 64561773.0,
      "step": 25475
    },
    {
      "entropy": 6.181817436218262,
      "epoch": 2.9405654933641086,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041523634647254595,
      "loss": 5.7545,
      "mean_token_accuracy": 0.1912101075053215,
      "num_tokens": 64576465.0,
      "step": 25480
    },
    {
      "entropy": 6.133107423782349,
      "epoch": 2.9411425274091174,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041520407520116633,
      "loss": 5.7739,
      "mean_token_accuracy": 0.1880086123943329,
      "num_tokens": 64589890.0,
      "step": 25485
    },
    {
      "entropy": 6.127618932723999,
      "epoch": 2.941719561454126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004151717992140923,
      "loss": 5.7295,
      "mean_token_accuracy": 0.19406392723321914,
      "num_tokens": 64601423.0,
      "step": 25490
    },
    {
      "entropy": 6.177945470809936,
      "epoch": 2.9422965954991342,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041513951851240947,
      "loss": 5.8098,
      "mean_token_accuracy": 0.1866205230355263,
      "num_tokens": 64614532.0,
      "step": 25495
    },
    {
      "entropy": 6.154408597946167,
      "epoch": 2.942873629544143,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041510723309720344,
      "loss": 5.7756,
      "mean_token_accuracy": 0.1892290785908699,
      "num_tokens": 64628260.0,
      "step": 25500
    },
    {
      "entropy": 6.171898221969604,
      "epoch": 2.943450663589152,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004150749429695603,
      "loss": 5.789,
      "mean_token_accuracy": 0.18955737650394439,
      "num_tokens": 64640913.0,
      "step": 25505
    },
    {
      "entropy": 6.085756587982178,
      "epoch": 2.9440276976341604,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041504264813056605,
      "loss": 5.7245,
      "mean_token_accuracy": 0.19082649946212768,
      "num_tokens": 64653438.0,
      "step": 25510
    },
    {
      "entropy": 6.114610624313355,
      "epoch": 2.944604731679169,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041501034858130683,
      "loss": 5.6753,
      "mean_token_accuracy": 0.19089840203523636,
      "num_tokens": 64665753.0,
      "step": 25515
    },
    {
      "entropy": 6.235386991500855,
      "epoch": 2.9451817657241777,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041497804432286914,
      "loss": 5.8835,
      "mean_token_accuracy": 0.1835867464542389,
      "num_tokens": 64676872.0,
      "step": 25520
    },
    {
      "entropy": 6.075920820236206,
      "epoch": 2.9457587997691865,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004149457353563394,
      "loss": 5.7124,
      "mean_token_accuracy": 0.20068821758031846,
      "num_tokens": 64690216.0,
      "step": 25525
    },
    {
      "entropy": 6.135827112197876,
      "epoch": 2.946335833814195,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00041491342168280444,
      "loss": 5.7751,
      "mean_token_accuracy": 0.19033685475587844,
      "num_tokens": 64704210.0,
      "step": 25530
    },
    {
      "entropy": 6.237697887420654,
      "epoch": 2.946912867859204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041488110330335093,
      "loss": 5.7925,
      "mean_token_accuracy": 0.18624559342861174,
      "num_tokens": 64717142.0,
      "step": 25535
    },
    {
      "entropy": 6.102040576934814,
      "epoch": 2.9474899019042122,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041484878021906605,
      "loss": 5.6541,
      "mean_token_accuracy": 0.19631390869617463,
      "num_tokens": 64729783.0,
      "step": 25540
    },
    {
      "entropy": 6.130129957199097,
      "epoch": 2.948066935949221,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041481645243103695,
      "loss": 5.8099,
      "mean_token_accuracy": 0.1839154615998268,
      "num_tokens": 64742179.0,
      "step": 25545
    },
    {
      "entropy": 6.236168098449707,
      "epoch": 2.9486439699942295,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041478411994035086,
      "loss": 5.8002,
      "mean_token_accuracy": 0.18524857461452485,
      "num_tokens": 64754925.0,
      "step": 25550
    },
    {
      "entropy": 6.103754138946533,
      "epoch": 2.9492210040392384,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041475178274809544,
      "loss": 5.7593,
      "mean_token_accuracy": 0.19133903086185455,
      "num_tokens": 64768341.0,
      "step": 25555
    },
    {
      "entropy": 6.17221040725708,
      "epoch": 2.9497980380842472,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041471944085535805,
      "loss": 5.8188,
      "mean_token_accuracy": 0.18680384755134583,
      "num_tokens": 64780780.0,
      "step": 25560
    },
    {
      "entropy": 6.236133241653443,
      "epoch": 2.9503750721292556,
      "grad_norm": 1.0,
      "learning_rate": 0.0004146870942632268,
      "loss": 5.8429,
      "mean_token_accuracy": 0.18068062663078308,
      "num_tokens": 64792495.0,
      "step": 25565
    },
    {
      "entropy": 6.240812015533447,
      "epoch": 2.950952106174264,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041465474297278955,
      "loss": 5.7729,
      "mean_token_accuracy": 0.19019680470228195,
      "num_tokens": 64804696.0,
      "step": 25570
    },
    {
      "entropy": 6.125988340377807,
      "epoch": 2.951529140219273,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004146223869851343,
      "loss": 5.7084,
      "mean_token_accuracy": 0.19415107518434524,
      "num_tokens": 64817146.0,
      "step": 25575
    },
    {
      "entropy": 6.111772966384888,
      "epoch": 2.952106174264282,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004145900263013494,
      "loss": 5.7821,
      "mean_token_accuracy": 0.18447566628456116,
      "num_tokens": 64828833.0,
      "step": 25580
    },
    {
      "entropy": 6.1927549839019775,
      "epoch": 2.95268320830929,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004145576609225234,
      "loss": 5.7733,
      "mean_token_accuracy": 0.1915595769882202,
      "num_tokens": 64843811.0,
      "step": 25585
    },
    {
      "entropy": 6.073775911331177,
      "epoch": 2.953260242354299,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004145252908497448,
      "loss": 5.6878,
      "mean_token_accuracy": 0.1998982161283493,
      "num_tokens": 64856829.0,
      "step": 25590
    },
    {
      "entropy": 6.034066200256348,
      "epoch": 2.9538372763993075,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041449291608410227,
      "loss": 5.6365,
      "mean_token_accuracy": 0.1892428144812584,
      "num_tokens": 64869044.0,
      "step": 25595
    },
    {
      "entropy": 6.217815971374511,
      "epoch": 2.9544143104443163,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041446053662668483,
      "loss": 5.8742,
      "mean_token_accuracy": 0.18574898391962053,
      "num_tokens": 64882330.0,
      "step": 25600
    },
    {
      "entropy": 6.197320604324341,
      "epoch": 2.9549913444893248,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004144281524785815,
      "loss": 5.7815,
      "mean_token_accuracy": 0.19254770427942275,
      "num_tokens": 64894609.0,
      "step": 25605
    },
    {
      "entropy": 6.07693452835083,
      "epoch": 2.9555683785343336,
      "grad_norm": 1.0,
      "learning_rate": 0.0004143957636408815,
      "loss": 5.7649,
      "mean_token_accuracy": 0.19174301326274873,
      "num_tokens": 64907251.0,
      "step": 25610
    },
    {
      "entropy": 6.2032591819763185,
      "epoch": 2.956145412579342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041436337011467426,
      "loss": 5.7889,
      "mean_token_accuracy": 0.19160752147436141,
      "num_tokens": 64919114.0,
      "step": 25615
    },
    {
      "entropy": 6.18372368812561,
      "epoch": 2.956722446624351,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041433097190104936,
      "loss": 5.8438,
      "mean_token_accuracy": 0.17706149071455002,
      "num_tokens": 64933064.0,
      "step": 25620
    },
    {
      "entropy": 6.16905460357666,
      "epoch": 2.9572994806693593,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041429856900109636,
      "loss": 5.7907,
      "mean_token_accuracy": 0.19139077961444856,
      "num_tokens": 64946625.0,
      "step": 25625
    },
    {
      "entropy": 6.129648017883301,
      "epoch": 2.957876514714368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004142661614159051,
      "loss": 5.662,
      "mean_token_accuracy": 0.19387938231229782,
      "num_tokens": 64958577.0,
      "step": 25630
    },
    {
      "entropy": 6.179660940170288,
      "epoch": 2.958453548759377,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041423374914656587,
      "loss": 5.8645,
      "mean_token_accuracy": 0.1814304083585739,
      "num_tokens": 64971588.0,
      "step": 25635
    },
    {
      "entropy": 6.174642705917359,
      "epoch": 2.9590305828043855,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004142013321941685,
      "loss": 5.8224,
      "mean_token_accuracy": 0.18751190453767777,
      "num_tokens": 64984448.0,
      "step": 25640
    },
    {
      "entropy": 6.168317461013794,
      "epoch": 2.959607616849394,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041416891055980347,
      "loss": 5.768,
      "mean_token_accuracy": 0.1889842540025711,
      "num_tokens": 64996759.0,
      "step": 25645
    },
    {
      "entropy": 6.2113720893859865,
      "epoch": 2.9601846508944027,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041413648424456134,
      "loss": 5.829,
      "mean_token_accuracy": 0.18671383410692216,
      "num_tokens": 65009732.0,
      "step": 25650
    },
    {
      "entropy": 6.182887029647827,
      "epoch": 2.9607616849394116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041410405324953265,
      "loss": 5.864,
      "mean_token_accuracy": 0.1773914724588394,
      "num_tokens": 65021479.0,
      "step": 25655
    },
    {
      "entropy": 6.23884220123291,
      "epoch": 2.96133871898442,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004140716175758082,
      "loss": 5.8286,
      "mean_token_accuracy": 0.18784796446561813,
      "num_tokens": 65034782.0,
      "step": 25660
    },
    {
      "entropy": 6.155757141113281,
      "epoch": 2.961915753029429,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00041403917722447896,
      "loss": 5.814,
      "mean_token_accuracy": 0.18590835481882095,
      "num_tokens": 65047427.0,
      "step": 25665
    },
    {
      "entropy": 6.166055059432983,
      "epoch": 2.9624927870744373,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004140067321966361,
      "loss": 5.7648,
      "mean_token_accuracy": 0.19290081411600113,
      "num_tokens": 65060867.0,
      "step": 25670
    },
    {
      "entropy": 6.143200159072876,
      "epoch": 2.963069821119446,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004139742824933709,
      "loss": 5.7128,
      "mean_token_accuracy": 0.19452529698610305,
      "num_tokens": 65074394.0,
      "step": 25675
    },
    {
      "entropy": 6.074012660980225,
      "epoch": 2.9636468551644546,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004139418281157747,
      "loss": 5.71,
      "mean_token_accuracy": 0.1997675970196724,
      "num_tokens": 65088352.0,
      "step": 25680
    },
    {
      "entropy": 6.175407838821411,
      "epoch": 2.9642238892094634,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004139093690649391,
      "loss": 5.7782,
      "mean_token_accuracy": 0.19033325463533401,
      "num_tokens": 65101113.0,
      "step": 25685
    },
    {
      "entropy": 6.19824628829956,
      "epoch": 2.964800923254472,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041387690534195584,
      "loss": 5.7832,
      "mean_token_accuracy": 0.18356478810310364,
      "num_tokens": 65112894.0,
      "step": 25690
    },
    {
      "entropy": 6.150338745117187,
      "epoch": 2.9653779572994807,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004138444369479169,
      "loss": 5.7755,
      "mean_token_accuracy": 0.18152814209461213,
      "num_tokens": 65125087.0,
      "step": 25695
    },
    {
      "entropy": 6.161756467819214,
      "epoch": 2.965954991344489,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004138119638839143,
      "loss": 5.7422,
      "mean_token_accuracy": 0.19545589983463288,
      "num_tokens": 65138345.0,
      "step": 25700
    },
    {
      "entropy": 6.206123399734497,
      "epoch": 2.966532025389498,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041377948615104024,
      "loss": 5.7826,
      "mean_token_accuracy": 0.18727729320526124,
      "num_tokens": 65151337.0,
      "step": 25705
    },
    {
      "entropy": 6.024678421020508,
      "epoch": 2.967109059434507,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004137470037503871,
      "loss": 5.6566,
      "mean_token_accuracy": 0.1973340168595314,
      "num_tokens": 65163061.0,
      "step": 25710
    },
    {
      "entropy": 6.185911798477173,
      "epoch": 2.9676860934795153,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004137145166830474,
      "loss": 5.7662,
      "mean_token_accuracy": 0.18733941167593002,
      "num_tokens": 65176358.0,
      "step": 25715
    },
    {
      "entropy": 6.182557010650635,
      "epoch": 2.9682631275245237,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004136820249501139,
      "loss": 5.7879,
      "mean_token_accuracy": 0.1872107282280922,
      "num_tokens": 65188189.0,
      "step": 25720
    },
    {
      "entropy": 6.17884578704834,
      "epoch": 2.9688401615695326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004136495285526792,
      "loss": 5.7779,
      "mean_token_accuracy": 0.18638554960489273,
      "num_tokens": 65200748.0,
      "step": 25725
    },
    {
      "entropy": 6.172512245178223,
      "epoch": 2.9694171956145414,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041361702749183663,
      "loss": 5.7582,
      "mean_token_accuracy": 0.19288298785686492,
      "num_tokens": 65212871.0,
      "step": 25730
    },
    {
      "entropy": 6.083426523208618,
      "epoch": 2.96999422965955,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004135845217686791,
      "loss": 5.6906,
      "mean_token_accuracy": 0.20035747587680816,
      "num_tokens": 65224604.0,
      "step": 25735
    },
    {
      "entropy": 6.139448785781861,
      "epoch": 2.9705712637045587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004135520113843001,
      "loss": 5.7365,
      "mean_token_accuracy": 0.19355759024620056,
      "num_tokens": 65236757.0,
      "step": 25740
    },
    {
      "entropy": 6.210270118713379,
      "epoch": 2.971148297749567,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000413519496339793,
      "loss": 5.7494,
      "mean_token_accuracy": 0.18610137403011323,
      "num_tokens": 65250735.0,
      "step": 25745
    },
    {
      "entropy": 6.021968412399292,
      "epoch": 2.971725331794576,
      "grad_norm": 0.84375,
      "learning_rate": 0.00041348697663625144,
      "loss": 5.6415,
      "mean_token_accuracy": 0.20317477881908416,
      "num_tokens": 65263543.0,
      "step": 25750
    },
    {
      "entropy": 6.137899923324585,
      "epoch": 2.9723023658395844,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004134544522747692,
      "loss": 5.7988,
      "mean_token_accuracy": 0.18780627399682998,
      "num_tokens": 65275637.0,
      "step": 25755
    },
    {
      "entropy": 6.11231803894043,
      "epoch": 2.9728793998845933,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041342192325644024,
      "loss": 5.6483,
      "mean_token_accuracy": 0.202757366001606,
      "num_tokens": 65287564.0,
      "step": 25760
    },
    {
      "entropy": 6.134401655197143,
      "epoch": 2.9734564339296017,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004133893895823586,
      "loss": 5.8118,
      "mean_token_accuracy": 0.19222568422555925,
      "num_tokens": 65299255.0,
      "step": 25765
    },
    {
      "entropy": 6.225218820571899,
      "epoch": 2.9740334679746105,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004133568512536187,
      "loss": 5.7886,
      "mean_token_accuracy": 0.1872355595231056,
      "num_tokens": 65311749.0,
      "step": 25770
    },
    {
      "entropy": 6.10582127571106,
      "epoch": 2.974610502019619,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041332430827131474,
      "loss": 5.7158,
      "mean_token_accuracy": 0.19686691612005233,
      "num_tokens": 65325391.0,
      "step": 25775
    },
    {
      "entropy": 6.182678937911987,
      "epoch": 2.975187536064628,
      "grad_norm": 1.171875,
      "learning_rate": 0.00041329176063654137,
      "loss": 5.8569,
      "mean_token_accuracy": 0.17983190566301346,
      "num_tokens": 65338640.0,
      "step": 25780
    },
    {
      "entropy": 6.114098978042603,
      "epoch": 2.9757645701096367,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041325920835039344,
      "loss": 5.6331,
      "mean_token_accuracy": 0.19274860620498657,
      "num_tokens": 65351378.0,
      "step": 25785
    },
    {
      "entropy": 6.227236366271972,
      "epoch": 2.976341604154645,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004132266514139656,
      "loss": 5.8398,
      "mean_token_accuracy": 0.18156189024448394,
      "num_tokens": 65364072.0,
      "step": 25790
    },
    {
      "entropy": 6.145966720581055,
      "epoch": 2.9769186381996535,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041319408982835306,
      "loss": 5.7311,
      "mean_token_accuracy": 0.19451336562633514,
      "num_tokens": 65377019.0,
      "step": 25795
    },
    {
      "entropy": 6.126743173599243,
      "epoch": 2.9774956722446624,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041316152359465087,
      "loss": 5.7244,
      "mean_token_accuracy": 0.1898722916841507,
      "num_tokens": 65390465.0,
      "step": 25800
    },
    {
      "entropy": 6.194520139694214,
      "epoch": 2.9780727062896712,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041312895271395457,
      "loss": 5.7457,
      "mean_token_accuracy": 0.1883865177631378,
      "num_tokens": 65402947.0,
      "step": 25805
    },
    {
      "entropy": 6.161165428161621,
      "epoch": 2.9786497403346797,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004130963771873595,
      "loss": 5.8258,
      "mean_token_accuracy": 0.17923434376716613,
      "num_tokens": 65414725.0,
      "step": 25810
    },
    {
      "entropy": 6.213643026351929,
      "epoch": 2.9792267743796885,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041306379701596134,
      "loss": 5.7162,
      "mean_token_accuracy": 0.1917021244764328,
      "num_tokens": 65427417.0,
      "step": 25815
    },
    {
      "entropy": 6.0661111831665036,
      "epoch": 2.979803808424697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041303121220085607,
      "loss": 5.751,
      "mean_token_accuracy": 0.19381363987922667,
      "num_tokens": 65440151.0,
      "step": 25820
    },
    {
      "entropy": 6.128991174697876,
      "epoch": 2.980380842469706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004129986227431394,
      "loss": 5.7996,
      "mean_token_accuracy": 0.19028945565223693,
      "num_tokens": 65453774.0,
      "step": 25825
    },
    {
      "entropy": 6.251294040679932,
      "epoch": 2.9809578765147142,
      "grad_norm": 1.0,
      "learning_rate": 0.0004129660286439077,
      "loss": 5.8405,
      "mean_token_accuracy": 0.18535294830799104,
      "num_tokens": 65466523.0,
      "step": 25830
    },
    {
      "entropy": 6.242503595352173,
      "epoch": 2.981534910559723,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041293342990425714,
      "loss": 5.9239,
      "mean_token_accuracy": 0.1738576591014862,
      "num_tokens": 65480728.0,
      "step": 25835
    },
    {
      "entropy": 6.164182090759278,
      "epoch": 2.982111944604732,
      "grad_norm": 0.890625,
      "learning_rate": 0.00041290082652528415,
      "loss": 5.8636,
      "mean_token_accuracy": 0.18174926936626434,
      "num_tokens": 65494086.0,
      "step": 25840
    },
    {
      "entropy": 6.144267940521241,
      "epoch": 2.9826889786497404,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004128682185080854,
      "loss": 5.7004,
      "mean_token_accuracy": 0.19243649542331695,
      "num_tokens": 65508079.0,
      "step": 25845
    },
    {
      "entropy": 6.1460533618927,
      "epoch": 2.983266012694749,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004128356058537576,
      "loss": 5.7212,
      "mean_token_accuracy": 0.1948441118001938,
      "num_tokens": 65523511.0,
      "step": 25850
    },
    {
      "entropy": 6.07674012184143,
      "epoch": 2.9838430467397576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041280298856339763,
      "loss": 5.6909,
      "mean_token_accuracy": 0.19042058736085893,
      "num_tokens": 65535968.0,
      "step": 25855
    },
    {
      "entropy": 6.101561164855957,
      "epoch": 2.9844200807847665,
      "grad_norm": 1.0,
      "learning_rate": 0.00041277036663810266,
      "loss": 5.7901,
      "mean_token_accuracy": 0.18664904981851577,
      "num_tokens": 65548009.0,
      "step": 25860
    },
    {
      "entropy": 6.217222261428833,
      "epoch": 2.984997114829775,
      "grad_norm": 1.203125,
      "learning_rate": 0.00041273774007896983,
      "loss": 5.7239,
      "mean_token_accuracy": 0.19082981795072557,
      "num_tokens": 65561076.0,
      "step": 25865
    },
    {
      "entropy": 6.090844583511353,
      "epoch": 2.9855741488747833,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004127051088870964,
      "loss": 5.6381,
      "mean_token_accuracy": 0.20527324974536895,
      "num_tokens": 65573982.0,
      "step": 25870
    },
    {
      "entropy": 6.170323371887207,
      "epoch": 2.986151182919792,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004126724730635801,
      "loss": 5.8406,
      "mean_token_accuracy": 0.18809180110692977,
      "num_tokens": 65585923.0,
      "step": 25875
    },
    {
      "entropy": 6.141154384613037,
      "epoch": 2.986728216964801,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041263983260951857,
      "loss": 5.7218,
      "mean_token_accuracy": 0.18971893340349197,
      "num_tokens": 65597366.0,
      "step": 25880
    },
    {
      "entropy": 6.244146919250488,
      "epoch": 2.9873052510098095,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004126071875260097,
      "loss": 5.871,
      "mean_token_accuracy": 0.188421767950058,
      "num_tokens": 65610200.0,
      "step": 25885
    },
    {
      "entropy": 6.134675121307373,
      "epoch": 2.9878822850548183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004125745378141513,
      "loss": 5.7744,
      "mean_token_accuracy": 0.19529479146003723,
      "num_tokens": 65623270.0,
      "step": 25890
    },
    {
      "entropy": 6.048443794250488,
      "epoch": 2.9884593190998268,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004125418834750418,
      "loss": 5.6646,
      "mean_token_accuracy": 0.20220012366771697,
      "num_tokens": 65637443.0,
      "step": 25895
    },
    {
      "entropy": 6.239281082153321,
      "epoch": 2.9890363531448356,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004125092245097793,
      "loss": 5.8275,
      "mean_token_accuracy": 0.1816248133778572,
      "num_tokens": 65650004.0,
      "step": 25900
    },
    {
      "entropy": 6.171496438980102,
      "epoch": 2.989613387189844,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041247656091946235,
      "loss": 5.7348,
      "mean_token_accuracy": 0.18765859603881835,
      "num_tokens": 65662832.0,
      "step": 25905
    },
    {
      "entropy": 6.141176700592041,
      "epoch": 2.990190421234853,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004124438927051895,
      "loss": 5.7863,
      "mean_token_accuracy": 0.18915551006793976,
      "num_tokens": 65675545.0,
      "step": 25910
    },
    {
      "entropy": 6.2161157608032225,
      "epoch": 2.9907674552798618,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004124112198680597,
      "loss": 5.8068,
      "mean_token_accuracy": 0.18392547369003295,
      "num_tokens": 65688194.0,
      "step": 25915
    },
    {
      "entropy": 6.269847440719604,
      "epoch": 2.99134448932487,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004123785424091717,
      "loss": 5.8361,
      "mean_token_accuracy": 0.18446499556303025,
      "num_tokens": 65701410.0,
      "step": 25920
    },
    {
      "entropy": 6.207719373703003,
      "epoch": 2.9919215233698786,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004123458603296247,
      "loss": 5.8049,
      "mean_token_accuracy": 0.19142566323280336,
      "num_tokens": 65714540.0,
      "step": 25925
    },
    {
      "entropy": 6.205081939697266,
      "epoch": 2.9924985574148875,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004123131736305178,
      "loss": 5.8251,
      "mean_token_accuracy": 0.1890667736530304,
      "num_tokens": 65727774.0,
      "step": 25930
    },
    {
      "entropy": 6.196783399581909,
      "epoch": 2.9930755914598963,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004122804823129507,
      "loss": 5.851,
      "mean_token_accuracy": 0.18182846903800964,
      "num_tokens": 65740040.0,
      "step": 25935
    },
    {
      "entropy": 6.184090566635132,
      "epoch": 2.9936526255049047,
      "grad_norm": 1.0,
      "learning_rate": 0.0004122477863780227,
      "loss": 5.8251,
      "mean_token_accuracy": 0.18570015132427214,
      "num_tokens": 65751188.0,
      "step": 25940
    },
    {
      "entropy": 6.212691831588745,
      "epoch": 2.9942296595499136,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004122150858268335,
      "loss": 5.7821,
      "mean_token_accuracy": 0.18940315544605255,
      "num_tokens": 65762960.0,
      "step": 25945
    },
    {
      "entropy": 6.296775913238525,
      "epoch": 2.994806693594922,
      "grad_norm": 0.90625,
      "learning_rate": 0.00041218238066048315,
      "loss": 5.8341,
      "mean_token_accuracy": 0.18633692115545272,
      "num_tokens": 65775069.0,
      "step": 25950
    },
    {
      "entropy": 6.19817886352539,
      "epoch": 2.995383727639931,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004121496708800715,
      "loss": 5.8558,
      "mean_token_accuracy": 0.18426271080970763,
      "num_tokens": 65788181.0,
      "step": 25955
    },
    {
      "entropy": 6.121626520156861,
      "epoch": 2.9959607616849393,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00041211695648669884,
      "loss": 5.719,
      "mean_token_accuracy": 0.2072511687874794,
      "num_tokens": 65801038.0,
      "step": 25960
    },
    {
      "entropy": 6.108192586898804,
      "epoch": 2.996537795729948,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004120842374814654,
      "loss": 5.7022,
      "mean_token_accuracy": 0.19959067851305007,
      "num_tokens": 65813631.0,
      "step": 25965
    },
    {
      "entropy": 6.190539789199829,
      "epoch": 2.9971148297749566,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041205151386547174,
      "loss": 5.7796,
      "mean_token_accuracy": 0.18970050066709518,
      "num_tokens": 65825345.0,
      "step": 25970
    },
    {
      "entropy": 6.1861536502838135,
      "epoch": 2.9976918638199654,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041201878563981855,
      "loss": 5.7587,
      "mean_token_accuracy": 0.1962239846587181,
      "num_tokens": 65837144.0,
      "step": 25975
    },
    {
      "entropy": 6.0960300922393795,
      "epoch": 2.998268897864974,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041198605280560655,
      "loss": 5.7218,
      "mean_token_accuracy": 0.2002798691391945,
      "num_tokens": 65850458.0,
      "step": 25980
    },
    {
      "entropy": 6.268774461746216,
      "epoch": 2.9988459319099827,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004119533153639367,
      "loss": 5.8235,
      "mean_token_accuracy": 0.18734786212444304,
      "num_tokens": 65863659.0,
      "step": 25985
    },
    {
      "entropy": 6.124815797805786,
      "epoch": 2.9994229659549916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041192057331591,
      "loss": 5.7715,
      "mean_token_accuracy": 0.1886030152440071,
      "num_tokens": 65877129.0,
      "step": 25990
    },
    {
      "entropy": 6.2515641212463375,
      "epoch": 3.0,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000411887826662628,
      "loss": 5.8443,
      "mean_token_accuracy": 0.1817408338189125,
      "num_tokens": 65889351.0,
      "step": 25995
    },
    {
      "entropy": 6.103536176681518,
      "epoch": 3.000577034045009,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004118550754051919,
      "loss": 5.6964,
      "mean_token_accuracy": 0.19620571434497833,
      "num_tokens": 65901967.0,
      "step": 26000
    },
    {
      "entropy": 6.070988512039184,
      "epoch": 3.0011540680900173,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004118223195447033,
      "loss": 5.5774,
      "mean_token_accuracy": 0.20379541218280792,
      "num_tokens": 65915259.0,
      "step": 26005
    },
    {
      "entropy": 6.180714750289917,
      "epoch": 3.001731102135026,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041178955908226396,
      "loss": 5.7063,
      "mean_token_accuracy": 0.19393839538097382,
      "num_tokens": 65928236.0,
      "step": 26010
    },
    {
      "entropy": 6.130123281478882,
      "epoch": 3.0023081361800346,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004117567940189757,
      "loss": 5.7146,
      "mean_token_accuracy": 0.19025546908378602,
      "num_tokens": 65941290.0,
      "step": 26015
    },
    {
      "entropy": 6.146258878707886,
      "epoch": 3.0028851702250434,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004117240243559406,
      "loss": 5.7777,
      "mean_token_accuracy": 0.17930715382099152,
      "num_tokens": 65954301.0,
      "step": 26020
    },
    {
      "entropy": 6.244616603851318,
      "epoch": 3.003462204270052,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004116912500942609,
      "loss": 5.7581,
      "mean_token_accuracy": 0.19302790910005568,
      "num_tokens": 65966406.0,
      "step": 26025
    },
    {
      "entropy": 6.099221801757812,
      "epoch": 3.0040392383150607,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004116584712350389,
      "loss": 5.6754,
      "mean_token_accuracy": 0.19372874200344087,
      "num_tokens": 65979177.0,
      "step": 26030
    },
    {
      "entropy": 6.05877366065979,
      "epoch": 3.004616272360069,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004116256877793771,
      "loss": 5.6509,
      "mean_token_accuracy": 0.1997956305742264,
      "num_tokens": 65992326.0,
      "step": 26035
    },
    {
      "entropy": 6.231365919113159,
      "epoch": 3.005193306405078,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041159289972837813,
      "loss": 5.6941,
      "mean_token_accuracy": 0.19710197001695634,
      "num_tokens": 66003481.0,
      "step": 26040
    },
    {
      "entropy": 6.215103340148926,
      "epoch": 3.0057703404500864,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041156010708314487,
      "loss": 5.7768,
      "mean_token_accuracy": 0.18236880451440812,
      "num_tokens": 66015552.0,
      "step": 26045
    },
    {
      "entropy": 6.192623519897461,
      "epoch": 3.0063473744950953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004115273098447802,
      "loss": 5.7336,
      "mean_token_accuracy": 0.19326211959123613,
      "num_tokens": 66028506.0,
      "step": 26050
    },
    {
      "entropy": 6.187188911437988,
      "epoch": 3.0069244085401037,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041149450801438727,
      "loss": 5.6924,
      "mean_token_accuracy": 0.18974484354257584,
      "num_tokens": 66040938.0,
      "step": 26055
    },
    {
      "entropy": 6.104094076156616,
      "epoch": 3.0075014425851125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041146170159306946,
      "loss": 5.6314,
      "mean_token_accuracy": 0.18857638090848922,
      "num_tokens": 66052838.0,
      "step": 26060
    },
    {
      "entropy": 6.176004314422608,
      "epoch": 3.008078476630121,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041142889058193,
      "loss": 5.6849,
      "mean_token_accuracy": 0.19574964046478271,
      "num_tokens": 66065286.0,
      "step": 26065
    },
    {
      "entropy": 6.1532289505004885,
      "epoch": 3.00865551067513,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041139607498207265,
      "loss": 5.7334,
      "mean_token_accuracy": 0.19025952816009523,
      "num_tokens": 66077174.0,
      "step": 26070
    },
    {
      "entropy": 6.170625448226929,
      "epoch": 3.0092325447201387,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000411363254794601,
      "loss": 5.7704,
      "mean_token_accuracy": 0.19564634412527085,
      "num_tokens": 66092510.0,
      "step": 26075
    },
    {
      "entropy": 6.188890075683593,
      "epoch": 3.009809578765147,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000411330430020619,
      "loss": 5.6948,
      "mean_token_accuracy": 0.2006187155842781,
      "num_tokens": 66105001.0,
      "step": 26080
    },
    {
      "entropy": 6.146265125274658,
      "epoch": 3.010386612810156,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004112976006612308,
      "loss": 5.6913,
      "mean_token_accuracy": 0.1987504929304123,
      "num_tokens": 66117264.0,
      "step": 26085
    },
    {
      "entropy": 6.111363410949707,
      "epoch": 3.0109636468551644,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041126476671754037,
      "loss": 5.7288,
      "mean_token_accuracy": 0.19112110882997513,
      "num_tokens": 66130954.0,
      "step": 26090
    },
    {
      "entropy": 6.189563894271851,
      "epoch": 3.0115406809001732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041123192819065234,
      "loss": 5.7194,
      "mean_token_accuracy": 0.1923195481300354,
      "num_tokens": 66143560.0,
      "step": 26095
    },
    {
      "entropy": 6.256980752944946,
      "epoch": 3.0121177149451817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000411199085081671,
      "loss": 5.7744,
      "mean_token_accuracy": 0.19028388261795043,
      "num_tokens": 66157169.0,
      "step": 26100
    },
    {
      "entropy": 6.167681169509888,
      "epoch": 3.0126947489901905,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041116623739170105,
      "loss": 5.7122,
      "mean_token_accuracy": 0.18988695442676545,
      "num_tokens": 66171331.0,
      "step": 26105
    },
    {
      "entropy": 6.182020854949951,
      "epoch": 3.013271783035199,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004111333851218474,
      "loss": 5.7432,
      "mean_token_accuracy": 0.193082532286644,
      "num_tokens": 66184428.0,
      "step": 26110
    },
    {
      "entropy": 6.191115522384644,
      "epoch": 3.013848817080208,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041110052827321496,
      "loss": 5.7007,
      "mean_token_accuracy": 0.19426756352186203,
      "num_tokens": 66196602.0,
      "step": 26115
    },
    {
      "entropy": 6.069936084747314,
      "epoch": 3.014425851125216,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041106766684690884,
      "loss": 5.6559,
      "mean_token_accuracy": 0.19884906560182572,
      "num_tokens": 66209603.0,
      "step": 26120
    },
    {
      "entropy": 6.13335018157959,
      "epoch": 3.015002885170225,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004110348008440344,
      "loss": 5.7338,
      "mean_token_accuracy": 0.1877104803919792,
      "num_tokens": 66221587.0,
      "step": 26125
    },
    {
      "entropy": 6.274585247039795,
      "epoch": 3.0155799192152335,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041100193026569686,
      "loss": 5.7735,
      "mean_token_accuracy": 0.18737897872924805,
      "num_tokens": 66233243.0,
      "step": 26130
    },
    {
      "entropy": 6.096971464157105,
      "epoch": 3.0161569532602424,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041096905511300207,
      "loss": 5.6865,
      "mean_token_accuracy": 0.19875524789094925,
      "num_tokens": 66245831.0,
      "step": 26135
    },
    {
      "entropy": 6.241936302185058,
      "epoch": 3.016733987305251,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004109361753870556,
      "loss": 5.7666,
      "mean_token_accuracy": 0.18903570473194123,
      "num_tokens": 66259113.0,
      "step": 26140
    },
    {
      "entropy": 6.1352049827575685,
      "epoch": 3.0173110213502596,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004109032910889635,
      "loss": 5.6728,
      "mean_token_accuracy": 0.19991651475429534,
      "num_tokens": 66272633.0,
      "step": 26145
    },
    {
      "entropy": 6.2056357860565186,
      "epoch": 3.0178880553952685,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004108704022198316,
      "loss": 5.6933,
      "mean_token_accuracy": 0.1884486123919487,
      "num_tokens": 66286100.0,
      "step": 26150
    },
    {
      "entropy": 6.129799842834473,
      "epoch": 3.018465089440277,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041083750878076627,
      "loss": 5.6322,
      "mean_token_accuracy": 0.1919589802622795,
      "num_tokens": 66297623.0,
      "step": 26155
    },
    {
      "entropy": 6.100574254989624,
      "epoch": 3.0190421234852858,
      "grad_norm": 1.125,
      "learning_rate": 0.0004108046107728737,
      "loss": 5.7118,
      "mean_token_accuracy": 0.1900741547346115,
      "num_tokens": 66311688.0,
      "step": 26160
    },
    {
      "entropy": 6.1915263652801515,
      "epoch": 3.019619157530294,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041077170819726064,
      "loss": 5.7294,
      "mean_token_accuracy": 0.18557717502117158,
      "num_tokens": 66324317.0,
      "step": 26165
    },
    {
      "entropy": 6.144586896896362,
      "epoch": 3.020196191575303,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041073880105503346,
      "loss": 5.6758,
      "mean_token_accuracy": 0.1939310312271118,
      "num_tokens": 66337152.0,
      "step": 26170
    },
    {
      "entropy": 6.261649751663208,
      "epoch": 3.0207732256203115,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041070588934729923,
      "loss": 5.7897,
      "mean_token_accuracy": 0.186767840385437,
      "num_tokens": 66349043.0,
      "step": 26175
    },
    {
      "entropy": 6.060703134536743,
      "epoch": 3.0213502596653203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004106729730751649,
      "loss": 5.5948,
      "mean_token_accuracy": 0.20910189747810365,
      "num_tokens": 66362283.0,
      "step": 26180
    },
    {
      "entropy": 6.065590906143188,
      "epoch": 3.0219272937103288,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00041064005223973737,
      "loss": 5.5758,
      "mean_token_accuracy": 0.20630859583616257,
      "num_tokens": 66376919.0,
      "step": 26185
    },
    {
      "entropy": 6.063845252990722,
      "epoch": 3.0225043277553376,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004106071268421241,
      "loss": 5.5982,
      "mean_token_accuracy": 0.2077121764421463,
      "num_tokens": 66389285.0,
      "step": 26190
    },
    {
      "entropy": 6.156057548522949,
      "epoch": 3.023081361800346,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041057419688343255,
      "loss": 5.7779,
      "mean_token_accuracy": 0.19199763089418412,
      "num_tokens": 66402196.0,
      "step": 26195
    },
    {
      "entropy": 6.036690139770508,
      "epoch": 3.023658395845355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004105412623647702,
      "loss": 5.6393,
      "mean_token_accuracy": 0.1942216619849205,
      "num_tokens": 66414334.0,
      "step": 26200
    },
    {
      "entropy": 6.16955156326294,
      "epoch": 3.0242354298903633,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041050832328724476,
      "loss": 5.7505,
      "mean_token_accuracy": 0.1834244728088379,
      "num_tokens": 66425975.0,
      "step": 26205
    },
    {
      "entropy": 6.120315504074097,
      "epoch": 3.024812463935372,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004104753796519642,
      "loss": 5.6315,
      "mean_token_accuracy": 0.19852339923381807,
      "num_tokens": 66438841.0,
      "step": 26210
    },
    {
      "entropy": 6.251455450057984,
      "epoch": 3.025389497980381,
      "grad_norm": 0.875,
      "learning_rate": 0.0004104424314600365,
      "loss": 5.7502,
      "mean_token_accuracy": 0.18712622076272964,
      "num_tokens": 66451100.0,
      "step": 26215
    },
    {
      "entropy": 6.155909872055053,
      "epoch": 3.0259665320253895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041040947871257,
      "loss": 5.7105,
      "mean_token_accuracy": 0.1956273227930069,
      "num_tokens": 66463613.0,
      "step": 26220
    },
    {
      "entropy": 6.141160011291504,
      "epoch": 3.0265435660703983,
      "grad_norm": 0.921875,
      "learning_rate": 0.00041037652141067294,
      "loss": 5.6759,
      "mean_token_accuracy": 0.19774365723133086,
      "num_tokens": 66476676.0,
      "step": 26225
    },
    {
      "entropy": 6.224869012832642,
      "epoch": 3.0271206001154067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041034355955545375,
      "loss": 5.7508,
      "mean_token_accuracy": 0.1915663704276085,
      "num_tokens": 66489577.0,
      "step": 26230
    },
    {
      "entropy": 6.147612047195435,
      "epoch": 3.0276976341604156,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041031059314802114,
      "loss": 5.7023,
      "mean_token_accuracy": 0.18792953342199326,
      "num_tokens": 66500732.0,
      "step": 26235
    },
    {
      "entropy": 6.105926656723023,
      "epoch": 3.028274668205424,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041027762218948403,
      "loss": 5.6576,
      "mean_token_accuracy": 0.20472156256437302,
      "num_tokens": 66512342.0,
      "step": 26240
    },
    {
      "entropy": 6.1010932445526125,
      "epoch": 3.028851702250433,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041024464668095123,
      "loss": 5.7144,
      "mean_token_accuracy": 0.19461794048547745,
      "num_tokens": 66526042.0,
      "step": 26245
    },
    {
      "entropy": 6.151596403121948,
      "epoch": 3.0294287362954413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000410211666623532,
      "loss": 5.6801,
      "mean_token_accuracy": 0.19064646810293198,
      "num_tokens": 66537995.0,
      "step": 26250
    },
    {
      "entropy": 6.108174753189087,
      "epoch": 3.03000577034045,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00041017868201833545,
      "loss": 5.7791,
      "mean_token_accuracy": 0.18986818939447403,
      "num_tokens": 66550684.0,
      "step": 26255
    },
    {
      "entropy": 6.143332004547119,
      "epoch": 3.0305828043854586,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004101456928664711,
      "loss": 5.6746,
      "mean_token_accuracy": 0.19494170397520066,
      "num_tokens": 66563940.0,
      "step": 26260
    },
    {
      "entropy": 6.075266313552857,
      "epoch": 3.0311598384304674,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004101126991690485,
      "loss": 5.6444,
      "mean_token_accuracy": 0.1995875656604767,
      "num_tokens": 66577411.0,
      "step": 26265
    },
    {
      "entropy": 6.175258445739746,
      "epoch": 3.031736872475476,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004100797009271774,
      "loss": 5.757,
      "mean_token_accuracy": 0.19028777331113816,
      "num_tokens": 66589520.0,
      "step": 26270
    },
    {
      "entropy": 6.179210090637207,
      "epoch": 3.0323139065204847,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004100466981419676,
      "loss": 5.716,
      "mean_token_accuracy": 0.19614706933498383,
      "num_tokens": 66602573.0,
      "step": 26275
    },
    {
      "entropy": 6.035876131057739,
      "epoch": 3.0328909405654936,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041001369081452926,
      "loss": 5.5739,
      "mean_token_accuracy": 0.20045224130153655,
      "num_tokens": 66614739.0,
      "step": 26280
    },
    {
      "entropy": 6.1427703380584715,
      "epoch": 3.033467974610502,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040998067894597247,
      "loss": 5.7963,
      "mean_token_accuracy": 0.1862247109413147,
      "num_tokens": 66628635.0,
      "step": 26285
    },
    {
      "entropy": 6.247771215438843,
      "epoch": 3.034045008655511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040994766253740756,
      "loss": 5.8089,
      "mean_token_accuracy": 0.18791309297084807,
      "num_tokens": 66640911.0,
      "step": 26290
    },
    {
      "entropy": 6.048626375198364,
      "epoch": 3.0346220427005193,
      "grad_norm": 0.80859375,
      "learning_rate": 0.00040991464158994506,
      "loss": 5.5785,
      "mean_token_accuracy": 0.2054696187376976,
      "num_tokens": 66654150.0,
      "step": 26295
    },
    {
      "entropy": 6.089620685577392,
      "epoch": 3.035199076745528,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004098816161046956,
      "loss": 5.7522,
      "mean_token_accuracy": 0.18896272033452988,
      "num_tokens": 66666026.0,
      "step": 26300
    },
    {
      "entropy": 6.139586591720581,
      "epoch": 3.0357761107905366,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040984858608277,
      "loss": 5.6296,
      "mean_token_accuracy": 0.19342040419578552,
      "num_tokens": 66679446.0,
      "step": 26305
    },
    {
      "entropy": 6.2065582275390625,
      "epoch": 3.0363531448355454,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004098155515252793,
      "loss": 5.7985,
      "mean_token_accuracy": 0.19067757874727248,
      "num_tokens": 66692689.0,
      "step": 26310
    },
    {
      "entropy": 6.176087236404419,
      "epoch": 3.036930178880554,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004097825124333343,
      "loss": 5.6241,
      "mean_token_accuracy": 0.20387111455202103,
      "num_tokens": 66705765.0,
      "step": 26315
    },
    {
      "entropy": 6.092849206924439,
      "epoch": 3.0375072129255627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004097494688080466,
      "loss": 5.6624,
      "mean_token_accuracy": 0.1970837026834488,
      "num_tokens": 66718856.0,
      "step": 26320
    },
    {
      "entropy": 6.196405363082886,
      "epoch": 3.038084246970571,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040971642065052734,
      "loss": 5.7207,
      "mean_token_accuracy": 0.19337523132562637,
      "num_tokens": 66731330.0,
      "step": 26325
    },
    {
      "entropy": 6.159465599060058,
      "epoch": 3.03866128101558,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004096833679618882,
      "loss": 5.7269,
      "mean_token_accuracy": 0.19103855788707733,
      "num_tokens": 66743676.0,
      "step": 26330
    },
    {
      "entropy": 6.112085294723511,
      "epoch": 3.0392383150605884,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040965031074324094,
      "loss": 5.6569,
      "mean_token_accuracy": 0.195459346473217,
      "num_tokens": 66757663.0,
      "step": 26335
    },
    {
      "entropy": 6.12399468421936,
      "epoch": 3.0398153491055973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004096172489956973,
      "loss": 5.6958,
      "mean_token_accuracy": 0.1900702640414238,
      "num_tokens": 66770149.0,
      "step": 26340
    },
    {
      "entropy": 6.143730020523071,
      "epoch": 3.0403923831506057,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004095841827203694,
      "loss": 5.6652,
      "mean_token_accuracy": 0.19444374591112137,
      "num_tokens": 66781754.0,
      "step": 26345
    },
    {
      "entropy": 6.040595626831054,
      "epoch": 3.0409694171956145,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004095511119183693,
      "loss": 5.5352,
      "mean_token_accuracy": 0.20391818284988403,
      "num_tokens": 66795213.0,
      "step": 26350
    },
    {
      "entropy": 6.186661577224731,
      "epoch": 3.0415464512406234,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004095180365908094,
      "loss": 5.6991,
      "mean_token_accuracy": 0.19509964436292648,
      "num_tokens": 66808514.0,
      "step": 26355
    },
    {
      "entropy": 5.970676851272583,
      "epoch": 3.042123485285632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004094849567388022,
      "loss": 5.5257,
      "mean_token_accuracy": 0.2070065051317215,
      "num_tokens": 66821665.0,
      "step": 26360
    },
    {
      "entropy": 6.129239511489868,
      "epoch": 3.0427005193306407,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004094518723634603,
      "loss": 5.7015,
      "mean_token_accuracy": 0.1878933623433113,
      "num_tokens": 66834447.0,
      "step": 26365
    },
    {
      "entropy": 6.204967737197876,
      "epoch": 3.043277553375649,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004094187834658964,
      "loss": 5.7395,
      "mean_token_accuracy": 0.18758202642202376,
      "num_tokens": 66846199.0,
      "step": 26370
    },
    {
      "entropy": 6.151585578918457,
      "epoch": 3.043854587420658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004093856900472236,
      "loss": 5.7064,
      "mean_token_accuracy": 0.19432766139507293,
      "num_tokens": 66858055.0,
      "step": 26375
    },
    {
      "entropy": 6.178366994857788,
      "epoch": 3.0444316214656664,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040935259210855477,
      "loss": 5.6956,
      "mean_token_accuracy": 0.1904509335756302,
      "num_tokens": 66870779.0,
      "step": 26380
    },
    {
      "entropy": 6.074148511886596,
      "epoch": 3.0450086555106752,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004093194896510033,
      "loss": 5.6791,
      "mean_token_accuracy": 0.20142536014318466,
      "num_tokens": 66883212.0,
      "step": 26385
    },
    {
      "entropy": 6.107582664489746,
      "epoch": 3.0455856895556837,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040928638267568245,
      "loss": 5.631,
      "mean_token_accuracy": 0.20406220853328705,
      "num_tokens": 66897019.0,
      "step": 26390
    },
    {
      "entropy": 6.180645990371704,
      "epoch": 3.0461627236006925,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040925327118370586,
      "loss": 5.7394,
      "mean_token_accuracy": 0.19447785317897798,
      "num_tokens": 66909415.0,
      "step": 26395
    },
    {
      "entropy": 6.11536922454834,
      "epoch": 3.046739757645701,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040922015517618724,
      "loss": 5.726,
      "mean_token_accuracy": 0.19340649098157883,
      "num_tokens": 66921459.0,
      "step": 26400
    },
    {
      "entropy": 6.12619161605835,
      "epoch": 3.04731679169071,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004091870346542403,
      "loss": 5.6164,
      "mean_token_accuracy": 0.20126460492610931,
      "num_tokens": 66933558.0,
      "step": 26405
    },
    {
      "entropy": 6.111900901794433,
      "epoch": 3.047893825735718,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004091539096189791,
      "loss": 5.7159,
      "mean_token_accuracy": 0.19719034135341645,
      "num_tokens": 66945950.0,
      "step": 26410
    },
    {
      "entropy": 6.10090069770813,
      "epoch": 3.048470859780727,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040912078007151776,
      "loss": 5.6672,
      "mean_token_accuracy": 0.19310089200735092,
      "num_tokens": 66958848.0,
      "step": 26415
    },
    {
      "entropy": 6.097727203369141,
      "epoch": 3.049047893825736,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040908764601297064,
      "loss": 5.5949,
      "mean_token_accuracy": 0.20191023051738738,
      "num_tokens": 66970302.0,
      "step": 26420
    },
    {
      "entropy": 6.144336843490601,
      "epoch": 3.0496249278707444,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004090545074444522,
      "loss": 5.7238,
      "mean_token_accuracy": 0.19061239957809448,
      "num_tokens": 66984465.0,
      "step": 26425
    },
    {
      "entropy": 6.136382913589477,
      "epoch": 3.050201961915753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040902136436707696,
      "loss": 5.6472,
      "mean_token_accuracy": 0.193617807328701,
      "num_tokens": 66995555.0,
      "step": 26430
    },
    {
      "entropy": 6.152734041213989,
      "epoch": 3.0507789959607616,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004089882167819596,
      "loss": 5.7297,
      "mean_token_accuracy": 0.19138612002134323,
      "num_tokens": 67007823.0,
      "step": 26435
    },
    {
      "entropy": 6.1525016784667965,
      "epoch": 3.0513560300057705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004089550646902153,
      "loss": 5.6498,
      "mean_token_accuracy": 0.19544131308794022,
      "num_tokens": 67019963.0,
      "step": 26440
    },
    {
      "entropy": 5.967006015777588,
      "epoch": 3.051933064050779,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040892190809295876,
      "loss": 5.5305,
      "mean_token_accuracy": 0.21141353398561477,
      "num_tokens": 67032721.0,
      "step": 26445
    },
    {
      "entropy": 6.125797080993652,
      "epoch": 3.0525100980957878,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004088887469913055,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1992385521531105,
      "num_tokens": 67046713.0,
      "step": 26450
    },
    {
      "entropy": 6.186389064788818,
      "epoch": 3.053087132140796,
      "grad_norm": 1.0,
      "learning_rate": 0.0004088555813863707,
      "loss": 5.749,
      "mean_token_accuracy": 0.18884366899728774,
      "num_tokens": 67059874.0,
      "step": 26455
    },
    {
      "entropy": 6.20120530128479,
      "epoch": 3.053664166185805,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004088224112792699,
      "loss": 5.7115,
      "mean_token_accuracy": 0.19516452252864838,
      "num_tokens": 67071897.0,
      "step": 26460
    },
    {
      "entropy": 6.025647926330566,
      "epoch": 3.0542412002308135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040878923667111874,
      "loss": 5.6048,
      "mean_token_accuracy": 0.19705188274383545,
      "num_tokens": 67084489.0,
      "step": 26465
    },
    {
      "entropy": 6.131008768081665,
      "epoch": 3.0548182342758223,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040875605756303314,
      "loss": 5.7014,
      "mean_token_accuracy": 0.1943628817796707,
      "num_tokens": 67095812.0,
      "step": 26470
    },
    {
      "entropy": 6.1213202476501465,
      "epoch": 3.0553952683208307,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004087228739561289,
      "loss": 5.6182,
      "mean_token_accuracy": 0.20195652395486832,
      "num_tokens": 67108551.0,
      "step": 26475
    },
    {
      "entropy": 6.104694986343384,
      "epoch": 3.0559723023658396,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004086896858515223,
      "loss": 5.6982,
      "mean_token_accuracy": 0.19177381247282027,
      "num_tokens": 67120887.0,
      "step": 26480
    },
    {
      "entropy": 6.12642388343811,
      "epoch": 3.056549336410848,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004086564932503294,
      "loss": 5.6792,
      "mean_token_accuracy": 0.19861301630735398,
      "num_tokens": 67133758.0,
      "step": 26485
    },
    {
      "entropy": 6.126889753341675,
      "epoch": 3.057126370455857,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004086232961536669,
      "loss": 5.6201,
      "mean_token_accuracy": 0.19907300621271135,
      "num_tokens": 67145880.0,
      "step": 26490
    },
    {
      "entropy": 6.125477409362793,
      "epoch": 3.0577034045008658,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040859009456265113,
      "loss": 5.6763,
      "mean_token_accuracy": 0.19218284785747528,
      "num_tokens": 67158104.0,
      "step": 26495
    },
    {
      "entropy": 6.170765066146851,
      "epoch": 3.058280438545874,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040855688847839884,
      "loss": 5.7127,
      "mean_token_accuracy": 0.18946211636066437,
      "num_tokens": 67170179.0,
      "step": 26500
    },
    {
      "entropy": 6.198104286193848,
      "epoch": 3.058857472590883,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00040852367790202693,
      "loss": 5.783,
      "mean_token_accuracy": 0.1860157936811447,
      "num_tokens": 67182318.0,
      "step": 26505
    },
    {
      "entropy": 6.208402824401856,
      "epoch": 3.0594345066358914,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004084904628346525,
      "loss": 5.7536,
      "mean_token_accuracy": 0.1904895156621933,
      "num_tokens": 67195302.0,
      "step": 26510
    },
    {
      "entropy": 6.2347949028015135,
      "epoch": 3.0600115406809003,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040845724327739273,
      "loss": 5.7325,
      "mean_token_accuracy": 0.19044644236564637,
      "num_tokens": 67208191.0,
      "step": 26515
    },
    {
      "entropy": 6.079102611541748,
      "epoch": 3.0605885747259087,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040842401923136477,
      "loss": 5.6731,
      "mean_token_accuracy": 0.19952663034200668,
      "num_tokens": 67221410.0,
      "step": 26520
    },
    {
      "entropy": 6.111173009872436,
      "epoch": 3.0611656087709176,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040839079069768627,
      "loss": 5.6919,
      "mean_token_accuracy": 0.19796486794948578,
      "num_tokens": 67233958.0,
      "step": 26525
    },
    {
      "entropy": 6.090298223495483,
      "epoch": 3.061742642815926,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004083575576774747,
      "loss": 5.634,
      "mean_token_accuracy": 0.20558057129383087,
      "num_tokens": 67246622.0,
      "step": 26530
    },
    {
      "entropy": 6.1435261249542235,
      "epoch": 3.062319676860935,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040832432017184797,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19658363461494446,
      "num_tokens": 67258942.0,
      "step": 26535
    },
    {
      "entropy": 6.201191711425781,
      "epoch": 3.0628967109059433,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000408291078181924,
      "loss": 5.7049,
      "mean_token_accuracy": 0.18951683342456818,
      "num_tokens": 67271094.0,
      "step": 26540
    },
    {
      "entropy": 6.137856531143188,
      "epoch": 3.063473744950952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040825783170882077,
      "loss": 5.7076,
      "mean_token_accuracy": 0.1904965028166771,
      "num_tokens": 67284715.0,
      "step": 26545
    },
    {
      "entropy": 6.185946798324585,
      "epoch": 3.0640507789959606,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040822458075365654,
      "loss": 5.6824,
      "mean_token_accuracy": 0.19310538619756698,
      "num_tokens": 67298096.0,
      "step": 26550
    },
    {
      "entropy": 6.084507417678833,
      "epoch": 3.0646278130409694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004081913253175498,
      "loss": 5.6769,
      "mean_token_accuracy": 0.1873644009232521,
      "num_tokens": 67310548.0,
      "step": 26555
    },
    {
      "entropy": 6.147056770324707,
      "epoch": 3.065204847085978,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040815806540161895,
      "loss": 5.687,
      "mean_token_accuracy": 0.1979440465569496,
      "num_tokens": 67322771.0,
      "step": 26560
    },
    {
      "entropy": 6.093401670455933,
      "epoch": 3.0657818811309867,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004081248010069827,
      "loss": 5.6095,
      "mean_token_accuracy": 0.20410916060209275,
      "num_tokens": 67336565.0,
      "step": 26565
    },
    {
      "entropy": 6.051231718063354,
      "epoch": 3.0663589151759956,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040809153213475983,
      "loss": 5.679,
      "mean_token_accuracy": 0.19722304940223695,
      "num_tokens": 67348390.0,
      "step": 26570
    },
    {
      "entropy": 6.108557510375976,
      "epoch": 3.066935949221004,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040805825878606944,
      "loss": 5.6483,
      "mean_token_accuracy": 0.19871712177991868,
      "num_tokens": 67360686.0,
      "step": 26575
    },
    {
      "entropy": 6.1745765686035154,
      "epoch": 3.067512983266013,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004080249809620307,
      "loss": 5.7234,
      "mean_token_accuracy": 0.19749303311109542,
      "num_tokens": 67372092.0,
      "step": 26580
    },
    {
      "entropy": 6.113530492782592,
      "epoch": 3.0680900173110213,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040799169866376266,
      "loss": 5.6383,
      "mean_token_accuracy": 0.1920008584856987,
      "num_tokens": 67386017.0,
      "step": 26585
    },
    {
      "entropy": 6.130209827423096,
      "epoch": 3.06866705135603,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040795841189238504,
      "loss": 5.7445,
      "mean_token_accuracy": 0.19548604041337966,
      "num_tokens": 67397987.0,
      "step": 26590
    },
    {
      "entropy": 6.232554578781128,
      "epoch": 3.0692440854010385,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040792512064901715,
      "loss": 5.8177,
      "mean_token_accuracy": 0.18661231100559234,
      "num_tokens": 67410756.0,
      "step": 26595
    },
    {
      "entropy": 6.268054389953614,
      "epoch": 3.0698211194460474,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040789182493477887,
      "loss": 5.7909,
      "mean_token_accuracy": 0.19176454693078995,
      "num_tokens": 67423528.0,
      "step": 26600
    },
    {
      "entropy": 6.112860822677613,
      "epoch": 3.070398153491056,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040785852475079017,
      "loss": 5.7464,
      "mean_token_accuracy": 0.19084911197423934,
      "num_tokens": 67436131.0,
      "step": 26605
    },
    {
      "entropy": 6.06038646697998,
      "epoch": 3.0709751875360647,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004078252200981709,
      "loss": 5.7182,
      "mean_token_accuracy": 0.19264955818653107,
      "num_tokens": 67449422.0,
      "step": 26610
    },
    {
      "entropy": 6.239610767364502,
      "epoch": 3.071552221581073,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040779191097804133,
      "loss": 5.725,
      "mean_token_accuracy": 0.1877145528793335,
      "num_tokens": 67461997.0,
      "step": 26615
    },
    {
      "entropy": 6.255790758132934,
      "epoch": 3.072129255626082,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040775859739152187,
      "loss": 5.8136,
      "mean_token_accuracy": 0.18494182229042053,
      "num_tokens": 67474758.0,
      "step": 26620
    },
    {
      "entropy": 6.040949821472168,
      "epoch": 3.0727062896710904,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004077252793397329,
      "loss": 5.6302,
      "mean_token_accuracy": 0.19428464025259018,
      "num_tokens": 67487437.0,
      "step": 26625
    },
    {
      "entropy": 6.114373779296875,
      "epoch": 3.0732833237160992,
      "grad_norm": 1.015625,
      "learning_rate": 0.000407691956823795,
      "loss": 5.6769,
      "mean_token_accuracy": 0.19175334572792052,
      "num_tokens": 67500850.0,
      "step": 26630
    },
    {
      "entropy": 6.155919218063355,
      "epoch": 3.073860357761108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040765862984482916,
      "loss": 5.6348,
      "mean_token_accuracy": 0.1979771926999092,
      "num_tokens": 67512303.0,
      "step": 26635
    },
    {
      "entropy": 6.176354455947876,
      "epoch": 3.0744373918061165,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004076252984039561,
      "loss": 5.7717,
      "mean_token_accuracy": 0.1902720347046852,
      "num_tokens": 67524561.0,
      "step": 26640
    },
    {
      "entropy": 6.177938413619995,
      "epoch": 3.0750144258511254,
      "grad_norm": 0.90625,
      "learning_rate": 0.00040759196250229706,
      "loss": 5.745,
      "mean_token_accuracy": 0.184525465965271,
      "num_tokens": 67536859.0,
      "step": 26645
    },
    {
      "entropy": 6.1658063411712645,
      "epoch": 3.075591459896134,
      "grad_norm": 1.265625,
      "learning_rate": 0.00040755862214097326,
      "loss": 5.7477,
      "mean_token_accuracy": 0.18829769790172576,
      "num_tokens": 67549265.0,
      "step": 26650
    },
    {
      "entropy": 6.098076868057251,
      "epoch": 3.0761684939411427,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000407525277321106,
      "loss": 5.6981,
      "mean_token_accuracy": 0.1875301256775856,
      "num_tokens": 67562630.0,
      "step": 26655
    },
    {
      "entropy": 6.229063844680786,
      "epoch": 3.076745527986151,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004074919280438169,
      "loss": 5.7247,
      "mean_token_accuracy": 0.19259355068206788,
      "num_tokens": 67574358.0,
      "step": 26660
    },
    {
      "entropy": 6.207111883163452,
      "epoch": 3.07732256203116,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040745857431022755,
      "loss": 5.8207,
      "mean_token_accuracy": 0.1826626479625702,
      "num_tokens": 67587514.0,
      "step": 26665
    },
    {
      "entropy": 6.229054260253906,
      "epoch": 3.0778995960761684,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040742521612145986,
      "loss": 5.8285,
      "mean_token_accuracy": 0.1867430552840233,
      "num_tokens": 67600406.0,
      "step": 26670
    },
    {
      "entropy": 6.113597679138183,
      "epoch": 3.0784766301211772,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040739185347863586,
      "loss": 5.6345,
      "mean_token_accuracy": 0.20075120478868486,
      "num_tokens": 67612988.0,
      "step": 26675
    },
    {
      "entropy": 6.109706878662109,
      "epoch": 3.0790536641661856,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004073584863828776,
      "loss": 5.6924,
      "mean_token_accuracy": 0.19725033342838288,
      "num_tokens": 67625949.0,
      "step": 26680
    },
    {
      "entropy": 6.148903846740723,
      "epoch": 3.0796306982111945,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004073251148353074,
      "loss": 5.6832,
      "mean_token_accuracy": 0.19956444352865219,
      "num_tokens": 67637971.0,
      "step": 26685
    },
    {
      "entropy": 6.128827619552612,
      "epoch": 3.080207732256203,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004072917388370477,
      "loss": 5.6621,
      "mean_token_accuracy": 0.19446104764938354,
      "num_tokens": 67650311.0,
      "step": 26690
    },
    {
      "entropy": 6.115025806427002,
      "epoch": 3.080784766301212,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040725835838922105,
      "loss": 5.7023,
      "mean_token_accuracy": 0.1935811772942543,
      "num_tokens": 67662812.0,
      "step": 26695
    },
    {
      "entropy": 6.1605547904968265,
      "epoch": 3.0813618003462206,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004072249734929502,
      "loss": 5.7693,
      "mean_token_accuracy": 0.18660773932933808,
      "num_tokens": 67674878.0,
      "step": 26700
    },
    {
      "entropy": 6.180890083312988,
      "epoch": 3.081938834391229,
      "grad_norm": 0.90625,
      "learning_rate": 0.00040719158414935806,
      "loss": 5.7122,
      "mean_token_accuracy": 0.19786154627799987,
      "num_tokens": 67687562.0,
      "step": 26705
    },
    {
      "entropy": 6.266283941268921,
      "epoch": 3.082515868436238,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004071581903595677,
      "loss": 5.8158,
      "mean_token_accuracy": 0.18455517441034316,
      "num_tokens": 67699546.0,
      "step": 26710
    },
    {
      "entropy": 6.114570188522339,
      "epoch": 3.0830929024812463,
      "grad_norm": 1.171875,
      "learning_rate": 0.00040712479212470223,
      "loss": 5.5815,
      "mean_token_accuracy": 0.20124581307172776,
      "num_tokens": 67711247.0,
      "step": 26715
    },
    {
      "entropy": 6.066486644744873,
      "epoch": 3.083669936526255,
      "grad_norm": 0.9375,
      "learning_rate": 0.000407091389445885,
      "loss": 5.7208,
      "mean_token_accuracy": 0.18863172829151154,
      "num_tokens": 67724305.0,
      "step": 26720
    },
    {
      "entropy": 6.120010900497436,
      "epoch": 3.0842469705712636,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004070579823242395,
      "loss": 5.7166,
      "mean_token_accuracy": 0.19494422227144242,
      "num_tokens": 67737015.0,
      "step": 26725
    },
    {
      "entropy": 6.209732532501221,
      "epoch": 3.0848240046162725,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040702457076088937,
      "loss": 5.757,
      "mean_token_accuracy": 0.18922561258077622,
      "num_tokens": 67748883.0,
      "step": 26730
    },
    {
      "entropy": 6.1650079727172855,
      "epoch": 3.085401038661281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004069911547569584,
      "loss": 5.6933,
      "mean_token_accuracy": 0.19975563883781433,
      "num_tokens": 67761278.0,
      "step": 26735
    },
    {
      "entropy": 6.141102981567383,
      "epoch": 3.0859780727062898,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040695773431357034,
      "loss": 5.7219,
      "mean_token_accuracy": 0.18901601433753967,
      "num_tokens": 67772823.0,
      "step": 26740
    },
    {
      "entropy": 6.14684796333313,
      "epoch": 3.086555106751298,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004069243094318496,
      "loss": 5.7379,
      "mean_token_accuracy": 0.19299978613853455,
      "num_tokens": 67786490.0,
      "step": 26745
    },
    {
      "entropy": 6.150284290313721,
      "epoch": 3.087132140796307,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040689088011292024,
      "loss": 5.6758,
      "mean_token_accuracy": 0.19601108133792877,
      "num_tokens": 67799123.0,
      "step": 26750
    },
    {
      "entropy": 6.112508058547974,
      "epoch": 3.0877091748413155,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004068574463579066,
      "loss": 5.6202,
      "mean_token_accuracy": 0.192739400267601,
      "num_tokens": 67811860.0,
      "step": 26755
    },
    {
      "entropy": 6.093053340911865,
      "epoch": 3.0882862088863243,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040682400816793326,
      "loss": 5.7346,
      "mean_token_accuracy": 0.1894342765212059,
      "num_tokens": 67824251.0,
      "step": 26760
    },
    {
      "entropy": 6.243050622940063,
      "epoch": 3.0888632429313327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040679056554412485,
      "loss": 5.8139,
      "mean_token_accuracy": 0.18186367750167848,
      "num_tokens": 67837316.0,
      "step": 26765
    },
    {
      "entropy": 6.272124624252319,
      "epoch": 3.0894402769763416,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004067571184876063,
      "loss": 5.8028,
      "mean_token_accuracy": 0.19087813049554825,
      "num_tokens": 67850257.0,
      "step": 26770
    },
    {
      "entropy": 6.171517467498779,
      "epoch": 3.0900173110213505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040672366699950245,
      "loss": 5.74,
      "mean_token_accuracy": 0.19003414660692214,
      "num_tokens": 67862715.0,
      "step": 26775
    },
    {
      "entropy": 6.210683107376099,
      "epoch": 3.090594345066359,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040669021108093856,
      "loss": 5.7095,
      "mean_token_accuracy": 0.19171876460313797,
      "num_tokens": 67875282.0,
      "step": 26780
    },
    {
      "entropy": 6.120309114456177,
      "epoch": 3.0911713791113677,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004066567507330398,
      "loss": 5.6681,
      "mean_token_accuracy": 0.1940741464495659,
      "num_tokens": 67886797.0,
      "step": 26785
    },
    {
      "entropy": 6.016139268875122,
      "epoch": 3.091748413156376,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004066232859569316,
      "loss": 5.6429,
      "mean_token_accuracy": 0.19710179567337036,
      "num_tokens": 67899581.0,
      "step": 26790
    },
    {
      "entropy": 6.109029245376587,
      "epoch": 3.092325447201385,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004065898167537396,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19441143572330474,
      "num_tokens": 67911640.0,
      "step": 26795
    },
    {
      "entropy": 6.17063159942627,
      "epoch": 3.0929024812463934,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004065563431245895,
      "loss": 5.7205,
      "mean_token_accuracy": 0.1947186291217804,
      "num_tokens": 67925094.0,
      "step": 26800
    },
    {
      "entropy": 6.08087797164917,
      "epoch": 3.0934795152914023,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004065228650706071,
      "loss": 5.67,
      "mean_token_accuracy": 0.19813878685235978,
      "num_tokens": 67937774.0,
      "step": 26805
    },
    {
      "entropy": 6.165895557403564,
      "epoch": 3.0940565493364107,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040648938259291845,
      "loss": 5.6739,
      "mean_token_accuracy": 0.19740253537893296,
      "num_tokens": 67950240.0,
      "step": 26810
    },
    {
      "entropy": 6.167765760421753,
      "epoch": 3.0946335833814196,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004064558956926497,
      "loss": 5.6649,
      "mean_token_accuracy": 0.1949150264263153,
      "num_tokens": 67963503.0,
      "step": 26815
    },
    {
      "entropy": 6.140791893005371,
      "epoch": 3.095210617426428,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004064224043709273,
      "loss": 5.7213,
      "mean_token_accuracy": 0.19097324460744858,
      "num_tokens": 67974610.0,
      "step": 26820
    },
    {
      "entropy": 6.094710206985473,
      "epoch": 3.095787651471437,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004063889086288776,
      "loss": 5.6301,
      "mean_token_accuracy": 0.20011126697063447,
      "num_tokens": 67988172.0,
      "step": 26825
    },
    {
      "entropy": 6.116142749786377,
      "epoch": 3.0963646855164453,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00040635540846762713,
      "loss": 5.6781,
      "mean_token_accuracy": 0.19725194722414016,
      "num_tokens": 68001489.0,
      "step": 26830
    },
    {
      "entropy": 6.146768426895141,
      "epoch": 3.096941719561454,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004063219038883028,
      "loss": 5.7275,
      "mean_token_accuracy": 0.19413065761327744,
      "num_tokens": 68014291.0,
      "step": 26835
    },
    {
      "entropy": 6.159466171264649,
      "epoch": 3.0975187536064626,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00040628839489203144,
      "loss": 5.7027,
      "mean_token_accuracy": 0.18735834062099457,
      "num_tokens": 68026735.0,
      "step": 26840
    },
    {
      "entropy": 6.221561431884766,
      "epoch": 3.0980957876514714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040625488147994025,
      "loss": 5.8162,
      "mean_token_accuracy": 0.18988598585128785,
      "num_tokens": 68039943.0,
      "step": 26845
    },
    {
      "entropy": 6.087126350402832,
      "epoch": 3.0986728216964803,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040622136365315616,
      "loss": 5.6801,
      "mean_token_accuracy": 0.1903586581349373,
      "num_tokens": 68053358.0,
      "step": 26850
    },
    {
      "entropy": 6.160599279403686,
      "epoch": 3.0992498557414887,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040618784141280674,
      "loss": 5.6497,
      "mean_token_accuracy": 0.19572605639696122,
      "num_tokens": 68066002.0,
      "step": 26855
    },
    {
      "entropy": 6.123009252548218,
      "epoch": 3.0998268897864976,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040615431476001947,
      "loss": 5.646,
      "mean_token_accuracy": 0.20059366375207902,
      "num_tokens": 68079221.0,
      "step": 26860
    },
    {
      "entropy": 6.174410152435303,
      "epoch": 3.100403923831506,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004061207836959219,
      "loss": 5.7259,
      "mean_token_accuracy": 0.18995708227157593,
      "num_tokens": 68091388.0,
      "step": 26865
    },
    {
      "entropy": 6.194659042358398,
      "epoch": 3.100980957876515,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040608724822164193,
      "loss": 5.7886,
      "mean_token_accuracy": 0.18564740121364592,
      "num_tokens": 68104016.0,
      "step": 26870
    },
    {
      "entropy": 6.177530670166016,
      "epoch": 3.1015579919215233,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004060537083383076,
      "loss": 5.6287,
      "mean_token_accuracy": 0.2020028918981552,
      "num_tokens": 68117762.0,
      "step": 26875
    },
    {
      "entropy": 6.115578079223633,
      "epoch": 3.102135025966532,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040602016404704666,
      "loss": 5.6791,
      "mean_token_accuracy": 0.20041325688362122,
      "num_tokens": 68129913.0,
      "step": 26880
    },
    {
      "entropy": 6.1328596591949465,
      "epoch": 3.1027120600115405,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004059866153489878,
      "loss": 5.735,
      "mean_token_accuracy": 0.19435964077711104,
      "num_tokens": 68142582.0,
      "step": 26885
    },
    {
      "entropy": 6.231853914260864,
      "epoch": 3.1032890940565494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040595306224525904,
      "loss": 5.7308,
      "mean_token_accuracy": 0.19070784151554107,
      "num_tokens": 68156233.0,
      "step": 26890
    },
    {
      "entropy": 6.166604375839233,
      "epoch": 3.103866128101558,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004059195047369892,
      "loss": 5.6914,
      "mean_token_accuracy": 0.19259731322526932,
      "num_tokens": 68168672.0,
      "step": 26895
    },
    {
      "entropy": 6.145269966125488,
      "epoch": 3.1044431621465667,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004058859428253068,
      "loss": 5.7659,
      "mean_token_accuracy": 0.18957958817481996,
      "num_tokens": 68180934.0,
      "step": 26900
    },
    {
      "entropy": 6.152492523193359,
      "epoch": 3.105020196191575,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004058523765113407,
      "loss": 5.7027,
      "mean_token_accuracy": 0.19278210401535034,
      "num_tokens": 68193788.0,
      "step": 26905
    },
    {
      "entropy": 6.1945576667785645,
      "epoch": 3.105597230236584,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040581880579621996,
      "loss": 5.6802,
      "mean_token_accuracy": 0.19895323365926743,
      "num_tokens": 68204908.0,
      "step": 26910
    },
    {
      "entropy": 6.161127614974975,
      "epoch": 3.106174264281593,
      "grad_norm": 0.890625,
      "learning_rate": 0.00040578523068107367,
      "loss": 5.6974,
      "mean_token_accuracy": 0.1886743575334549,
      "num_tokens": 68217956.0,
      "step": 26915
    },
    {
      "entropy": 6.114270782470703,
      "epoch": 3.1067512983266012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040575165116703103,
      "loss": 5.6303,
      "mean_token_accuracy": 0.199447900056839,
      "num_tokens": 68229755.0,
      "step": 26920
    },
    {
      "entropy": 6.205735492706299,
      "epoch": 3.10732833237161,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040571806725522164,
      "loss": 5.7513,
      "mean_token_accuracy": 0.1872178941965103,
      "num_tokens": 68242620.0,
      "step": 26925
    },
    {
      "entropy": 6.036647510528565,
      "epoch": 3.1079053664166185,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040568447894677494,
      "loss": 5.6305,
      "mean_token_accuracy": 0.1973928615450859,
      "num_tokens": 68256045.0,
      "step": 26930
    },
    {
      "entropy": 6.119180202484131,
      "epoch": 3.1084824004616274,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040565088624282074,
      "loss": 5.6344,
      "mean_token_accuracy": 0.2013043314218521,
      "num_tokens": 68269378.0,
      "step": 26935
    },
    {
      "entropy": 6.179052352905273,
      "epoch": 3.109059434506636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004056172891444889,
      "loss": 5.6994,
      "mean_token_accuracy": 0.1950487896800041,
      "num_tokens": 68281575.0,
      "step": 26940
    },
    {
      "entropy": 6.142611742019653,
      "epoch": 3.1096364685516447,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040558368765290933,
      "loss": 5.6492,
      "mean_token_accuracy": 0.198221755027771,
      "num_tokens": 68293690.0,
      "step": 26945
    },
    {
      "entropy": 6.178358268737793,
      "epoch": 3.110213502596653,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040555008176921235,
      "loss": 5.7682,
      "mean_token_accuracy": 0.18755603730678558,
      "num_tokens": 68305200.0,
      "step": 26950
    },
    {
      "entropy": 6.1204619884490965,
      "epoch": 3.110790536641662,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004055164714945282,
      "loss": 5.6574,
      "mean_token_accuracy": 0.19494971483945847,
      "num_tokens": 68317259.0,
      "step": 26955
    },
    {
      "entropy": 6.102750539779663,
      "epoch": 3.1113675706866704,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040548285682998736,
      "loss": 5.6853,
      "mean_token_accuracy": 0.19803863018751144,
      "num_tokens": 68330101.0,
      "step": 26960
    },
    {
      "entropy": 6.143659782409668,
      "epoch": 3.111944604731679,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040544923777672046,
      "loss": 5.7166,
      "mean_token_accuracy": 0.19461364597082137,
      "num_tokens": 68342438.0,
      "step": 26965
    },
    {
      "entropy": 6.080714988708496,
      "epoch": 3.1125216387766876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004054156143358583,
      "loss": 5.648,
      "mean_token_accuracy": 0.20194706022739412,
      "num_tokens": 68355202.0,
      "step": 26970
    },
    {
      "entropy": 6.112496852874756,
      "epoch": 3.1130986728216965,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040538198650853156,
      "loss": 5.6267,
      "mean_token_accuracy": 0.19803008437156677,
      "num_tokens": 68366586.0,
      "step": 26975
    },
    {
      "entropy": 6.1533894538879395,
      "epoch": 3.1136757068667054,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004053483542958717,
      "loss": 5.7281,
      "mean_token_accuracy": 0.1937193751335144,
      "num_tokens": 68381132.0,
      "step": 26980
    },
    {
      "entropy": 6.163727331161499,
      "epoch": 3.114252740911714,
      "grad_norm": 1.0,
      "learning_rate": 0.00040531471769900953,
      "loss": 5.7361,
      "mean_token_accuracy": 0.1895759552717209,
      "num_tokens": 68392691.0,
      "step": 26985
    },
    {
      "entropy": 6.153784561157226,
      "epoch": 3.1148297749567226,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004052810767190766,
      "loss": 5.6966,
      "mean_token_accuracy": 0.19710321426391603,
      "num_tokens": 68405499.0,
      "step": 26990
    },
    {
      "entropy": 6.139175462722778,
      "epoch": 3.115406809001731,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040524743135720437,
      "loss": 5.7602,
      "mean_token_accuracy": 0.19449110925197602,
      "num_tokens": 68418485.0,
      "step": 26995
    },
    {
      "entropy": 6.17497501373291,
      "epoch": 3.11598384304674,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004052137816145245,
      "loss": 5.7286,
      "mean_token_accuracy": 0.19408611953258514,
      "num_tokens": 68432409.0,
      "step": 27000
    },
    {
      "epoch": 3.11598384304674,
      "eval_entropy": 5.987836458007144,
      "eval_loss": 5.857382297515869,
      "eval_mean_token_accuracy": 0.19283583052911846,
      "eval_num_tokens": 68432409.0,
      "eval_runtime": 22.748,
      "eval_samples_per_second": 1236.856,
      "eval_steps_per_second": 154.607,
      "step": 27000
    },
    {
      "entropy": 6.116707515716553,
      "epoch": 3.1165608770917483,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004051801274921688,
      "loss": 5.6403,
      "mean_token_accuracy": 0.19760415703058243,
      "num_tokens": 68444516.0,
      "step": 27005
    },
    {
      "entropy": 6.186652040481567,
      "epoch": 3.117137911136757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040514646899126905,
      "loss": 5.7633,
      "mean_token_accuracy": 0.1909859895706177,
      "num_tokens": 68458326.0,
      "step": 27010
    },
    {
      "entropy": 6.194436073303223,
      "epoch": 3.1177149451817656,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004051128061129576,
      "loss": 5.6982,
      "mean_token_accuracy": 0.18686971813440323,
      "num_tokens": 68472500.0,
      "step": 27015
    },
    {
      "entropy": 6.152800178527832,
      "epoch": 3.1182919792267745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040507913885836646,
      "loss": 5.7544,
      "mean_token_accuracy": 0.19872835725545884,
      "num_tokens": 68485280.0,
      "step": 27020
    },
    {
      "entropy": 6.150189208984375,
      "epoch": 3.118869013271783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004050454672286281,
      "loss": 5.6948,
      "mean_token_accuracy": 0.1894780769944191,
      "num_tokens": 68497792.0,
      "step": 27025
    },
    {
      "entropy": 6.073841762542725,
      "epoch": 3.1194460473167918,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004050117912248751,
      "loss": 5.6322,
      "mean_token_accuracy": 0.20315015465021133,
      "num_tokens": 68510444.0,
      "step": 27030
    },
    {
      "entropy": 6.088042211532593,
      "epoch": 3.1200230813618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040497811084824005,
      "loss": 5.6425,
      "mean_token_accuracy": 0.19691527187824248,
      "num_tokens": 68522414.0,
      "step": 27035
    },
    {
      "entropy": 6.1867858409881595,
      "epoch": 3.120600115406809,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004049444260998558,
      "loss": 5.6901,
      "mean_token_accuracy": 0.19433800131082535,
      "num_tokens": 68534389.0,
      "step": 27040
    },
    {
      "entropy": 6.076087093353271,
      "epoch": 3.1211771494518175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004049107369808553,
      "loss": 5.6633,
      "mean_token_accuracy": 0.19283735305070876,
      "num_tokens": 68545881.0,
      "step": 27045
    },
    {
      "entropy": 6.231611680984497,
      "epoch": 3.1217541834968263,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040487704349237175,
      "loss": 5.723,
      "mean_token_accuracy": 0.19022363275289536,
      "num_tokens": 68558758.0,
      "step": 27050
    },
    {
      "entropy": 6.139404582977295,
      "epoch": 3.122331217541835,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040484334563553836,
      "loss": 5.6617,
      "mean_token_accuracy": 0.20138685256242753,
      "num_tokens": 68570551.0,
      "step": 27055
    },
    {
      "entropy": 6.113898754119873,
      "epoch": 3.1229082515868436,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004048096434114885,
      "loss": 5.7726,
      "mean_token_accuracy": 0.18614590167999268,
      "num_tokens": 68583012.0,
      "step": 27060
    },
    {
      "entropy": 6.098631095886231,
      "epoch": 3.1234852856318525,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00040477593682135575,
      "loss": 5.6838,
      "mean_token_accuracy": 0.19528522193431855,
      "num_tokens": 68594235.0,
      "step": 27065
    },
    {
      "entropy": 6.15474214553833,
      "epoch": 3.124062319676861,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004047422258662739,
      "loss": 5.7062,
      "mean_token_accuracy": 0.1918144553899765,
      "num_tokens": 68605978.0,
      "step": 27070
    },
    {
      "entropy": 6.1452208995819095,
      "epoch": 3.1246393537218697,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00040470851054737666,
      "loss": 5.7057,
      "mean_token_accuracy": 0.19720974415540696,
      "num_tokens": 68619385.0,
      "step": 27075
    },
    {
      "entropy": 6.080608129501343,
      "epoch": 3.125216387766878,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004046747908657982,
      "loss": 5.656,
      "mean_token_accuracy": 0.19955915808677674,
      "num_tokens": 68631986.0,
      "step": 27080
    },
    {
      "entropy": 6.151964426040649,
      "epoch": 3.125793421811887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040464106682267236,
      "loss": 5.6859,
      "mean_token_accuracy": 0.19174427539110184,
      "num_tokens": 68644262.0,
      "step": 27085
    },
    {
      "entropy": 6.129231882095337,
      "epoch": 3.1263704558568954,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004046073384191338,
      "loss": 5.7095,
      "mean_token_accuracy": 0.1935574308037758,
      "num_tokens": 68656941.0,
      "step": 27090
    },
    {
      "entropy": 6.157695579528808,
      "epoch": 3.1269474899019043,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040457360565631674,
      "loss": 5.7965,
      "mean_token_accuracy": 0.18418788611888887,
      "num_tokens": 68669231.0,
      "step": 27095
    },
    {
      "entropy": 6.1517857074737545,
      "epoch": 3.1275245239469127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040453986853535587,
      "loss": 5.6884,
      "mean_token_accuracy": 0.19333977550268172,
      "num_tokens": 68681882.0,
      "step": 27100
    },
    {
      "entropy": 6.1447900295257565,
      "epoch": 3.1281015579919216,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004045061270573858,
      "loss": 5.5659,
      "mean_token_accuracy": 0.2177928566932678,
      "num_tokens": 68693970.0,
      "step": 27105
    },
    {
      "entropy": 6.028097820281983,
      "epoch": 3.12867859203693,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040447238122354146,
      "loss": 5.6438,
      "mean_token_accuracy": 0.195227912068367,
      "num_tokens": 68706367.0,
      "step": 27110
    },
    {
      "entropy": 6.2195396900177,
      "epoch": 3.129255626081939,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004044386310349579,
      "loss": 5.8069,
      "mean_token_accuracy": 0.18571565449237823,
      "num_tokens": 68718644.0,
      "step": 27115
    },
    {
      "entropy": 6.126541423797607,
      "epoch": 3.1298326601269473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040440487649277037,
      "loss": 5.633,
      "mean_token_accuracy": 0.1973568856716156,
      "num_tokens": 68730545.0,
      "step": 27120
    },
    {
      "entropy": 6.160487508773803,
      "epoch": 3.130409694171956,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040437111759811397,
      "loss": 5.7312,
      "mean_token_accuracy": 0.19705376625061036,
      "num_tokens": 68743093.0,
      "step": 27125
    },
    {
      "entropy": 6.126802062988281,
      "epoch": 3.130986728216965,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040433735435212436,
      "loss": 5.7251,
      "mean_token_accuracy": 0.19497810304164886,
      "num_tokens": 68756437.0,
      "step": 27130
    },
    {
      "entropy": 6.08596363067627,
      "epoch": 3.1315637622619734,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004043035867559371,
      "loss": 5.6938,
      "mean_token_accuracy": 0.1972505435347557,
      "num_tokens": 68769380.0,
      "step": 27135
    },
    {
      "entropy": 6.177377033233642,
      "epoch": 3.1321407963069823,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040426981481068785,
      "loss": 5.6979,
      "mean_token_accuracy": 0.19830397069454192,
      "num_tokens": 68782390.0,
      "step": 27140
    },
    {
      "entropy": 6.099611759185791,
      "epoch": 3.1327178303519907,
      "grad_norm": 1.140625,
      "learning_rate": 0.00040423603851751273,
      "loss": 5.5893,
      "mean_token_accuracy": 0.2013296142220497,
      "num_tokens": 68793931.0,
      "step": 27145
    },
    {
      "entropy": 6.077713203430176,
      "epoch": 3.1332948643969996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040420225787754754,
      "loss": 5.7283,
      "mean_token_accuracy": 0.19569722414016724,
      "num_tokens": 68806020.0,
      "step": 27150
    },
    {
      "entropy": 6.1687322616577145,
      "epoch": 3.133871898442008,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004041684728919286,
      "loss": 5.7726,
      "mean_token_accuracy": 0.1871213957667351,
      "num_tokens": 68818616.0,
      "step": 27155
    },
    {
      "entropy": 6.177014255523682,
      "epoch": 3.134448932487017,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040413468356179224,
      "loss": 5.7204,
      "mean_token_accuracy": 0.19158955216407775,
      "num_tokens": 68831549.0,
      "step": 27160
    },
    {
      "entropy": 6.20126633644104,
      "epoch": 3.1350259665320253,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000404100889888275,
      "loss": 5.6866,
      "mean_token_accuracy": 0.19375785291194916,
      "num_tokens": 68842551.0,
      "step": 27165
    },
    {
      "entropy": 6.147972631454468,
      "epoch": 3.135603000577034,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004040670918725133,
      "loss": 5.7273,
      "mean_token_accuracy": 0.19284260720014573,
      "num_tokens": 68855525.0,
      "step": 27170
    },
    {
      "entropy": 6.108228158950806,
      "epoch": 3.1361800346220425,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004040332895156443,
      "loss": 5.7106,
      "mean_token_accuracy": 0.19711076021194457,
      "num_tokens": 68869570.0,
      "step": 27175
    },
    {
      "entropy": 6.18882303237915,
      "epoch": 3.1367570686670514,
      "grad_norm": 1.25,
      "learning_rate": 0.0004039994828188045,
      "loss": 5.7008,
      "mean_token_accuracy": 0.19134557098150254,
      "num_tokens": 68881392.0,
      "step": 27180
    },
    {
      "entropy": 6.0376002311706545,
      "epoch": 3.13733410271206,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040396567178313124,
      "loss": 5.6718,
      "mean_token_accuracy": 0.195427830517292,
      "num_tokens": 68894107.0,
      "step": 27185
    },
    {
      "entropy": 6.24177303314209,
      "epoch": 3.1379111367570687,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040393185640976174,
      "loss": 5.7615,
      "mean_token_accuracy": 0.19118544459342957,
      "num_tokens": 68905734.0,
      "step": 27190
    },
    {
      "entropy": 6.232033109664917,
      "epoch": 3.138488170802077,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004038980366998332,
      "loss": 5.7701,
      "mean_token_accuracy": 0.187275230884552,
      "num_tokens": 68918996.0,
      "step": 27195
    },
    {
      "entropy": 6.056928873062134,
      "epoch": 3.139065204847086,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004038642126544833,
      "loss": 5.6535,
      "mean_token_accuracy": 0.20074088275432586,
      "num_tokens": 68931381.0,
      "step": 27200
    },
    {
      "entropy": 6.149992561340332,
      "epoch": 3.139642238892095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004038303842748496,
      "loss": 5.7509,
      "mean_token_accuracy": 0.1898006409406662,
      "num_tokens": 68943615.0,
      "step": 27205
    },
    {
      "entropy": 6.124445962905884,
      "epoch": 3.1402192729371032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040379655156206983,
      "loss": 5.6156,
      "mean_token_accuracy": 0.19554246962070465,
      "num_tokens": 68955115.0,
      "step": 27210
    },
    {
      "entropy": 6.05906662940979,
      "epoch": 3.140796306982112,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004037627145172822,
      "loss": 5.6075,
      "mean_token_accuracy": 0.20224371403455735,
      "num_tokens": 68968042.0,
      "step": 27215
    },
    {
      "entropy": 6.12451171875,
      "epoch": 3.1413733410271205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004037288731416245,
      "loss": 5.6571,
      "mean_token_accuracy": 0.2077217683196068,
      "num_tokens": 68982178.0,
      "step": 27220
    },
    {
      "entropy": 6.094468832015991,
      "epoch": 3.1419503750721294,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004036950274362351,
      "loss": 5.7369,
      "mean_token_accuracy": 0.18722873628139497,
      "num_tokens": 68994546.0,
      "step": 27225
    },
    {
      "entropy": 6.16364541053772,
      "epoch": 3.142527409117138,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004036611774022524,
      "loss": 5.6819,
      "mean_token_accuracy": 0.19240596890449524,
      "num_tokens": 69007134.0,
      "step": 27230
    },
    {
      "entropy": 6.1228940963745115,
      "epoch": 3.1431044431621467,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004036273230408149,
      "loss": 5.6921,
      "mean_token_accuracy": 0.196954445540905,
      "num_tokens": 69019668.0,
      "step": 27235
    },
    {
      "entropy": 6.17612509727478,
      "epoch": 3.143681477207155,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004035934643530613,
      "loss": 5.7308,
      "mean_token_accuracy": 0.19115004688501358,
      "num_tokens": 69032127.0,
      "step": 27240
    },
    {
      "entropy": 6.162550163269043,
      "epoch": 3.144258511252164,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040355960134013036,
      "loss": 5.7307,
      "mean_token_accuracy": 0.1894148588180542,
      "num_tokens": 69044926.0,
      "step": 27245
    },
    {
      "entropy": 6.131181478500366,
      "epoch": 3.1448355452971724,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004035257340031611,
      "loss": 5.644,
      "mean_token_accuracy": 0.19221512973308563,
      "num_tokens": 69056326.0,
      "step": 27250
    },
    {
      "entropy": 6.168639469146728,
      "epoch": 3.145412579342181,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004034918623432926,
      "loss": 5.7516,
      "mean_token_accuracy": 0.1910143032670021,
      "num_tokens": 69069696.0,
      "step": 27255
    },
    {
      "entropy": 6.16878457069397,
      "epoch": 3.14598961338719,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040345798636166424,
      "loss": 5.7069,
      "mean_token_accuracy": 0.19244280010461806,
      "num_tokens": 69082517.0,
      "step": 27260
    },
    {
      "entropy": 6.08049168586731,
      "epoch": 3.1465666474321985,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040342410605941526,
      "loss": 5.6425,
      "mean_token_accuracy": 0.19867913275957108,
      "num_tokens": 69095774.0,
      "step": 27265
    },
    {
      "entropy": 6.1829994201660154,
      "epoch": 3.1471436814772074,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040339022143768524,
      "loss": 5.6909,
      "mean_token_accuracy": 0.1935090869665146,
      "num_tokens": 69109575.0,
      "step": 27270
    },
    {
      "entropy": 6.073039579391479,
      "epoch": 3.1477207155222158,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00040335633249761387,
      "loss": 5.6837,
      "mean_token_accuracy": 0.19323285520076752,
      "num_tokens": 69123365.0,
      "step": 27275
    },
    {
      "entropy": 6.162595129013061,
      "epoch": 3.1482977495672246,
      "grad_norm": 0.875,
      "learning_rate": 0.00040332243924034105,
      "loss": 5.7336,
      "mean_token_accuracy": 0.1946937620639801,
      "num_tokens": 69136187.0,
      "step": 27280
    },
    {
      "entropy": 6.097646045684814,
      "epoch": 3.148874783612233,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004032885416670068,
      "loss": 5.6746,
      "mean_token_accuracy": 0.20315638929605484,
      "num_tokens": 69150271.0,
      "step": 27285
    },
    {
      "entropy": 6.142451047897339,
      "epoch": 3.149451817657242,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004032546397787511,
      "loss": 5.7284,
      "mean_token_accuracy": 0.19465454667806625,
      "num_tokens": 69161906.0,
      "step": 27290
    },
    {
      "entropy": 6.171128273010254,
      "epoch": 3.1500288517022503,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004032207335767142,
      "loss": 5.6652,
      "mean_token_accuracy": 0.19658414870500565,
      "num_tokens": 69174367.0,
      "step": 27295
    },
    {
      "entropy": 6.101913928985596,
      "epoch": 3.150605885747259,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004031868230620368,
      "loss": 5.7316,
      "mean_token_accuracy": 0.19344296157360077,
      "num_tokens": 69186970.0,
      "step": 27300
    },
    {
      "entropy": 6.155976343154907,
      "epoch": 3.1511829197922676,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004031529082358592,
      "loss": 5.7205,
      "mean_token_accuracy": 0.19540518820285796,
      "num_tokens": 69199347.0,
      "step": 27305
    },
    {
      "entropy": 6.0524097919464115,
      "epoch": 3.1517599538372765,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004031189890993222,
      "loss": 5.6576,
      "mean_token_accuracy": 0.19292837083339692,
      "num_tokens": 69211710.0,
      "step": 27310
    },
    {
      "entropy": 6.11064658164978,
      "epoch": 3.152336987882285,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040308506565356664,
      "loss": 5.6851,
      "mean_token_accuracy": 0.19722912907600404,
      "num_tokens": 69224153.0,
      "step": 27315
    },
    {
      "entropy": 6.184148359298706,
      "epoch": 3.1529140219272938,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040305113789973354,
      "loss": 5.719,
      "mean_token_accuracy": 0.18931434750556947,
      "num_tokens": 69236046.0,
      "step": 27320
    },
    {
      "entropy": 6.10137906074524,
      "epoch": 3.153491055972302,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004030172058389639,
      "loss": 5.661,
      "mean_token_accuracy": 0.1945364996790886,
      "num_tokens": 69247997.0,
      "step": 27325
    },
    {
      "entropy": 6.120266151428223,
      "epoch": 3.154068090017311,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040298326947239935,
      "loss": 5.673,
      "mean_token_accuracy": 0.2005025789141655,
      "num_tokens": 69261189.0,
      "step": 27330
    },
    {
      "entropy": 6.12740535736084,
      "epoch": 3.15464512406232,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004029493288011809,
      "loss": 5.7153,
      "mean_token_accuracy": 0.19623589366674424,
      "num_tokens": 69274039.0,
      "step": 27335
    },
    {
      "entropy": 6.009217548370361,
      "epoch": 3.1552221581073283,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004029153838264504,
      "loss": 5.6667,
      "mean_token_accuracy": 0.2013350784778595,
      "num_tokens": 69286759.0,
      "step": 27340
    },
    {
      "entropy": 6.131100177764893,
      "epoch": 3.155799192152337,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004028814345493496,
      "loss": 5.6996,
      "mean_token_accuracy": 0.19019987136125566,
      "num_tokens": 69299619.0,
      "step": 27345
    },
    {
      "entropy": 6.076996326446533,
      "epoch": 3.1563762261973456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00040284748097102013,
      "loss": 5.5706,
      "mean_token_accuracy": 0.2010394126176834,
      "num_tokens": 69311593.0,
      "step": 27350
    },
    {
      "entropy": 6.149994850158691,
      "epoch": 3.1569532602423545,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040281352309260414,
      "loss": 5.7157,
      "mean_token_accuracy": 0.1869497463107109,
      "num_tokens": 69324686.0,
      "step": 27355
    },
    {
      "entropy": 6.121657848358154,
      "epoch": 3.157530294287363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004027795609152439,
      "loss": 5.6574,
      "mean_token_accuracy": 0.19951744228601456,
      "num_tokens": 69337042.0,
      "step": 27360
    },
    {
      "entropy": 6.031878662109375,
      "epoch": 3.1581073283323717,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004027455944400815,
      "loss": 5.5803,
      "mean_token_accuracy": 0.2017280265688896,
      "num_tokens": 69350244.0,
      "step": 27365
    },
    {
      "entropy": 6.1586519241333,
      "epoch": 3.15868436237738,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004027116236682596,
      "loss": 5.7628,
      "mean_token_accuracy": 0.18867990821599961,
      "num_tokens": 69362813.0,
      "step": 27370
    },
    {
      "entropy": 6.249085521697998,
      "epoch": 3.159261396422389,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004026776486009206,
      "loss": 5.7815,
      "mean_token_accuracy": 0.1933159902691841,
      "num_tokens": 69375397.0,
      "step": 27375
    },
    {
      "entropy": 6.158890342712402,
      "epoch": 3.1598384304673974,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00040264366923920735,
      "loss": 5.7501,
      "mean_token_accuracy": 0.18520111739635467,
      "num_tokens": 69388107.0,
      "step": 27380
    },
    {
      "entropy": 6.107667922973633,
      "epoch": 3.1604154645124063,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040260968558426265,
      "loss": 5.7069,
      "mean_token_accuracy": 0.19940268993377686,
      "num_tokens": 69401894.0,
      "step": 27385
    },
    {
      "entropy": 6.1452759265899655,
      "epoch": 3.1609924985574147,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004025756976372296,
      "loss": 5.6449,
      "mean_token_accuracy": 0.20127691626548766,
      "num_tokens": 69414581.0,
      "step": 27390
    },
    {
      "entropy": 6.146435308456421,
      "epoch": 3.1615695326024236,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004025417053992513,
      "loss": 5.6883,
      "mean_token_accuracy": 0.1978426143527031,
      "num_tokens": 69428943.0,
      "step": 27395
    },
    {
      "entropy": 6.183188009262085,
      "epoch": 3.162146566647432,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004025077088714711,
      "loss": 5.7551,
      "mean_token_accuracy": 0.1931930214166641,
      "num_tokens": 69442670.0,
      "step": 27400
    },
    {
      "entropy": 6.147484397888183,
      "epoch": 3.162723600692441,
      "grad_norm": 1.0,
      "learning_rate": 0.0004024737080550324,
      "loss": 5.6901,
      "mean_token_accuracy": 0.19142468571662902,
      "num_tokens": 69454784.0,
      "step": 27405
    },
    {
      "entropy": 6.062293624877929,
      "epoch": 3.1633006347374497,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004024397029510789,
      "loss": 5.6586,
      "mean_token_accuracy": 0.20212570428848267,
      "num_tokens": 69467683.0,
      "step": 27410
    },
    {
      "entropy": 6.136881971359253,
      "epoch": 3.163877668782458,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004024056935607543,
      "loss": 5.6431,
      "mean_token_accuracy": 0.2008592501282692,
      "num_tokens": 69480741.0,
      "step": 27415
    },
    {
      "entropy": 6.170301675796509,
      "epoch": 3.164454702827467,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040237167988520245,
      "loss": 5.7207,
      "mean_token_accuracy": 0.18927052319049836,
      "num_tokens": 69493130.0,
      "step": 27420
    },
    {
      "entropy": 6.0939685821533205,
      "epoch": 3.1650317368724754,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004023376619255674,
      "loss": 5.6918,
      "mean_token_accuracy": 0.20142821371555328,
      "num_tokens": 69504962.0,
      "step": 27425
    },
    {
      "entropy": 6.04275336265564,
      "epoch": 3.1656087709174843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040230363968299344,
      "loss": 5.6502,
      "mean_token_accuracy": 0.19614885449409486,
      "num_tokens": 69517295.0,
      "step": 27430
    },
    {
      "entropy": 6.176437282562256,
      "epoch": 3.1661858049624927,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004022696131586247,
      "loss": 5.7175,
      "mean_token_accuracy": 0.19070173054933548,
      "num_tokens": 69529768.0,
      "step": 27435
    },
    {
      "entropy": 6.123783254623413,
      "epoch": 3.1667628390075016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004022355823536058,
      "loss": 5.7042,
      "mean_token_accuracy": 0.20241218507289888,
      "num_tokens": 69541746.0,
      "step": 27440
    },
    {
      "entropy": 6.164044761657715,
      "epoch": 3.16733987305251,
      "grad_norm": 0.9375,
      "learning_rate": 0.00040220154726908127,
      "loss": 5.6968,
      "mean_token_accuracy": 0.19764174222946168,
      "num_tokens": 69553778.0,
      "step": 27445
    },
    {
      "entropy": 6.183361434936524,
      "epoch": 3.167916907097519,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004021675079061959,
      "loss": 5.7013,
      "mean_token_accuracy": 0.19016370326280593,
      "num_tokens": 69566317.0,
      "step": 27450
    },
    {
      "entropy": 6.105200529098511,
      "epoch": 3.1684939411425272,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00040213346426609453,
      "loss": 5.6403,
      "mean_token_accuracy": 0.2026249185204506,
      "num_tokens": 69577656.0,
      "step": 27455
    },
    {
      "entropy": 6.113854598999024,
      "epoch": 3.169070975187536,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040209941634992226,
      "loss": 5.7087,
      "mean_token_accuracy": 0.1921801671385765,
      "num_tokens": 69589020.0,
      "step": 27460
    },
    {
      "entropy": 6.186026191711425,
      "epoch": 3.1696480092325445,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004020653641588242,
      "loss": 5.6498,
      "mean_token_accuracy": 0.19333481341600417,
      "num_tokens": 69602003.0,
      "step": 27465
    },
    {
      "entropy": 6.191118383407593,
      "epoch": 3.1702250432775534,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004020313076939458,
      "loss": 5.7352,
      "mean_token_accuracy": 0.18799963891506194,
      "num_tokens": 69614123.0,
      "step": 27470
    },
    {
      "entropy": 6.093089866638183,
      "epoch": 3.170802077322562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004019972469564325,
      "loss": 5.5549,
      "mean_token_accuracy": 0.20680705159902574,
      "num_tokens": 69628490.0,
      "step": 27475
    },
    {
      "entropy": 6.093897008895874,
      "epoch": 3.1713791113675707,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040196318194742975,
      "loss": 5.6792,
      "mean_token_accuracy": 0.19264387339353561,
      "num_tokens": 69641313.0,
      "step": 27480
    },
    {
      "entropy": 6.08471269607544,
      "epoch": 3.1719561454125795,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040192911266808356,
      "loss": 5.5489,
      "mean_token_accuracy": 0.20377940088510513,
      "num_tokens": 69652849.0,
      "step": 27485
    },
    {
      "entropy": 6.088257503509522,
      "epoch": 3.172533179457588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004018950391195397,
      "loss": 5.7017,
      "mean_token_accuracy": 0.19392950981855392,
      "num_tokens": 69664396.0,
      "step": 27490
    },
    {
      "entropy": 6.1147418975830075,
      "epoch": 3.173110213502597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004018609613029442,
      "loss": 5.6611,
      "mean_token_accuracy": 0.19233012199401855,
      "num_tokens": 69677022.0,
      "step": 27495
    },
    {
      "entropy": 6.125004243850708,
      "epoch": 3.1736872475476052,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004018268792194433,
      "loss": 5.695,
      "mean_token_accuracy": 0.1952821910381317,
      "num_tokens": 69689132.0,
      "step": 27500
    },
    {
      "entropy": 6.169435977935791,
      "epoch": 3.174264281592614,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00040179279287018334,
      "loss": 5.7129,
      "mean_token_accuracy": 0.1946713373064995,
      "num_tokens": 69701704.0,
      "step": 27505
    },
    {
      "entropy": 6.156677865982056,
      "epoch": 3.1748413156376225,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00040175870225631066,
      "loss": 5.7784,
      "mean_token_accuracy": 0.19002797305583954,
      "num_tokens": 69715362.0,
      "step": 27510
    },
    {
      "entropy": 6.157092666625976,
      "epoch": 3.1754183496826314,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004017246073789721,
      "loss": 5.6943,
      "mean_token_accuracy": 0.1891680344939232,
      "num_tokens": 69726745.0,
      "step": 27515
    },
    {
      "entropy": 6.092809247970581,
      "epoch": 3.17599538372764,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004016905082393143,
      "loss": 5.6202,
      "mean_token_accuracy": 0.19824722558259963,
      "num_tokens": 69740074.0,
      "step": 27520
    },
    {
      "entropy": 6.0835113525390625,
      "epoch": 3.1765724177726486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040165640483848415,
      "loss": 5.6956,
      "mean_token_accuracy": 0.20157905220985411,
      "num_tokens": 69751220.0,
      "step": 27525
    },
    {
      "entropy": 6.111167526245117,
      "epoch": 3.177149451817657,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040162229717762883,
      "loss": 5.7535,
      "mean_token_accuracy": 0.19212746173143386,
      "num_tokens": 69764532.0,
      "step": 27530
    },
    {
      "entropy": 6.1592050075531,
      "epoch": 3.177726485862666,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004015881852578953,
      "loss": 5.6895,
      "mean_token_accuracy": 0.195404751598835,
      "num_tokens": 69776862.0,
      "step": 27535
    },
    {
      "entropy": 6.094680500030518,
      "epoch": 3.1783035199076743,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040155406908043114,
      "loss": 5.6608,
      "mean_token_accuracy": 0.19931994527578353,
      "num_tokens": 69789667.0,
      "step": 27540
    },
    {
      "entropy": 6.104642629623413,
      "epoch": 3.178880553952683,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004015199486463837,
      "loss": 5.682,
      "mean_token_accuracy": 0.1939501479268074,
      "num_tokens": 69802112.0,
      "step": 27545
    },
    {
      "entropy": 6.108482503890992,
      "epoch": 3.179457587997692,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004014858239569006,
      "loss": 5.705,
      "mean_token_accuracy": 0.1937064126133919,
      "num_tokens": 69813908.0,
      "step": 27550
    },
    {
      "entropy": 6.108286333084107,
      "epoch": 3.1800346220427005,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040145169501312966,
      "loss": 5.6166,
      "mean_token_accuracy": 0.20078689604997635,
      "num_tokens": 69825920.0,
      "step": 27555
    },
    {
      "entropy": 6.146532011032105,
      "epoch": 3.1806116560877093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004014175618162188,
      "loss": 5.6735,
      "mean_token_accuracy": 0.19559755474328994,
      "num_tokens": 69839657.0,
      "step": 27560
    },
    {
      "entropy": 6.139887762069702,
      "epoch": 3.1811886901327178,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00040138342436731603,
      "loss": 5.6782,
      "mean_token_accuracy": 0.19961002320051194,
      "num_tokens": 69851778.0,
      "step": 27565
    },
    {
      "entropy": 6.114302396774292,
      "epoch": 3.1817657241777266,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040134928266756954,
      "loss": 5.7056,
      "mean_token_accuracy": 0.19824820905923843,
      "num_tokens": 69865011.0,
      "step": 27570
    },
    {
      "entropy": 6.140068292617798,
      "epoch": 3.182342758222735,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004013151367181277,
      "loss": 5.7391,
      "mean_token_accuracy": 0.1893114075064659,
      "num_tokens": 69877740.0,
      "step": 27575
    },
    {
      "entropy": 6.200154972076416,
      "epoch": 3.182919792267744,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004012809865201391,
      "loss": 5.7108,
      "mean_token_accuracy": 0.1902512013912201,
      "num_tokens": 69890102.0,
      "step": 27580
    },
    {
      "entropy": 6.090849781036377,
      "epoch": 3.1834968263127523,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040124683207475205,
      "loss": 5.6367,
      "mean_token_accuracy": 0.19401753395795823,
      "num_tokens": 69902500.0,
      "step": 27585
    },
    {
      "entropy": 6.120217800140381,
      "epoch": 3.184073860357761,
      "grad_norm": 0.921875,
      "learning_rate": 0.00040121267338311556,
      "loss": 5.7241,
      "mean_token_accuracy": 0.19053917676210402,
      "num_tokens": 69915013.0,
      "step": 27590
    },
    {
      "entropy": 6.148542022705078,
      "epoch": 3.1846508944027696,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040117851044637855,
      "loss": 5.5684,
      "mean_token_accuracy": 0.21735063940286636,
      "num_tokens": 69927149.0,
      "step": 27595
    },
    {
      "entropy": 6.162502908706665,
      "epoch": 3.1852279284477785,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040114434326569005,
      "loss": 5.6982,
      "mean_token_accuracy": 0.1941469430923462,
      "num_tokens": 69939924.0,
      "step": 27600
    },
    {
      "entropy": 6.063241052627563,
      "epoch": 3.185804962492787,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00040111017184219915,
      "loss": 5.6569,
      "mean_token_accuracy": 0.19664947241544722,
      "num_tokens": 69952068.0,
      "step": 27605
    },
    {
      "entropy": 6.0914960384368895,
      "epoch": 3.1863819965377957,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040107599617705527,
      "loss": 5.6537,
      "mean_token_accuracy": 0.1954856261610985,
      "num_tokens": 69965270.0,
      "step": 27610
    },
    {
      "entropy": 6.204733514785767,
      "epoch": 3.1869590305828046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000401041816271408,
      "loss": 5.7672,
      "mean_token_accuracy": 0.18885004371404648,
      "num_tokens": 69977138.0,
      "step": 27615
    },
    {
      "entropy": 6.179504776000977,
      "epoch": 3.187536064627813,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040100763212640666,
      "loss": 5.7339,
      "mean_token_accuracy": 0.18840832859277726,
      "num_tokens": 69989599.0,
      "step": 27620
    },
    {
      "entropy": 6.190329599380493,
      "epoch": 3.188113098672822,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004009734437432013,
      "loss": 5.7558,
      "mean_token_accuracy": 0.1908293694257736,
      "num_tokens": 70000887.0,
      "step": 27625
    },
    {
      "entropy": 6.067487621307373,
      "epoch": 3.1886901327178303,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040093925112294173,
      "loss": 5.6362,
      "mean_token_accuracy": 0.197028449177742,
      "num_tokens": 70013868.0,
      "step": 27630
    },
    {
      "entropy": 6.186224365234375,
      "epoch": 3.189267166762839,
      "grad_norm": 0.890625,
      "learning_rate": 0.000400905054266778,
      "loss": 5.6757,
      "mean_token_accuracy": 0.19788626581430435,
      "num_tokens": 70026307.0,
      "step": 27635
    },
    {
      "entropy": 6.0445051193237305,
      "epoch": 3.1898442008078476,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004008708531758603,
      "loss": 5.6291,
      "mean_token_accuracy": 0.21009810119867325,
      "num_tokens": 70040481.0,
      "step": 27640
    },
    {
      "entropy": 6.059706497192383,
      "epoch": 3.1904212348528564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000400836647851339,
      "loss": 5.6197,
      "mean_token_accuracy": 0.1999824598431587,
      "num_tokens": 70052067.0,
      "step": 27645
    },
    {
      "entropy": 6.1772455215454105,
      "epoch": 3.190998268897865,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004008024382943645,
      "loss": 5.773,
      "mean_token_accuracy": 0.18657677620649338,
      "num_tokens": 70064513.0,
      "step": 27650
    },
    {
      "entropy": 6.083112096786499,
      "epoch": 3.1915753029428737,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00040076822450608757,
      "loss": 5.6362,
      "mean_token_accuracy": 0.19627652764320375,
      "num_tokens": 70077496.0,
      "step": 27655
    },
    {
      "entropy": 6.140889978408813,
      "epoch": 3.192152336987882,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004007340064876588,
      "loss": 5.7466,
      "mean_token_accuracy": 0.19131832122802733,
      "num_tokens": 70091035.0,
      "step": 27660
    },
    {
      "entropy": 6.203101968765258,
      "epoch": 3.192729371032891,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004006997842402292,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1944819062948227,
      "num_tokens": 70104773.0,
      "step": 27665
    },
    {
      "entropy": 6.126469659805298,
      "epoch": 3.1933064050778994,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004006655577649498,
      "loss": 5.7176,
      "mean_token_accuracy": 0.1977926805615425,
      "num_tokens": 70118019.0,
      "step": 27670
    },
    {
      "entropy": 6.081540441513061,
      "epoch": 3.1938834391229083,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004006313270629718,
      "loss": 5.6952,
      "mean_token_accuracy": 0.19859555661678313,
      "num_tokens": 70130642.0,
      "step": 27675
    },
    {
      "entropy": 6.128610134124756,
      "epoch": 3.1944604731679167,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040059709213544656,
      "loss": 5.5971,
      "mean_token_accuracy": 0.20044137835502623,
      "num_tokens": 70144053.0,
      "step": 27680
    },
    {
      "entropy": 6.1069989681243895,
      "epoch": 3.1950375072129256,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004005628529835255,
      "loss": 5.6836,
      "mean_token_accuracy": 0.19584569483995437,
      "num_tokens": 70156765.0,
      "step": 27685
    },
    {
      "entropy": 6.095725727081299,
      "epoch": 3.1956145412579344,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004005286096083602,
      "loss": 5.6617,
      "mean_token_accuracy": 0.19987715035676956,
      "num_tokens": 70169515.0,
      "step": 27690
    },
    {
      "entropy": 6.087531423568725,
      "epoch": 3.196191575302943,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00040049436201110246,
      "loss": 5.6744,
      "mean_token_accuracy": 0.1999328151345253,
      "num_tokens": 70183958.0,
      "step": 27695
    },
    {
      "entropy": 6.164551687240601,
      "epoch": 3.1967686093479517,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040046011019290417,
      "loss": 5.7439,
      "mean_token_accuracy": 0.18895459324121475,
      "num_tokens": 70196381.0,
      "step": 27700
    },
    {
      "entropy": 6.085721635818482,
      "epoch": 3.19734564339296,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040042585415491747,
      "loss": 5.6887,
      "mean_token_accuracy": 0.1930871933698654,
      "num_tokens": 70208685.0,
      "step": 27705
    },
    {
      "entropy": 6.181350183486939,
      "epoch": 3.197922677437969,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004003915938982944,
      "loss": 5.7457,
      "mean_token_accuracy": 0.18845361173152925,
      "num_tokens": 70220651.0,
      "step": 27710
    },
    {
      "entropy": 6.14144868850708,
      "epoch": 3.1984997114829774,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004003573294241875,
      "loss": 5.7247,
      "mean_token_accuracy": 0.18952881395816804,
      "num_tokens": 70233315.0,
      "step": 27715
    },
    {
      "entropy": 6.095832061767578,
      "epoch": 3.1990767455279863,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040032306073374893,
      "loss": 5.6499,
      "mean_token_accuracy": 0.19654903709888458,
      "num_tokens": 70244845.0,
      "step": 27720
    },
    {
      "entropy": 6.107531642913818,
      "epoch": 3.1996537795729947,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004002887878281315,
      "loss": 5.7119,
      "mean_token_accuracy": 0.1910250574350357,
      "num_tokens": 70258242.0,
      "step": 27725
    },
    {
      "entropy": 6.170880365371704,
      "epoch": 3.2002308136180035,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00040025451070848793,
      "loss": 5.7643,
      "mean_token_accuracy": 0.19094973653554917,
      "num_tokens": 70270568.0,
      "step": 27730
    },
    {
      "entropy": 6.138811492919922,
      "epoch": 3.200807847663012,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040022022937597114,
      "loss": 5.6635,
      "mean_token_accuracy": 0.1937575876712799,
      "num_tokens": 70284180.0,
      "step": 27735
    },
    {
      "entropy": 6.183554220199585,
      "epoch": 3.201384881708021,
      "grad_norm": 1.125,
      "learning_rate": 0.00040018594383173407,
      "loss": 5.7699,
      "mean_token_accuracy": 0.18765608668327333,
      "num_tokens": 70296970.0,
      "step": 27740
    },
    {
      "entropy": 6.1393286228179935,
      "epoch": 3.2019619157530292,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040015165407693005,
      "loss": 5.6642,
      "mean_token_accuracy": 0.195418281853199,
      "num_tokens": 70308814.0,
      "step": 27745
    },
    {
      "entropy": 6.093904209136963,
      "epoch": 3.202538949798038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004001173601127123,
      "loss": 5.6902,
      "mean_token_accuracy": 0.20344341844320296,
      "num_tokens": 70320608.0,
      "step": 27750
    },
    {
      "entropy": 6.03027720451355,
      "epoch": 3.2031159838430465,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040008306194023427,
      "loss": 5.6602,
      "mean_token_accuracy": 0.19518781155347825,
      "num_tokens": 70333550.0,
      "step": 27755
    },
    {
      "entropy": 6.1123401641845705,
      "epoch": 3.2036930178880554,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004000487595606496,
      "loss": 5.6059,
      "mean_token_accuracy": 0.20316351652145387,
      "num_tokens": 70346417.0,
      "step": 27760
    },
    {
      "entropy": 6.173285579681396,
      "epoch": 3.2042700519330642,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004000144529751119,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19262183010578154,
      "num_tokens": 70359820.0,
      "step": 27765
    },
    {
      "entropy": 6.0782546043396,
      "epoch": 3.2048470859780727,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003999801421847753,
      "loss": 5.7313,
      "mean_token_accuracy": 0.19237934350967406,
      "num_tokens": 70373835.0,
      "step": 27770
    },
    {
      "entropy": 6.135693025588989,
      "epoch": 3.2054241200230815,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039994582719079364,
      "loss": 5.655,
      "mean_token_accuracy": 0.19483501762151717,
      "num_tokens": 70387162.0,
      "step": 27775
    },
    {
      "entropy": 6.16173529624939,
      "epoch": 3.20600115406809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039991150799432117,
      "loss": 5.7053,
      "mean_token_accuracy": 0.19465212970972062,
      "num_tokens": 70399443.0,
      "step": 27780
    },
    {
      "entropy": 6.137966632843018,
      "epoch": 3.206578188113099,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003998771845965122,
      "loss": 5.6474,
      "mean_token_accuracy": 0.1973962590098381,
      "num_tokens": 70411256.0,
      "step": 27785
    },
    {
      "entropy": 6.067451763153076,
      "epoch": 3.207155222158107,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00039984285699852116,
      "loss": 5.6634,
      "mean_token_accuracy": 0.1996780127286911,
      "num_tokens": 70424327.0,
      "step": 27790
    },
    {
      "entropy": 6.12373161315918,
      "epoch": 3.207732256203116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039980852520150255,
      "loss": 5.6505,
      "mean_token_accuracy": 0.19766132086515426,
      "num_tokens": 70436869.0,
      "step": 27795
    },
    {
      "entropy": 6.118344259262085,
      "epoch": 3.2083092902481245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003997741892066113,
      "loss": 5.7234,
      "mean_token_accuracy": 0.19439528435468673,
      "num_tokens": 70449803.0,
      "step": 27800
    },
    {
      "entropy": 6.1401876449584964,
      "epoch": 3.2088863242931334,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003997398490150021,
      "loss": 5.6802,
      "mean_token_accuracy": 0.18971627503633498,
      "num_tokens": 70461822.0,
      "step": 27805
    },
    {
      "entropy": 6.131408262252807,
      "epoch": 3.209463358338142,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003997055046278301,
      "loss": 5.668,
      "mean_token_accuracy": 0.19413064569234847,
      "num_tokens": 70473906.0,
      "step": 27810
    },
    {
      "entropy": 6.034913158416748,
      "epoch": 3.2100403923831506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003996711560462503,
      "loss": 5.5759,
      "mean_token_accuracy": 0.20933042466640472,
      "num_tokens": 70486915.0,
      "step": 27815
    },
    {
      "entropy": 6.140358257293701,
      "epoch": 3.210617426428159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003996368032714182,
      "loss": 5.6901,
      "mean_token_accuracy": 0.20078474283218384,
      "num_tokens": 70500180.0,
      "step": 27820
    },
    {
      "entropy": 6.098399353027344,
      "epoch": 3.211194460473168,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003996024463044891,
      "loss": 5.6267,
      "mean_token_accuracy": 0.1992820158600807,
      "num_tokens": 70512785.0,
      "step": 27825
    },
    {
      "entropy": 6.169175863265991,
      "epoch": 3.2117714945181763,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003995680851466186,
      "loss": 5.7214,
      "mean_token_accuracy": 0.19246959686279297,
      "num_tokens": 70525187.0,
      "step": 27830
    },
    {
      "entropy": 6.138226556777954,
      "epoch": 3.212348528563185,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003995337197989624,
      "loss": 5.6955,
      "mean_token_accuracy": 0.18921154737472534,
      "num_tokens": 70538252.0,
      "step": 27835
    },
    {
      "entropy": 6.074362802505493,
      "epoch": 3.212925562608194,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039949935026267644,
      "loss": 5.6821,
      "mean_token_accuracy": 0.19398800879716874,
      "num_tokens": 70549826.0,
      "step": 27840
    },
    {
      "entropy": 6.203020668029785,
      "epoch": 3.2135025966532025,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039946497653891666,
      "loss": 5.6651,
      "mean_token_accuracy": 0.20169676840305328,
      "num_tokens": 70561580.0,
      "step": 27845
    },
    {
      "entropy": 6.205409336090088,
      "epoch": 3.2140796306982113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039943059862883925,
      "loss": 5.8202,
      "mean_token_accuracy": 0.18746827840805053,
      "num_tokens": 70575246.0,
      "step": 27850
    },
    {
      "entropy": 6.201395845413208,
      "epoch": 3.2146566647432198,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039939621653360045,
      "loss": 5.823,
      "mean_token_accuracy": 0.18765945732593536,
      "num_tokens": 70588457.0,
      "step": 27855
    },
    {
      "entropy": 6.170694875717163,
      "epoch": 3.2152336987882286,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003993618302543566,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19841423481702805,
      "num_tokens": 70599772.0,
      "step": 27860
    },
    {
      "entropy": 6.079241561889648,
      "epoch": 3.215810732833237,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003993274397922645,
      "loss": 5.6318,
      "mean_token_accuracy": 0.19536245316267015,
      "num_tokens": 70611901.0,
      "step": 27865
    },
    {
      "entropy": 6.0728436470031735,
      "epoch": 3.216387766878246,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003992930451484807,
      "loss": 5.6484,
      "mean_token_accuracy": 0.19923944324254989,
      "num_tokens": 70625613.0,
      "step": 27870
    },
    {
      "entropy": 6.20472092628479,
      "epoch": 3.2169648009232543,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000399258646324162,
      "loss": 5.8218,
      "mean_token_accuracy": 0.19123612642288207,
      "num_tokens": 70639286.0,
      "step": 27875
    },
    {
      "entropy": 6.172567749023438,
      "epoch": 3.217541834968263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003992242433204655,
      "loss": 5.7003,
      "mean_token_accuracy": 0.19108872562646867,
      "num_tokens": 70651021.0,
      "step": 27880
    },
    {
      "entropy": 6.180254888534546,
      "epoch": 3.2181188690132716,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039918983613854825,
      "loss": 5.6625,
      "mean_token_accuracy": 0.19786344468593597,
      "num_tokens": 70664417.0,
      "step": 27885
    },
    {
      "entropy": 6.132473468780518,
      "epoch": 3.2186959030582805,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003991554247795676,
      "loss": 5.6675,
      "mean_token_accuracy": 0.1980031371116638,
      "num_tokens": 70677517.0,
      "step": 27890
    },
    {
      "entropy": 6.099011039733886,
      "epoch": 3.2192729371032893,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003991210092446808,
      "loss": 5.6786,
      "mean_token_accuracy": 0.1918819233775139,
      "num_tokens": 70689746.0,
      "step": 27895
    },
    {
      "entropy": 6.114844942092896,
      "epoch": 3.2198499711482977,
      "grad_norm": 1.0,
      "learning_rate": 0.00039908658953504567,
      "loss": 5.7492,
      "mean_token_accuracy": 0.1947510063648224,
      "num_tokens": 70702356.0,
      "step": 27900
    },
    {
      "entropy": 6.009182405471802,
      "epoch": 3.2204270051933066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039905216565181963,
      "loss": 5.5593,
      "mean_token_accuracy": 0.21047858148813248,
      "num_tokens": 70715147.0,
      "step": 27905
    },
    {
      "entropy": 6.2126298427581785,
      "epoch": 3.221004039238315,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003990177375961607,
      "loss": 5.8137,
      "mean_token_accuracy": 0.1922590434551239,
      "num_tokens": 70727951.0,
      "step": 27910
    },
    {
      "entropy": 6.102776527404785,
      "epoch": 3.221581073283324,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039898330536922665,
      "loss": 5.6389,
      "mean_token_accuracy": 0.1978570118546486,
      "num_tokens": 70740823.0,
      "step": 27915
    },
    {
      "entropy": 6.147250699996948,
      "epoch": 3.2221581073283323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003989488689721758,
      "loss": 5.6846,
      "mean_token_accuracy": 0.19394493103027344,
      "num_tokens": 70753323.0,
      "step": 27920
    },
    {
      "entropy": 6.1066591262817385,
      "epoch": 3.222735141373341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003989144284061662,
      "loss": 5.6414,
      "mean_token_accuracy": 0.1983753561973572,
      "num_tokens": 70765546.0,
      "step": 27925
    },
    {
      "entropy": 6.052069234848022,
      "epoch": 3.2233121754183496,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003988799836723565,
      "loss": 5.6809,
      "mean_token_accuracy": 0.1958295002579689,
      "num_tokens": 70778869.0,
      "step": 27930
    },
    {
      "entropy": 6.12645788192749,
      "epoch": 3.2238892094633584,
      "grad_norm": 0.9375,
      "learning_rate": 0.000398845534771905,
      "loss": 5.6483,
      "mean_token_accuracy": 0.20348383039236068,
      "num_tokens": 70792227.0,
      "step": 27935
    },
    {
      "entropy": 6.229370880126953,
      "epoch": 3.224466243508367,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003988110817059705,
      "loss": 5.7592,
      "mean_token_accuracy": 0.19297787696123123,
      "num_tokens": 70804979.0,
      "step": 27940
    },
    {
      "entropy": 6.034563064575195,
      "epoch": 3.2250432775533757,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003987766244757117,
      "loss": 5.6722,
      "mean_token_accuracy": 0.20018650889396666,
      "num_tokens": 70817538.0,
      "step": 27945
    },
    {
      "entropy": 6.162887239456177,
      "epoch": 3.225620311598384,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039874216308228764,
      "loss": 5.7196,
      "mean_token_accuracy": 0.19467681795358657,
      "num_tokens": 70830186.0,
      "step": 27950
    },
    {
      "entropy": 6.132397127151489,
      "epoch": 3.226197345643393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039870769752685744,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19618593752384186,
      "num_tokens": 70841367.0,
      "step": 27955
    },
    {
      "entropy": 6.126722717285157,
      "epoch": 3.2267743796884014,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039867322781058024,
      "loss": 5.6594,
      "mean_token_accuracy": 0.197601780295372,
      "num_tokens": 70853630.0,
      "step": 27960
    },
    {
      "entropy": 6.209865760803223,
      "epoch": 3.2273514137334103,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039863875393461546,
      "loss": 5.7935,
      "mean_token_accuracy": 0.19246441423892974,
      "num_tokens": 70866966.0,
      "step": 27965
    },
    {
      "entropy": 6.1236084461212155,
      "epoch": 3.227928447778419,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003986042759001226,
      "loss": 5.6854,
      "mean_token_accuracy": 0.1907464936375618,
      "num_tokens": 70879287.0,
      "step": 27970
    },
    {
      "entropy": 6.095495986938476,
      "epoch": 3.2285054818234276,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003985697937082613,
      "loss": 5.6577,
      "mean_token_accuracy": 0.19225332736968995,
      "num_tokens": 70892749.0,
      "step": 27975
    },
    {
      "entropy": 6.150600481033325,
      "epoch": 3.2290825158684364,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039853530736019143,
      "loss": 5.649,
      "mean_token_accuracy": 0.19959256947040557,
      "num_tokens": 70905463.0,
      "step": 27980
    },
    {
      "entropy": 6.116052484512329,
      "epoch": 3.229659549913445,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003985008168570728,
      "loss": 5.6695,
      "mean_token_accuracy": 0.1962723046541214,
      "num_tokens": 70919630.0,
      "step": 27985
    },
    {
      "entropy": 6.0763952255249025,
      "epoch": 3.2302365839584537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003984663222000656,
      "loss": 5.62,
      "mean_token_accuracy": 0.19935242235660552,
      "num_tokens": 70931810.0,
      "step": 27990
    },
    {
      "entropy": 6.113321256637573,
      "epoch": 3.230813618003462,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039843182339032997,
      "loss": 5.6994,
      "mean_token_accuracy": 0.19373810887336732,
      "num_tokens": 70945003.0,
      "step": 27995
    },
    {
      "entropy": 6.123551750183106,
      "epoch": 3.231390652048471,
      "grad_norm": 1.0,
      "learning_rate": 0.0003983973204290263,
      "loss": 5.6773,
      "mean_token_accuracy": 0.20523401945829392,
      "num_tokens": 70958889.0,
      "step": 28000
    },
    {
      "entropy": 6.0941650390625,
      "epoch": 3.2319676860934794,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039836281331731495,
      "loss": 5.6101,
      "mean_token_accuracy": 0.19690290987491607,
      "num_tokens": 70972003.0,
      "step": 28005
    },
    {
      "entropy": 6.223172950744629,
      "epoch": 3.2325447201384883,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039832830205635675,
      "loss": 5.7321,
      "mean_token_accuracy": 0.18877068907022476,
      "num_tokens": 70984269.0,
      "step": 28010
    },
    {
      "entropy": 6.09482159614563,
      "epoch": 3.2331217541834967,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00039829378664731226,
      "loss": 5.6818,
      "mean_token_accuracy": 0.19647594690322875,
      "num_tokens": 70998602.0,
      "step": 28015
    },
    {
      "entropy": 6.098766469955445,
      "epoch": 3.2336987882285055,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039825926709134257,
      "loss": 5.6861,
      "mean_token_accuracy": 0.19949979484081268,
      "num_tokens": 71011095.0,
      "step": 28020
    },
    {
      "entropy": 6.190096616744995,
      "epoch": 3.234275822273514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003982247433896087,
      "loss": 5.7123,
      "mean_token_accuracy": 0.19168585687875747,
      "num_tokens": 71022610.0,
      "step": 28025
    },
    {
      "entropy": 6.168711805343628,
      "epoch": 3.234852856318523,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039819021554327174,
      "loss": 5.7533,
      "mean_token_accuracy": 0.1923417791724205,
      "num_tokens": 71035896.0,
      "step": 28030
    },
    {
      "entropy": 6.13963623046875,
      "epoch": 3.2354298903635312,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003981556835534931,
      "loss": 5.7043,
      "mean_token_accuracy": 0.19601708501577378,
      "num_tokens": 71047793.0,
      "step": 28035
    },
    {
      "entropy": 6.197778987884521,
      "epoch": 3.23600692440854,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003981211474214342,
      "loss": 5.7238,
      "mean_token_accuracy": 0.19796839058399202,
      "num_tokens": 71060481.0,
      "step": 28040
    },
    {
      "entropy": 6.168506240844726,
      "epoch": 3.236583958453549,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003980866071482566,
      "loss": 5.7084,
      "mean_token_accuracy": 0.197569739818573,
      "num_tokens": 71073757.0,
      "step": 28045
    },
    {
      "entropy": 6.103160762786866,
      "epoch": 3.2371609924985574,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003980520627351222,
      "loss": 5.7101,
      "mean_token_accuracy": 0.2021234005689621,
      "num_tokens": 71087158.0,
      "step": 28050
    },
    {
      "entropy": 6.153443288803101,
      "epoch": 3.2377380265435662,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003980175141831928,
      "loss": 5.7309,
      "mean_token_accuracy": 0.19064217805862427,
      "num_tokens": 71099242.0,
      "step": 28055
    },
    {
      "entropy": 6.159456396102906,
      "epoch": 3.2383150605885747,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039798296149363033,
      "loss": 5.6525,
      "mean_token_accuracy": 0.19336422085762023,
      "num_tokens": 71110971.0,
      "step": 28060
    },
    {
      "entropy": 6.11139554977417,
      "epoch": 3.2388920946335835,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039794840466759714,
      "loss": 5.7469,
      "mean_token_accuracy": 0.19323474466800689,
      "num_tokens": 71125024.0,
      "step": 28065
    },
    {
      "entropy": 6.115737581253052,
      "epoch": 3.239469128678592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039791384370625537,
      "loss": 5.5826,
      "mean_token_accuracy": 0.19599466621875763,
      "num_tokens": 71138030.0,
      "step": 28070
    },
    {
      "entropy": 6.173791599273682,
      "epoch": 3.240046162723601,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003978792786107675,
      "loss": 5.7519,
      "mean_token_accuracy": 0.19325887262821198,
      "num_tokens": 71150763.0,
      "step": 28075
    },
    {
      "entropy": 6.122519063949585,
      "epoch": 3.240623196768609,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039784470938229624,
      "loss": 5.6262,
      "mean_token_accuracy": 0.20141390562057496,
      "num_tokens": 71163235.0,
      "step": 28080
    },
    {
      "entropy": 6.023569345474243,
      "epoch": 3.241200230813618,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039781013602200406,
      "loss": 5.645,
      "mean_token_accuracy": 0.19679532200098038,
      "num_tokens": 71176119.0,
      "step": 28085
    },
    {
      "entropy": 6.059888696670532,
      "epoch": 3.2417772648586265,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003977755585310541,
      "loss": 5.6097,
      "mean_token_accuracy": 0.20407044589519502,
      "num_tokens": 71190919.0,
      "step": 28090
    },
    {
      "entropy": 6.191178321838379,
      "epoch": 3.2423542989036354,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003977409769106091,
      "loss": 5.7317,
      "mean_token_accuracy": 0.1891528382897377,
      "num_tokens": 71202589.0,
      "step": 28095
    },
    {
      "entropy": 6.096343898773194,
      "epoch": 3.2429313329486438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003977063911618323,
      "loss": 5.666,
      "mean_token_accuracy": 0.20136779993772508,
      "num_tokens": 71214632.0,
      "step": 28100
    },
    {
      "entropy": 6.0593907833099365,
      "epoch": 3.2435083669936526,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003976718012858871,
      "loss": 5.6188,
      "mean_token_accuracy": 0.20363556444644929,
      "num_tokens": 71226864.0,
      "step": 28105
    },
    {
      "entropy": 6.1296711444854735,
      "epoch": 3.244085401038661,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003976372072839367,
      "loss": 5.6825,
      "mean_token_accuracy": 0.19185557216405869,
      "num_tokens": 71239254.0,
      "step": 28110
    },
    {
      "entropy": 6.117106533050537,
      "epoch": 3.24466243508367,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003976026091571448,
      "loss": 5.7227,
      "mean_token_accuracy": 0.19308591037988662,
      "num_tokens": 71252215.0,
      "step": 28115
    },
    {
      "entropy": 6.119906997680664,
      "epoch": 3.2452394691286788,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039756800690667505,
      "loss": 5.6848,
      "mean_token_accuracy": 0.20117004513740538,
      "num_tokens": 71263733.0,
      "step": 28120
    },
    {
      "entropy": 6.094219303131103,
      "epoch": 3.245816503173687,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039753340053369116,
      "loss": 5.7188,
      "mean_token_accuracy": 0.19860861897468568,
      "num_tokens": 71277556.0,
      "step": 28125
    },
    {
      "entropy": 6.189479875564575,
      "epoch": 3.246393537218696,
      "grad_norm": 0.921875,
      "learning_rate": 0.00039749879003935743,
      "loss": 5.7394,
      "mean_token_accuracy": 0.19267741441726685,
      "num_tokens": 71289942.0,
      "step": 28130
    },
    {
      "entropy": 6.263569784164429,
      "epoch": 3.2469705712637045,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003974641754248375,
      "loss": 5.8222,
      "mean_token_accuracy": 0.18577689081430435,
      "num_tokens": 71302517.0,
      "step": 28135
    },
    {
      "entropy": 6.0751505374908445,
      "epoch": 3.2475476053087133,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039742955669129607,
      "loss": 5.6913,
      "mean_token_accuracy": 0.20028978139162062,
      "num_tokens": 71315022.0,
      "step": 28140
    },
    {
      "entropy": 6.042082166671753,
      "epoch": 3.2481246393537218,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00039739493383989714,
      "loss": 5.5883,
      "mean_token_accuracy": 0.19971897155046464,
      "num_tokens": 71329335.0,
      "step": 28145
    },
    {
      "entropy": 6.125881338119507,
      "epoch": 3.2487016733987306,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003973603068718055,
      "loss": 5.7168,
      "mean_token_accuracy": 0.19682153612375258,
      "num_tokens": 71342739.0,
      "step": 28150
    },
    {
      "entropy": 6.094817638397217,
      "epoch": 3.249278707443739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039732567578818574,
      "loss": 5.621,
      "mean_token_accuracy": 0.19936226904392243,
      "num_tokens": 71355367.0,
      "step": 28155
    },
    {
      "entropy": 6.131360960006714,
      "epoch": 3.249855741488748,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003972910405902026,
      "loss": 5.6528,
      "mean_token_accuracy": 0.19836146831512452,
      "num_tokens": 71367163.0,
      "step": 28160
    },
    {
      "entropy": 6.0587163925170895,
      "epoch": 3.2504327755337563,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003972564012790211,
      "loss": 5.6966,
      "mean_token_accuracy": 0.19916419237852095,
      "num_tokens": 71379424.0,
      "step": 28165
    },
    {
      "entropy": 6.183052682876587,
      "epoch": 3.251009809578765,
      "grad_norm": 0.890625,
      "learning_rate": 0.00039722175785580617,
      "loss": 5.7551,
      "mean_token_accuracy": 0.19467741698026658,
      "num_tokens": 71392115.0,
      "step": 28170
    },
    {
      "entropy": 6.117837429046631,
      "epoch": 3.251586843623774,
      "grad_norm": 0.875,
      "learning_rate": 0.0003971871103217232,
      "loss": 5.6812,
      "mean_token_accuracy": 0.1975853368639946,
      "num_tokens": 71404339.0,
      "step": 28175
    },
    {
      "entropy": 6.112961292266846,
      "epoch": 3.2521638776687825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039715245867793744,
      "loss": 5.6886,
      "mean_token_accuracy": 0.19479491263628007,
      "num_tokens": 71416331.0,
      "step": 28180
    },
    {
      "entropy": 6.116593885421753,
      "epoch": 3.252740911713791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003971178029256144,
      "loss": 5.6487,
      "mean_token_accuracy": 0.20387978702783585,
      "num_tokens": 71428511.0,
      "step": 28185
    },
    {
      "entropy": 6.136179065704345,
      "epoch": 3.2533179457587997,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039708314306591974,
      "loss": 5.6816,
      "mean_token_accuracy": 0.19546741545200347,
      "num_tokens": 71440708.0,
      "step": 28190
    },
    {
      "entropy": 6.0933012008667,
      "epoch": 3.2538949798038086,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039704847910001926,
      "loss": 5.6567,
      "mean_token_accuracy": 0.1944028839468956,
      "num_tokens": 71453363.0,
      "step": 28195
    },
    {
      "entropy": 6.094995832443237,
      "epoch": 3.254472013848817,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003970138110290787,
      "loss": 5.5986,
      "mean_token_accuracy": 0.20314981043338776,
      "num_tokens": 71465390.0,
      "step": 28200
    },
    {
      "entropy": 6.185604000091553,
      "epoch": 3.255049047893826,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00039697913885426424,
      "loss": 5.7902,
      "mean_token_accuracy": 0.1876772627234459,
      "num_tokens": 71478148.0,
      "step": 28205
    },
    {
      "entropy": 6.086142349243164,
      "epoch": 3.2556260819388343,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039694446257674203,
      "loss": 5.6375,
      "mean_token_accuracy": 0.2032093420624733,
      "num_tokens": 71490604.0,
      "step": 28210
    },
    {
      "entropy": 6.104994678497315,
      "epoch": 3.256203115983843,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039690978219767846,
      "loss": 5.7821,
      "mean_token_accuracy": 0.18981408774852754,
      "num_tokens": 71503379.0,
      "step": 28215
    },
    {
      "entropy": 6.226002311706543,
      "epoch": 3.2567801500288516,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003968750977182399,
      "loss": 5.6973,
      "mean_token_accuracy": 0.19314495623111724,
      "num_tokens": 71516350.0,
      "step": 28220
    },
    {
      "entropy": 6.1495274066925045,
      "epoch": 3.2573571840738604,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039684040913959295,
      "loss": 5.6489,
      "mean_token_accuracy": 0.19741449356079102,
      "num_tokens": 71528314.0,
      "step": 28225
    },
    {
      "entropy": 6.081626605987549,
      "epoch": 3.257934218118869,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003968057164629043,
      "loss": 5.6723,
      "mean_token_accuracy": 0.20468200892210006,
      "num_tokens": 71540987.0,
      "step": 28230
    },
    {
      "entropy": 6.063146162033081,
      "epoch": 3.2585112521638777,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000396771019689341,
      "loss": 5.5989,
      "mean_token_accuracy": 0.20137522667646407,
      "num_tokens": 71553605.0,
      "step": 28235
    },
    {
      "entropy": 6.126628971099853,
      "epoch": 3.259088286208886,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039673631882006986,
      "loss": 5.6954,
      "mean_token_accuracy": 0.19189148843288423,
      "num_tokens": 71565744.0,
      "step": 28240
    },
    {
      "entropy": 6.217823505401611,
      "epoch": 3.259665320253895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039670161385625815,
      "loss": 5.7657,
      "mean_token_accuracy": 0.19321134686470032,
      "num_tokens": 71577756.0,
      "step": 28245
    },
    {
      "entropy": 6.114751195907592,
      "epoch": 3.260242354298904,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039666690479907307,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19664305299520493,
      "num_tokens": 71590814.0,
      "step": 28250
    },
    {
      "entropy": 6.156188917160034,
      "epoch": 3.2608193883439123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039663219164968213,
      "loss": 5.7002,
      "mean_token_accuracy": 0.19467169493436814,
      "num_tokens": 71602960.0,
      "step": 28255
    },
    {
      "entropy": 6.231931829452515,
      "epoch": 3.261396422388921,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039659747440925277,
      "loss": 5.823,
      "mean_token_accuracy": 0.18830994069576262,
      "num_tokens": 71614568.0,
      "step": 28260
    },
    {
      "entropy": 6.139056444168091,
      "epoch": 3.2619734564339296,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039656275307895286,
      "loss": 5.7385,
      "mean_token_accuracy": 0.19024017453193665,
      "num_tokens": 71628103.0,
      "step": 28265
    },
    {
      "entropy": 6.047552013397217,
      "epoch": 3.2625504904789384,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039652802765995006,
      "loss": 5.5823,
      "mean_token_accuracy": 0.20624837428331375,
      "num_tokens": 71640343.0,
      "step": 28270
    },
    {
      "entropy": 6.103335952758789,
      "epoch": 3.263127524523947,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039649329815341243,
      "loss": 5.6129,
      "mean_token_accuracy": 0.1963995724916458,
      "num_tokens": 71653430.0,
      "step": 28275
    },
    {
      "entropy": 6.075761604309082,
      "epoch": 3.2637045585689557,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039645856456050813,
      "loss": 5.6257,
      "mean_token_accuracy": 0.20027599930763246,
      "num_tokens": 71665635.0,
      "step": 28280
    },
    {
      "entropy": 5.953153276443482,
      "epoch": 3.264281592613964,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003964238268824052,
      "loss": 5.546,
      "mean_token_accuracy": 0.2134902521967888,
      "num_tokens": 71679190.0,
      "step": 28285
    },
    {
      "entropy": 6.103391075134278,
      "epoch": 3.264858626658973,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0003963890851202723,
      "loss": 5.6743,
      "mean_token_accuracy": 0.19676871001720428,
      "num_tokens": 71693938.0,
      "step": 28290
    },
    {
      "entropy": 6.125302791595459,
      "epoch": 3.2654356607039814,
      "grad_norm": 0.875,
      "learning_rate": 0.00039635433927527776,
      "loss": 5.6028,
      "mean_token_accuracy": 0.2024545818567276,
      "num_tokens": 71706682.0,
      "step": 28295
    },
    {
      "entropy": 6.19836688041687,
      "epoch": 3.2660126947489903,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039631958934859023,
      "loss": 5.7153,
      "mean_token_accuracy": 0.19123188853263856,
      "num_tokens": 71718454.0,
      "step": 28300
    },
    {
      "entropy": 6.154357051849365,
      "epoch": 3.2665897287939987,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039628483534137865,
      "loss": 5.7262,
      "mean_token_accuracy": 0.20406693667173387,
      "num_tokens": 71731270.0,
      "step": 28305
    },
    {
      "entropy": 6.151256656646728,
      "epoch": 3.2671667628390075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039625007725481193,
      "loss": 5.7089,
      "mean_token_accuracy": 0.19369070678949357,
      "num_tokens": 71744086.0,
      "step": 28310
    },
    {
      "entropy": 6.13329176902771,
      "epoch": 3.267743796884016,
      "grad_norm": 0.984375,
      "learning_rate": 0.000396215315090059,
      "loss": 5.6796,
      "mean_token_accuracy": 0.19493555426597595,
      "num_tokens": 71756496.0,
      "step": 28315
    },
    {
      "entropy": 6.116254901885986,
      "epoch": 3.268320830929025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039618054884828924,
      "loss": 5.6604,
      "mean_token_accuracy": 0.20112556219100952,
      "num_tokens": 71768939.0,
      "step": 28320
    },
    {
      "entropy": 6.09043402671814,
      "epoch": 3.2688978649740337,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003961457785306719,
      "loss": 5.6318,
      "mean_token_accuracy": 0.1919059321284294,
      "num_tokens": 71780555.0,
      "step": 28325
    },
    {
      "entropy": 6.1364072322845455,
      "epoch": 3.269474899019042,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003961110041383764,
      "loss": 5.7971,
      "mean_token_accuracy": 0.19151452034711838,
      "num_tokens": 71795705.0,
      "step": 28330
    },
    {
      "entropy": 6.17677755355835,
      "epoch": 3.270051933064051,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003960762256725725,
      "loss": 5.6725,
      "mean_token_accuracy": 0.20107867270708085,
      "num_tokens": 71807582.0,
      "step": 28335
    },
    {
      "entropy": 6.189778184890747,
      "epoch": 3.2706289671090594,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039604144313442984,
      "loss": 5.7479,
      "mean_token_accuracy": 0.19078557342290878,
      "num_tokens": 71819349.0,
      "step": 28340
    },
    {
      "entropy": 6.154573726654053,
      "epoch": 3.2712060011540682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039600665652511836,
      "loss": 5.6923,
      "mean_token_accuracy": 0.20255973190069199,
      "num_tokens": 71832443.0,
      "step": 28345
    },
    {
      "entropy": 6.089715194702149,
      "epoch": 3.2717830351990767,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039597186584580814,
      "loss": 5.709,
      "mean_token_accuracy": 0.18834333270788192,
      "num_tokens": 71844961.0,
      "step": 28350
    },
    {
      "entropy": 6.155925941467285,
      "epoch": 3.2723600692440855,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003959370710976693,
      "loss": 5.7375,
      "mean_token_accuracy": 0.1984020322561264,
      "num_tokens": 71858158.0,
      "step": 28355
    },
    {
      "entropy": 6.153133201599121,
      "epoch": 3.272937103289094,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039590227228187213,
      "loss": 5.7066,
      "mean_token_accuracy": 0.20005539804697037,
      "num_tokens": 71870985.0,
      "step": 28360
    },
    {
      "entropy": 6.137960767745971,
      "epoch": 3.273514137334103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003958674693995871,
      "loss": 5.6939,
      "mean_token_accuracy": 0.19767895340919495,
      "num_tokens": 71883687.0,
      "step": 28365
    },
    {
      "entropy": 6.143102550506592,
      "epoch": 3.274091171379111,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003958326624519848,
      "loss": 5.7571,
      "mean_token_accuracy": 0.1857014700770378,
      "num_tokens": 71896678.0,
      "step": 28370
    },
    {
      "entropy": 6.108162927627563,
      "epoch": 3.27466820542412,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039579785144023595,
      "loss": 5.6663,
      "mean_token_accuracy": 0.20057436674833298,
      "num_tokens": 71910722.0,
      "step": 28375
    },
    {
      "entropy": 6.125950622558594,
      "epoch": 3.2752452394691285,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003957630363655113,
      "loss": 5.6714,
      "mean_token_accuracy": 0.19949873685836791,
      "num_tokens": 71922849.0,
      "step": 28380
    },
    {
      "entropy": 6.113627624511719,
      "epoch": 3.2758222735141374,
      "grad_norm": 1.0,
      "learning_rate": 0.00039572821722898185,
      "loss": 5.6253,
      "mean_token_accuracy": 0.19689620286226273,
      "num_tokens": 71937106.0,
      "step": 28385
    },
    {
      "entropy": 6.176388359069824,
      "epoch": 3.2763993075591458,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003956933940318189,
      "loss": 5.7059,
      "mean_token_accuracy": 0.19192722588777542,
      "num_tokens": 71949384.0,
      "step": 28390
    },
    {
      "entropy": 6.077463960647583,
      "epoch": 3.2769763416041546,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003956585667751935,
      "loss": 5.6517,
      "mean_token_accuracy": 0.19595678448677062,
      "num_tokens": 71961906.0,
      "step": 28395
    },
    {
      "entropy": 6.072017621994019,
      "epoch": 3.2775533756491635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039562373546027726,
      "loss": 5.6179,
      "mean_token_accuracy": 0.2000121220946312,
      "num_tokens": 71974684.0,
      "step": 28400
    },
    {
      "entropy": 6.1156532764434814,
      "epoch": 3.278130409694172,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003955889000882415,
      "loss": 5.6693,
      "mean_token_accuracy": 0.18992233872413636,
      "num_tokens": 71987410.0,
      "step": 28405
    },
    {
      "entropy": 6.11706428527832,
      "epoch": 3.2787074437391808,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039555406066025796,
      "loss": 5.6911,
      "mean_token_accuracy": 0.19453849792480468,
      "num_tokens": 72000155.0,
      "step": 28410
    },
    {
      "entropy": 6.160224676132202,
      "epoch": 3.279284477784189,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003955192171774986,
      "loss": 5.7166,
      "mean_token_accuracy": 0.19104048758745193,
      "num_tokens": 72013057.0,
      "step": 28415
    },
    {
      "entropy": 6.11291537284851,
      "epoch": 3.279861511829198,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003954843696411351,
      "loss": 5.653,
      "mean_token_accuracy": 0.18929742276668549,
      "num_tokens": 72024875.0,
      "step": 28420
    },
    {
      "entropy": 6.13860855102539,
      "epoch": 3.2804385458742065,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003954495180523397,
      "loss": 5.6667,
      "mean_token_accuracy": 0.19640615284442903,
      "num_tokens": 72036932.0,
      "step": 28425
    },
    {
      "entropy": 6.107240104675293,
      "epoch": 3.2810155799192153,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039541466241228466,
      "loss": 5.6867,
      "mean_token_accuracy": 0.19907524287700654,
      "num_tokens": 72049081.0,
      "step": 28430
    },
    {
      "entropy": 6.144339179992675,
      "epoch": 3.2815926139642237,
      "grad_norm": 2.96875,
      "learning_rate": 0.00039537980272214224,
      "loss": 5.6487,
      "mean_token_accuracy": 0.19673170894384384,
      "num_tokens": 72061334.0,
      "step": 28435
    },
    {
      "entropy": 6.102508497238159,
      "epoch": 3.2821696480092326,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003953449389830849,
      "loss": 5.6383,
      "mean_token_accuracy": 0.19665226042270662,
      "num_tokens": 72073446.0,
      "step": 28440
    },
    {
      "entropy": 6.094725513458252,
      "epoch": 3.282746682054241,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003953100711962853,
      "loss": 5.6225,
      "mean_token_accuracy": 0.2013086885213852,
      "num_tokens": 72086219.0,
      "step": 28445
    },
    {
      "entropy": 6.125956106185913,
      "epoch": 3.28332371609925,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039527519936291626,
      "loss": 5.7183,
      "mean_token_accuracy": 0.19325001984834672,
      "num_tokens": 72098442.0,
      "step": 28450
    },
    {
      "entropy": 6.141130638122559,
      "epoch": 3.2839007501442588,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039524032348415075,
      "loss": 5.6924,
      "mean_token_accuracy": 0.1980439692735672,
      "num_tokens": 72111416.0,
      "step": 28455
    },
    {
      "entropy": 6.068466281890869,
      "epoch": 3.284477784189267,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003952054435611615,
      "loss": 5.6284,
      "mean_token_accuracy": 0.20466866195201874,
      "num_tokens": 72124077.0,
      "step": 28460
    },
    {
      "entropy": 6.169934177398682,
      "epoch": 3.2850548182342756,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039517055959512195,
      "loss": 5.7232,
      "mean_token_accuracy": 0.19903453290462494,
      "num_tokens": 72136170.0,
      "step": 28465
    },
    {
      "entropy": 6.065824031829834,
      "epoch": 3.2856318522792844,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003951356715872053,
      "loss": 5.6374,
      "mean_token_accuracy": 0.20355813056230546,
      "num_tokens": 72148868.0,
      "step": 28470
    },
    {
      "entropy": 6.137563705444336,
      "epoch": 3.2862088863242933,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000395100779538585,
      "loss": 5.6672,
      "mean_token_accuracy": 0.19911954551935196,
      "num_tokens": 72162225.0,
      "step": 28475
    },
    {
      "entropy": 6.1495139598846436,
      "epoch": 3.2867859203693017,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003950658834504347,
      "loss": 5.7365,
      "mean_token_accuracy": 0.192801333963871,
      "num_tokens": 72174424.0,
      "step": 28480
    },
    {
      "entropy": 6.184037065505981,
      "epoch": 3.2873629544143106,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000395030983323928,
      "loss": 5.6659,
      "mean_token_accuracy": 0.19671981036663055,
      "num_tokens": 72187672.0,
      "step": 28485
    },
    {
      "entropy": 6.036303281784058,
      "epoch": 3.287939988459319,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039499607916023885,
      "loss": 5.5365,
      "mean_token_accuracy": 0.20413777679204942,
      "num_tokens": 72201575.0,
      "step": 28490
    },
    {
      "entropy": 6.018047094345093,
      "epoch": 3.288517022504328,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003949611709605411,
      "loss": 5.5415,
      "mean_token_accuracy": 0.2080550193786621,
      "num_tokens": 72215035.0,
      "step": 28495
    },
    {
      "entropy": 6.184219074249268,
      "epoch": 3.2890940565493363,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000394926258726009,
      "loss": 5.7228,
      "mean_token_accuracy": 0.19657255858182907,
      "num_tokens": 72227152.0,
      "step": 28500
    },
    {
      "entropy": 6.106447076797485,
      "epoch": 3.289671090594345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003948913424578168,
      "loss": 5.6705,
      "mean_token_accuracy": 0.19262617826461792,
      "num_tokens": 72240770.0,
      "step": 28505
    },
    {
      "entropy": 6.14880256652832,
      "epoch": 3.2902481246393536,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003948564221571388,
      "loss": 5.6493,
      "mean_token_accuracy": 0.20426878482103347,
      "num_tokens": 72253254.0,
      "step": 28510
    },
    {
      "entropy": 6.07218565940857,
      "epoch": 3.2908251586843624,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039482149782514955,
      "loss": 5.6529,
      "mean_token_accuracy": 0.20243048369884492,
      "num_tokens": 72267863.0,
      "step": 28515
    },
    {
      "entropy": 6.1394964218139645,
      "epoch": 3.291402192729371,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003947865694630238,
      "loss": 5.6506,
      "mean_token_accuracy": 0.19797384589910508,
      "num_tokens": 72279867.0,
      "step": 28520
    },
    {
      "entropy": 6.022223663330078,
      "epoch": 3.2919792267743797,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003947516370719362,
      "loss": 5.513,
      "mean_token_accuracy": 0.21671516299247742,
      "num_tokens": 72293347.0,
      "step": 28525
    },
    {
      "entropy": 6.127876472473145,
      "epoch": 3.2925562608193886,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003947167006530618,
      "loss": 5.724,
      "mean_token_accuracy": 0.19348247945308686,
      "num_tokens": 72306216.0,
      "step": 28530
    },
    {
      "entropy": 6.149895286560058,
      "epoch": 3.293133294864397,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039468176020757573,
      "loss": 5.7082,
      "mean_token_accuracy": 0.19451941400766373,
      "num_tokens": 72318253.0,
      "step": 28535
    },
    {
      "entropy": 6.091161394119263,
      "epoch": 3.293710328909406,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000394646815736653,
      "loss": 5.697,
      "mean_token_accuracy": 0.1967163398861885,
      "num_tokens": 72331451.0,
      "step": 28540
    },
    {
      "entropy": 6.175706768035889,
      "epoch": 3.2942873629544143,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003946118672414692,
      "loss": 5.7118,
      "mean_token_accuracy": 0.19456691145896912,
      "num_tokens": 72343532.0,
      "step": 28545
    },
    {
      "entropy": 6.1448719024658205,
      "epoch": 3.294864396999423,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039457691472319953,
      "loss": 5.7237,
      "mean_token_accuracy": 0.18762239068746567,
      "num_tokens": 72355803.0,
      "step": 28550
    },
    {
      "entropy": 6.122489500045776,
      "epoch": 3.2954414310444315,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003945419581830197,
      "loss": 5.6205,
      "mean_token_accuracy": 0.19542174488306047,
      "num_tokens": 72370056.0,
      "step": 28555
    },
    {
      "entropy": 6.079394197463989,
      "epoch": 3.2960184650894404,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039450699762210556,
      "loss": 5.5515,
      "mean_token_accuracy": 0.20370566695928574,
      "num_tokens": 72382444.0,
      "step": 28560
    },
    {
      "entropy": 6.063602304458618,
      "epoch": 3.296595499134449,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039447203304163295,
      "loss": 5.6254,
      "mean_token_accuracy": 0.20646788328886032,
      "num_tokens": 72396773.0,
      "step": 28565
    },
    {
      "entropy": 6.068623018264771,
      "epoch": 3.2971725331794577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003944370644427777,
      "loss": 5.5642,
      "mean_token_accuracy": 0.20680238008499147,
      "num_tokens": 72409922.0,
      "step": 28570
    },
    {
      "entropy": 6.178087949752808,
      "epoch": 3.297749567224466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003944020918267163,
      "loss": 5.6803,
      "mean_token_accuracy": 0.20366569608449936,
      "num_tokens": 72423011.0,
      "step": 28575
    },
    {
      "entropy": 6.0516369342803955,
      "epoch": 3.298326601269475,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039436711519462474,
      "loss": 5.622,
      "mean_token_accuracy": 0.198382431268692,
      "num_tokens": 72434888.0,
      "step": 28580
    },
    {
      "entropy": 6.122730875015259,
      "epoch": 3.2989036353144834,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003943321345476796,
      "loss": 5.6493,
      "mean_token_accuracy": 0.2048882484436035,
      "num_tokens": 72447202.0,
      "step": 28585
    },
    {
      "entropy": 5.954018020629883,
      "epoch": 3.2994806693594922,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039429714988705735,
      "loss": 5.5965,
      "mean_token_accuracy": 0.20587584674358367,
      "num_tokens": 72460232.0,
      "step": 28590
    },
    {
      "entropy": 6.149437093734742,
      "epoch": 3.3000577034045007,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039426216121393477,
      "loss": 5.6619,
      "mean_token_accuracy": 0.19850059300661088,
      "num_tokens": 72472665.0,
      "step": 28595
    },
    {
      "entropy": 6.094783163070678,
      "epoch": 3.3006347374495095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003942271685294886,
      "loss": 5.6789,
      "mean_token_accuracy": 0.20599209815263747,
      "num_tokens": 72484766.0,
      "step": 28600
    },
    {
      "entropy": 6.212546396255493,
      "epoch": 3.3012117714945184,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00039419217183489594,
      "loss": 5.7195,
      "mean_token_accuracy": 0.19344098418951033,
      "num_tokens": 72497352.0,
      "step": 28605
    },
    {
      "entropy": 6.200887823104859,
      "epoch": 3.301788805539527,
      "grad_norm": 1.125,
      "learning_rate": 0.00039415717113133356,
      "loss": 5.7599,
      "mean_token_accuracy": 0.19536616653203964,
      "num_tokens": 72509057.0,
      "step": 28610
    },
    {
      "entropy": 6.200597190856934,
      "epoch": 3.3023658395845357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003941221664199791,
      "loss": 5.7398,
      "mean_token_accuracy": 0.1925618976354599,
      "num_tokens": 72521624.0,
      "step": 28615
    },
    {
      "entropy": 6.123512125015258,
      "epoch": 3.302942873629544,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00039408715770200976,
      "loss": 5.6528,
      "mean_token_accuracy": 0.2027030423283577,
      "num_tokens": 72534078.0,
      "step": 28620
    },
    {
      "entropy": 6.0805620670318605,
      "epoch": 3.303519907674553,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039405214497860295,
      "loss": 5.6461,
      "mean_token_accuracy": 0.1997436463832855,
      "num_tokens": 72547310.0,
      "step": 28625
    },
    {
      "entropy": 6.187747478485107,
      "epoch": 3.3040969417195614,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003940171282509363,
      "loss": 5.6791,
      "mean_token_accuracy": 0.19879880994558335,
      "num_tokens": 72559758.0,
      "step": 28630
    },
    {
      "entropy": 6.120146179199219,
      "epoch": 3.3046739757645702,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003939821075201878,
      "loss": 5.701,
      "mean_token_accuracy": 0.19482448399066926,
      "num_tokens": 72571365.0,
      "step": 28635
    },
    {
      "entropy": 6.19158902168274,
      "epoch": 3.3052510098095786,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003939470827875352,
      "loss": 5.7396,
      "mean_token_accuracy": 0.19038573354482652,
      "num_tokens": 72583425.0,
      "step": 28640
    },
    {
      "entropy": 6.138459205627441,
      "epoch": 3.3058280438545875,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003939120540541565,
      "loss": 5.6512,
      "mean_token_accuracy": 0.2009557455778122,
      "num_tokens": 72595743.0,
      "step": 28645
    },
    {
      "entropy": 6.148765945434571,
      "epoch": 3.306405077899596,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039387702132122993,
      "loss": 5.7282,
      "mean_token_accuracy": 0.1975986823439598,
      "num_tokens": 72608510.0,
      "step": 28650
    },
    {
      "entropy": 6.068108558654785,
      "epoch": 3.306982111944605,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039384198458993386,
      "loss": 5.589,
      "mean_token_accuracy": 0.20037676990032197,
      "num_tokens": 72620681.0,
      "step": 28655
    },
    {
      "entropy": 6.164902544021606,
      "epoch": 3.307559145989613,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039380694386144665,
      "loss": 5.7064,
      "mean_token_accuracy": 0.19251110255718232,
      "num_tokens": 72632970.0,
      "step": 28660
    },
    {
      "entropy": 6.148618888854981,
      "epoch": 3.308136180034622,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039377189913694687,
      "loss": 5.7357,
      "mean_token_accuracy": 0.19509387463331224,
      "num_tokens": 72647953.0,
      "step": 28665
    },
    {
      "entropy": 6.226953792572021,
      "epoch": 3.3087132140796305,
      "grad_norm": 1.0,
      "learning_rate": 0.00039373685041761323,
      "loss": 5.6941,
      "mean_token_accuracy": 0.19258223921060563,
      "num_tokens": 72660333.0,
      "step": 28670
    },
    {
      "entropy": 6.107355117797852,
      "epoch": 3.3092902481246393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003937017977046247,
      "loss": 5.6652,
      "mean_token_accuracy": 0.19594338089227675,
      "num_tokens": 72673620.0,
      "step": 28675
    },
    {
      "entropy": 6.143393802642822,
      "epoch": 3.309867282169648,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039366674099916007,
      "loss": 5.7698,
      "mean_token_accuracy": 0.1875929519534111,
      "num_tokens": 72686285.0,
      "step": 28680
    },
    {
      "entropy": 6.012445116043091,
      "epoch": 3.3104443162146566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003936316803023986,
      "loss": 5.4968,
      "mean_token_accuracy": 0.20871724486351012,
      "num_tokens": 72699209.0,
      "step": 28685
    },
    {
      "entropy": 6.080950927734375,
      "epoch": 3.3110213502596655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039359661561551946,
      "loss": 5.6787,
      "mean_token_accuracy": 0.19370285421609879,
      "num_tokens": 72711825.0,
      "step": 28690
    },
    {
      "entropy": 6.058858108520508,
      "epoch": 3.311598384304674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039356154693970214,
      "loss": 5.6221,
      "mean_token_accuracy": 0.19335783123970032,
      "num_tokens": 72724908.0,
      "step": 28695
    },
    {
      "entropy": 6.143716335296631,
      "epoch": 3.3121754183496828,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039352647427612597,
      "loss": 5.6474,
      "mean_token_accuracy": 0.2015886574983597,
      "num_tokens": 72738431.0,
      "step": 28700
    },
    {
      "entropy": 6.117425203323364,
      "epoch": 3.312752452394691,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003934913976259709,
      "loss": 5.6698,
      "mean_token_accuracy": 0.19382237643003464,
      "num_tokens": 72750800.0,
      "step": 28705
    },
    {
      "entropy": 6.119411468505859,
      "epoch": 3.3133294864397,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003934563169904164,
      "loss": 5.6372,
      "mean_token_accuracy": 0.20595670044422149,
      "num_tokens": 72763343.0,
      "step": 28710
    },
    {
      "entropy": 6.1254864692687985,
      "epoch": 3.3139065204847085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003934212323706425,
      "loss": 5.625,
      "mean_token_accuracy": 0.20556642413139342,
      "num_tokens": 72775865.0,
      "step": 28715
    },
    {
      "entropy": 6.046801519393921,
      "epoch": 3.3144835545297173,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003933861437678292,
      "loss": 5.6872,
      "mean_token_accuracy": 0.195333631336689,
      "num_tokens": 72789579.0,
      "step": 28720
    },
    {
      "entropy": 6.069760179519653,
      "epoch": 3.3150605885747257,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003933510511831569,
      "loss": 5.6082,
      "mean_token_accuracy": 0.20962165743112565,
      "num_tokens": 72802952.0,
      "step": 28725
    },
    {
      "entropy": 6.093776845932007,
      "epoch": 3.3156376226197346,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039331595461780574,
      "loss": 5.6516,
      "mean_token_accuracy": 0.19898153692483903,
      "num_tokens": 72816571.0,
      "step": 28730
    },
    {
      "entropy": 6.103912353515625,
      "epoch": 3.3162146566647435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039328085407295616,
      "loss": 5.5928,
      "mean_token_accuracy": 0.20080455243587494,
      "num_tokens": 72829077.0,
      "step": 28735
    },
    {
      "entropy": 6.0752137184143065,
      "epoch": 3.316791690709752,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039324574954978885,
      "loss": 5.6545,
      "mean_token_accuracy": 0.19743015021085739,
      "num_tokens": 72842530.0,
      "step": 28740
    },
    {
      "entropy": 6.1761727809906,
      "epoch": 3.3173687247547603,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039321064104948456,
      "loss": 5.731,
      "mean_token_accuracy": 0.18902297765016557,
      "num_tokens": 72854534.0,
      "step": 28745
    },
    {
      "entropy": 6.070225954055786,
      "epoch": 3.317945758799769,
      "grad_norm": 1.0,
      "learning_rate": 0.00039317552857322404,
      "loss": 5.7465,
      "mean_token_accuracy": 0.19280608296394347,
      "num_tokens": 72866860.0,
      "step": 28750
    },
    {
      "entropy": 6.2420454025268555,
      "epoch": 3.318522792844778,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039314041212218826,
      "loss": 5.7338,
      "mean_token_accuracy": 0.19116558283567428,
      "num_tokens": 72879074.0,
      "step": 28755
    },
    {
      "entropy": 6.097189903259277,
      "epoch": 3.3190998268897864,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003931052916975584,
      "loss": 5.5829,
      "mean_token_accuracy": 0.20671233534812927,
      "num_tokens": 72890981.0,
      "step": 28760
    },
    {
      "entropy": 6.075610494613647,
      "epoch": 3.3196768609347953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039307016730051576,
      "loss": 5.6852,
      "mean_token_accuracy": 0.19994308352470397,
      "num_tokens": 72903068.0,
      "step": 28765
    },
    {
      "entropy": 6.069404888153076,
      "epoch": 3.3202538949798037,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003930350389322417,
      "loss": 5.6715,
      "mean_token_accuracy": 0.19757690280675888,
      "num_tokens": 72917805.0,
      "step": 28770
    },
    {
      "entropy": 6.116404581069946,
      "epoch": 3.3208309290248126,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003929999065939177,
      "loss": 5.625,
      "mean_token_accuracy": 0.20093947947025298,
      "num_tokens": 72930665.0,
      "step": 28775
    },
    {
      "entropy": 6.147107839584351,
      "epoch": 3.321407963069821,
      "grad_norm": 0.890625,
      "learning_rate": 0.00039296477028672545,
      "loss": 5.6852,
      "mean_token_accuracy": 0.19490035623311996,
      "num_tokens": 72943841.0,
      "step": 28780
    },
    {
      "entropy": 6.134539842605591,
      "epoch": 3.32198499711483,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039292963001184676,
      "loss": 5.7218,
      "mean_token_accuracy": 0.19435421824455262,
      "num_tokens": 72957886.0,
      "step": 28785
    },
    {
      "entropy": 6.157153940200805,
      "epoch": 3.3225620311598383,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003928944857704636,
      "loss": 5.6561,
      "mean_token_accuracy": 0.20160724818706513,
      "num_tokens": 72969698.0,
      "step": 28790
    },
    {
      "entropy": 6.072511386871338,
      "epoch": 3.323139065204847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039285933756375794,
      "loss": 5.5837,
      "mean_token_accuracy": 0.20369360893964766,
      "num_tokens": 72983356.0,
      "step": 28795
    },
    {
      "entropy": 6.03831672668457,
      "epoch": 3.3237160992498556,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003928241853929119,
      "loss": 5.6639,
      "mean_token_accuracy": 0.20169093012809752,
      "num_tokens": 72995839.0,
      "step": 28800
    },
    {
      "entropy": 6.064248037338257,
      "epoch": 3.3242931332948644,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039278902925910815,
      "loss": 5.6469,
      "mean_token_accuracy": 0.19969455897808075,
      "num_tokens": 73008337.0,
      "step": 28805
    },
    {
      "entropy": 6.1179803848266605,
      "epoch": 3.3248701673398733,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039275386916352866,
      "loss": 5.7074,
      "mean_token_accuracy": 0.19655280113220214,
      "num_tokens": 73020262.0,
      "step": 28810
    },
    {
      "entropy": 6.159458255767822,
      "epoch": 3.3254472013848817,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003927187051073564,
      "loss": 5.7707,
      "mean_token_accuracy": 0.18897237330675126,
      "num_tokens": 73034142.0,
      "step": 28815
    },
    {
      "entropy": 6.182563066482544,
      "epoch": 3.32602423542989,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039268353709177395,
      "loss": 5.7086,
      "mean_token_accuracy": 0.1969788894057274,
      "num_tokens": 73046973.0,
      "step": 28820
    },
    {
      "entropy": 6.1018201351165775,
      "epoch": 3.326601269474899,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003926483651179642,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1932616800069809,
      "num_tokens": 73060036.0,
      "step": 28825
    },
    {
      "entropy": 6.030070018768311,
      "epoch": 3.327178303519908,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003926131891871101,
      "loss": 5.6104,
      "mean_token_accuracy": 0.1999261423945427,
      "num_tokens": 73072017.0,
      "step": 28830
    },
    {
      "entropy": 6.0897620677947994,
      "epoch": 3.3277553375649163,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039257800930039485,
      "loss": 5.5744,
      "mean_token_accuracy": 0.20028634816408158,
      "num_tokens": 73083937.0,
      "step": 28835
    },
    {
      "entropy": 6.0535084247589115,
      "epoch": 3.328332371609925,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003925428254590016,
      "loss": 5.7055,
      "mean_token_accuracy": 0.19825107455253602,
      "num_tokens": 73096084.0,
      "step": 28840
    },
    {
      "entropy": 6.186150979995728,
      "epoch": 3.3289094056549335,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00039250763766411384,
      "loss": 5.7882,
      "mean_token_accuracy": 0.19582255631685258,
      "num_tokens": 73109109.0,
      "step": 28845
    },
    {
      "entropy": 6.127252912521362,
      "epoch": 3.3294864396999424,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039247244591691504,
      "loss": 5.6153,
      "mean_token_accuracy": 0.20720813125371934,
      "num_tokens": 73121843.0,
      "step": 28850
    },
    {
      "entropy": 6.1502635955810545,
      "epoch": 3.330063473744951,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039243725021858894,
      "loss": 5.7425,
      "mean_token_accuracy": 0.18707681745290755,
      "num_tokens": 73135102.0,
      "step": 28855
    },
    {
      "entropy": 6.093119859695435,
      "epoch": 3.3306405077899597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003924020505703191,
      "loss": 5.6274,
      "mean_token_accuracy": 0.203123739361763,
      "num_tokens": 73148575.0,
      "step": 28860
    },
    {
      "entropy": 6.1708704948425295,
      "epoch": 3.331217541834968,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039236684697328974,
      "loss": 5.7501,
      "mean_token_accuracy": 0.1887365072965622,
      "num_tokens": 73160795.0,
      "step": 28865
    },
    {
      "entropy": 6.08322958946228,
      "epoch": 3.331794575879977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039233163942868475,
      "loss": 5.6556,
      "mean_token_accuracy": 0.20242613255977632,
      "num_tokens": 73173407.0,
      "step": 28870
    },
    {
      "entropy": 6.101664066314697,
      "epoch": 3.3323716099249854,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003922964279376883,
      "loss": 5.6329,
      "mean_token_accuracy": 0.20785931795835494,
      "num_tokens": 73187781.0,
      "step": 28875
    },
    {
      "entropy": 6.158063793182373,
      "epoch": 3.3329486439699942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003922612125014848,
      "loss": 5.7243,
      "mean_token_accuracy": 0.19410327523946763,
      "num_tokens": 73201701.0,
      "step": 28880
    },
    {
      "entropy": 6.062568521499633,
      "epoch": 3.333525678015003,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039222599312125874,
      "loss": 5.7067,
      "mean_token_accuracy": 0.1908780500292778,
      "num_tokens": 73214144.0,
      "step": 28885
    },
    {
      "entropy": 6.087817859649658,
      "epoch": 3.3341027120600115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003921907697981946,
      "loss": 5.5624,
      "mean_token_accuracy": 0.2039722129702568,
      "num_tokens": 73227314.0,
      "step": 28890
    },
    {
      "entropy": 6.139893865585327,
      "epoch": 3.3346797461050204,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00039215554253347706,
      "loss": 5.6709,
      "mean_token_accuracy": 0.19554894417524338,
      "num_tokens": 73240285.0,
      "step": 28895
    },
    {
      "entropy": 6.117650079727173,
      "epoch": 3.335256780150029,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003921203113282911,
      "loss": 5.7253,
      "mean_token_accuracy": 0.19506211578845978,
      "num_tokens": 73252839.0,
      "step": 28900
    },
    {
      "entropy": 6.077865839004517,
      "epoch": 3.3358338141950377,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003920850761838216,
      "loss": 5.6353,
      "mean_token_accuracy": 0.2037528470158577,
      "num_tokens": 73266894.0,
      "step": 28905
    },
    {
      "entropy": 6.144443130493164,
      "epoch": 3.336410848240046,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039204983710125377,
      "loss": 5.7025,
      "mean_token_accuracy": 0.1994447484612465,
      "num_tokens": 73279867.0,
      "step": 28910
    },
    {
      "entropy": 6.138685655593872,
      "epoch": 3.336987882285055,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039201459408177275,
      "loss": 5.703,
      "mean_token_accuracy": 0.19226930290460587,
      "num_tokens": 73292080.0,
      "step": 28915
    },
    {
      "entropy": 6.110139989852906,
      "epoch": 3.3375649163300634,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003919793471265641,
      "loss": 5.6134,
      "mean_token_accuracy": 0.19952233433723449,
      "num_tokens": 73303502.0,
      "step": 28920
    },
    {
      "entropy": 6.036711406707764,
      "epoch": 3.338141950375072,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039194409623681303,
      "loss": 5.5699,
      "mean_token_accuracy": 0.2070951670408249,
      "num_tokens": 73316169.0,
      "step": 28925
    },
    {
      "entropy": 6.16759467124939,
      "epoch": 3.3387189844200806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039190884141370553,
      "loss": 5.6585,
      "mean_token_accuracy": 0.20364924371242524,
      "num_tokens": 73327929.0,
      "step": 28930
    },
    {
      "entropy": 6.016793632507325,
      "epoch": 3.3392960184650895,
      "grad_norm": 1.0,
      "learning_rate": 0.00039187358265842714,
      "loss": 5.6084,
      "mean_token_accuracy": 0.20275096446275712,
      "num_tokens": 73339951.0,
      "step": 28935
    },
    {
      "entropy": 6.059000825881958,
      "epoch": 3.339873052510098,
      "grad_norm": 1.125,
      "learning_rate": 0.00039183831997216393,
      "loss": 5.5784,
      "mean_token_accuracy": 0.1988500952720642,
      "num_tokens": 73351959.0,
      "step": 28940
    },
    {
      "entropy": 6.012447357177734,
      "epoch": 3.340450086555107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003918030533561018,
      "loss": 5.5723,
      "mean_token_accuracy": 0.20370370894670486,
      "num_tokens": 73366727.0,
      "step": 28945
    },
    {
      "entropy": 6.1245378971099855,
      "epoch": 3.341027120600115,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039176778281142696,
      "loss": 5.6147,
      "mean_token_accuracy": 0.20102010667324066,
      "num_tokens": 73378429.0,
      "step": 28950
    },
    {
      "entropy": 6.09379734992981,
      "epoch": 3.341604154645124,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039173250833932576,
      "loss": 5.5839,
      "mean_token_accuracy": 0.20450549125671386,
      "num_tokens": 73391901.0,
      "step": 28955
    },
    {
      "entropy": 6.042441558837891,
      "epoch": 3.342181188690133,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039169722994098464,
      "loss": 5.6415,
      "mean_token_accuracy": 0.20028617829084397,
      "num_tokens": 73404827.0,
      "step": 28960
    },
    {
      "entropy": 6.149498128890992,
      "epoch": 3.3427582227351413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00039166194761759015,
      "loss": 5.6924,
      "mean_token_accuracy": 0.19924909770488738,
      "num_tokens": 73417941.0,
      "step": 28965
    },
    {
      "entropy": 6.1962748050689695,
      "epoch": 3.34333525678015,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00039162666137032906,
      "loss": 5.7091,
      "mean_token_accuracy": 0.19753915518522264,
      "num_tokens": 73429838.0,
      "step": 28970
    },
    {
      "entropy": 6.169246959686279,
      "epoch": 3.3439122908251586,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003915913712003882,
      "loss": 5.6863,
      "mean_token_accuracy": 0.19088172167539597,
      "num_tokens": 73441559.0,
      "step": 28975
    },
    {
      "entropy": 6.145298767089844,
      "epoch": 3.3444893248701675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003915560771089543,
      "loss": 5.736,
      "mean_token_accuracy": 0.18661364316940307,
      "num_tokens": 73454665.0,
      "step": 28980
    },
    {
      "entropy": 6.115599250793457,
      "epoch": 3.345066358915176,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003915207790972147,
      "loss": 5.6211,
      "mean_token_accuracy": 0.20005884468555452,
      "num_tokens": 73466969.0,
      "step": 28985
    },
    {
      "entropy": 6.151247835159301,
      "epoch": 3.3456433929601848,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003914854771663567,
      "loss": 5.706,
      "mean_token_accuracy": 0.19477426260709763,
      "num_tokens": 73481939.0,
      "step": 28990
    },
    {
      "entropy": 6.169728946685791,
      "epoch": 3.346220427005193,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00039145017131756745,
      "loss": 5.6837,
      "mean_token_accuracy": 0.19861585050821304,
      "num_tokens": 73496650.0,
      "step": 28995
    },
    {
      "entropy": 6.086485958099365,
      "epoch": 3.346797461050202,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003914148615520345,
      "loss": 5.6279,
      "mean_token_accuracy": 0.206010702252388,
      "num_tokens": 73508363.0,
      "step": 29000
    },
    {
      "entropy": 6.033454036712646,
      "epoch": 3.3473744950952105,
      "grad_norm": 0.890625,
      "learning_rate": 0.00039137954787094546,
      "loss": 5.615,
      "mean_token_accuracy": 0.19520912766456605,
      "num_tokens": 73522530.0,
      "step": 29005
    },
    {
      "entropy": 6.080453157424927,
      "epoch": 3.3479515291402193,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039134423027548825,
      "loss": 5.6086,
      "mean_token_accuracy": 0.20337537825107574,
      "num_tokens": 73535204.0,
      "step": 29010
    },
    {
      "entropy": 6.140754890441895,
      "epoch": 3.3485285631852277,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00039130890876685066,
      "loss": 5.6849,
      "mean_token_accuracy": 0.19961223900318145,
      "num_tokens": 73547783.0,
      "step": 29015
    },
    {
      "entropy": 6.067149591445923,
      "epoch": 3.3491055972302366,
      "grad_norm": 1.0,
      "learning_rate": 0.0003912735833462206,
      "loss": 5.7959,
      "mean_token_accuracy": 0.19198202341794968,
      "num_tokens": 73559707.0,
      "step": 29020
    },
    {
      "entropy": 6.12426905632019,
      "epoch": 3.349682631275245,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039123825401478633,
      "loss": 5.6733,
      "mean_token_accuracy": 0.20417928397655488,
      "num_tokens": 73571749.0,
      "step": 29025
    },
    {
      "entropy": 6.07002625465393,
      "epoch": 3.350259665320254,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003912029207737363,
      "loss": 5.6365,
      "mean_token_accuracy": 0.2024119019508362,
      "num_tokens": 73583876.0,
      "step": 29030
    },
    {
      "entropy": 5.989030838012695,
      "epoch": 3.3508366993652627,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039116758362425856,
      "loss": 5.5716,
      "mean_token_accuracy": 0.20426071137189866,
      "num_tokens": 73596899.0,
      "step": 29035
    },
    {
      "entropy": 6.091226053237915,
      "epoch": 3.351413733410271,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003911322425675419,
      "loss": 5.6258,
      "mean_token_accuracy": 0.20146955400705338,
      "num_tokens": 73609190.0,
      "step": 29040
    },
    {
      "entropy": 6.173632621765137,
      "epoch": 3.35199076745528,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003910968976047749,
      "loss": 5.6723,
      "mean_token_accuracy": 0.19326019585132598,
      "num_tokens": 73622410.0,
      "step": 29045
    },
    {
      "entropy": 5.932807922363281,
      "epoch": 3.3525678015002884,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003910615487371465,
      "loss": 5.4816,
      "mean_token_accuracy": 0.21557075828313826,
      "num_tokens": 73634540.0,
      "step": 29050
    },
    {
      "entropy": 6.0248185157775875,
      "epoch": 3.3531448355452973,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003910261959658455,
      "loss": 5.674,
      "mean_token_accuracy": 0.2020629957318306,
      "num_tokens": 73646392.0,
      "step": 29055
    },
    {
      "entropy": 6.099936580657959,
      "epoch": 3.3537218695903057,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003909908392920611,
      "loss": 5.6675,
      "mean_token_accuracy": 0.20144849866628647,
      "num_tokens": 73657929.0,
      "step": 29060
    },
    {
      "entropy": 6.061476564407348,
      "epoch": 3.3542989036353146,
      "grad_norm": 0.890625,
      "learning_rate": 0.00039095547871698236,
      "loss": 5.6626,
      "mean_token_accuracy": 0.2021855190396309,
      "num_tokens": 73670190.0,
      "step": 29065
    },
    {
      "entropy": 6.048135280609131,
      "epoch": 3.354875937680323,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039092011424179875,
      "loss": 5.6314,
      "mean_token_accuracy": 0.20407300293445588,
      "num_tokens": 73682634.0,
      "step": 29070
    },
    {
      "entropy": 6.162403774261475,
      "epoch": 3.355452971725332,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003908847458676996,
      "loss": 5.6159,
      "mean_token_accuracy": 0.20729955285787582,
      "num_tokens": 73694627.0,
      "step": 29075
    },
    {
      "entropy": 6.160260820388794,
      "epoch": 3.3560300057703403,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003908493735958747,
      "loss": 5.8224,
      "mean_token_accuracy": 0.18829547613859177,
      "num_tokens": 73707618.0,
      "step": 29080
    },
    {
      "entropy": 6.041449451446534,
      "epoch": 3.356607039815349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039081399742751363,
      "loss": 5.5586,
      "mean_token_accuracy": 0.20376075357198714,
      "num_tokens": 73719354.0,
      "step": 29085
    },
    {
      "entropy": 6.132792568206787,
      "epoch": 3.357184073860358,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039077861736380617,
      "loss": 5.6456,
      "mean_token_accuracy": 0.1984936699271202,
      "num_tokens": 73731586.0,
      "step": 29090
    },
    {
      "entropy": 6.12219123840332,
      "epoch": 3.3577611079053664,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039074323340594256,
      "loss": 5.7533,
      "mean_token_accuracy": 0.19600808322429658,
      "num_tokens": 73745907.0,
      "step": 29095
    },
    {
      "entropy": 6.000311422348022,
      "epoch": 3.358338141950375,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039070784555511276,
      "loss": 5.5001,
      "mean_token_accuracy": 0.21606164574623107,
      "num_tokens": 73759458.0,
      "step": 29100
    },
    {
      "entropy": 6.140221643447876,
      "epoch": 3.3589151759953837,
      "grad_norm": 1.0,
      "learning_rate": 0.000390672453812507,
      "loss": 5.6417,
      "mean_token_accuracy": 0.20022647827863693,
      "num_tokens": 73771116.0,
      "step": 29105
    },
    {
      "entropy": 6.069024467468262,
      "epoch": 3.3594922100403926,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039063705817931574,
      "loss": 5.7175,
      "mean_token_accuracy": 0.193771331012249,
      "num_tokens": 73784573.0,
      "step": 29110
    },
    {
      "entropy": 6.145127201080323,
      "epoch": 3.360069244085401,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003906016586567295,
      "loss": 5.6604,
      "mean_token_accuracy": 0.2027412474155426,
      "num_tokens": 73797402.0,
      "step": 29115
    },
    {
      "entropy": 6.223798513412476,
      "epoch": 3.36064627813041,
      "grad_norm": 0.828125,
      "learning_rate": 0.0003905662552459389,
      "loss": 5.7465,
      "mean_token_accuracy": 0.1938004106283188,
      "num_tokens": 73810924.0,
      "step": 29120
    },
    {
      "entropy": 6.11265230178833,
      "epoch": 3.3612233121754183,
      "grad_norm": 0.859375,
      "learning_rate": 0.00039053084794813466,
      "loss": 5.6766,
      "mean_token_accuracy": 0.1990928515791893,
      "num_tokens": 73824439.0,
      "step": 29125
    },
    {
      "entropy": 6.09653377532959,
      "epoch": 3.361800346220427,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039049543676450773,
      "loss": 5.6852,
      "mean_token_accuracy": 0.19784125685691833,
      "num_tokens": 73838322.0,
      "step": 29130
    },
    {
      "entropy": 6.145147848129272,
      "epoch": 3.3623773802654355,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003904600216962491,
      "loss": 5.6761,
      "mean_token_accuracy": 0.19254917800426483,
      "num_tokens": 73850598.0,
      "step": 29135
    },
    {
      "entropy": 6.116708850860595,
      "epoch": 3.3629544143104444,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003904246027445501,
      "loss": 5.6658,
      "mean_token_accuracy": 0.199068121612072,
      "num_tokens": 73863335.0,
      "step": 29140
    },
    {
      "entropy": 6.151806640625,
      "epoch": 3.363531448355453,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039038917991060187,
      "loss": 5.751,
      "mean_token_accuracy": 0.19046205133199692,
      "num_tokens": 73876638.0,
      "step": 29145
    },
    {
      "entropy": 6.110155344009399,
      "epoch": 3.3641084824004617,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003903537531955959,
      "loss": 5.7324,
      "mean_token_accuracy": 0.19594415575265883,
      "num_tokens": 73889640.0,
      "step": 29150
    },
    {
      "entropy": 6.109386587142945,
      "epoch": 3.36468551644547,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003903183226007237,
      "loss": 5.6865,
      "mean_token_accuracy": 0.2000272437930107,
      "num_tokens": 73901883.0,
      "step": 29155
    },
    {
      "entropy": 6.124618434906006,
      "epoch": 3.365262550490479,
      "grad_norm": 1.0,
      "learning_rate": 0.00039028288812717715,
      "loss": 5.6299,
      "mean_token_accuracy": 0.20795801728963853,
      "num_tokens": 73913717.0,
      "step": 29160
    },
    {
      "entropy": 6.07931547164917,
      "epoch": 3.365839584535488,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003902474497761478,
      "loss": 5.6274,
      "mean_token_accuracy": 0.19971617609262465,
      "num_tokens": 73928057.0,
      "step": 29165
    },
    {
      "entropy": 6.135083246231079,
      "epoch": 3.3664166185804962,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039021200754882774,
      "loss": 5.7324,
      "mean_token_accuracy": 0.19539804756641388,
      "num_tokens": 73941184.0,
      "step": 29170
    },
    {
      "entropy": 6.211198616027832,
      "epoch": 3.366993652625505,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000390176561446409,
      "loss": 5.7556,
      "mean_token_accuracy": 0.18804158568382262,
      "num_tokens": 73953607.0,
      "step": 29175
    },
    {
      "entropy": 6.146896409988403,
      "epoch": 3.3675706866705135,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003901411114700839,
      "loss": 5.6065,
      "mean_token_accuracy": 0.20742944926023482,
      "num_tokens": 73965238.0,
      "step": 29180
    },
    {
      "entropy": 6.093308782577514,
      "epoch": 3.3681477207155224,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003901056576210447,
      "loss": 5.6361,
      "mean_token_accuracy": 0.19821404069662094,
      "num_tokens": 73979213.0,
      "step": 29185
    },
    {
      "entropy": 6.123352336883545,
      "epoch": 3.368724754760531,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039007019990048387,
      "loss": 5.6683,
      "mean_token_accuracy": 0.19590487331151962,
      "num_tokens": 73991852.0,
      "step": 29190
    },
    {
      "entropy": 6.160832500457763,
      "epoch": 3.3693017888055397,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00039003473830959395,
      "loss": 5.7759,
      "mean_token_accuracy": 0.19638842344284058,
      "num_tokens": 74005244.0,
      "step": 29195
    },
    {
      "entropy": 5.995227432250976,
      "epoch": 3.369878822850548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038999927284956784,
      "loss": 5.4528,
      "mean_token_accuracy": 0.21742042303085327,
      "num_tokens": 74018163.0,
      "step": 29200
    },
    {
      "entropy": 6.10787353515625,
      "epoch": 3.370455856895557,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038996380352159833,
      "loss": 5.7078,
      "mean_token_accuracy": 0.1898741155862808,
      "num_tokens": 74030857.0,
      "step": 29205
    },
    {
      "entropy": 6.068977880477905,
      "epoch": 3.3710328909405654,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00038992833032687837,
      "loss": 5.5841,
      "mean_token_accuracy": 0.20431800335645675,
      "num_tokens": 74042659.0,
      "step": 29210
    },
    {
      "entropy": 6.213417482376099,
      "epoch": 3.371609924985574,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000389892853266601,
      "loss": 5.7026,
      "mean_token_accuracy": 0.19185577929019929,
      "num_tokens": 74055412.0,
      "step": 29215
    },
    {
      "entropy": 6.12598557472229,
      "epoch": 3.3721869590305826,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038985737234195975,
      "loss": 5.6655,
      "mean_token_accuracy": 0.20019135177135466,
      "num_tokens": 74068160.0,
      "step": 29220
    },
    {
      "entropy": 6.122549533843994,
      "epoch": 3.3727639930755915,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038982188755414774,
      "loss": 5.6403,
      "mean_token_accuracy": 0.2009674921631813,
      "num_tokens": 74080119.0,
      "step": 29225
    },
    {
      "entropy": 6.115718650817871,
      "epoch": 3.3733410271206,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038978639890435853,
      "loss": 5.7088,
      "mean_token_accuracy": 0.19333766251802445,
      "num_tokens": 74091768.0,
      "step": 29230
    },
    {
      "entropy": 6.140683126449585,
      "epoch": 3.3739180611656088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003897509063937859,
      "loss": 5.6983,
      "mean_token_accuracy": 0.1911182314157486,
      "num_tokens": 74104265.0,
      "step": 29235
    },
    {
      "entropy": 6.255893135070801,
      "epoch": 3.3744950952106176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003897154100236235,
      "loss": 5.7419,
      "mean_token_accuracy": 0.19009674936532975,
      "num_tokens": 74116322.0,
      "step": 29240
    },
    {
      "entropy": 6.206576490402222,
      "epoch": 3.375072129255626,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003896799097950653,
      "loss": 5.7686,
      "mean_token_accuracy": 0.18801714330911637,
      "num_tokens": 74127665.0,
      "step": 29245
    },
    {
      "entropy": 6.114025115966797,
      "epoch": 3.375649163300635,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003896444057093052,
      "loss": 5.6666,
      "mean_token_accuracy": 0.1956649050116539,
      "num_tokens": 74140860.0,
      "step": 29250
    },
    {
      "entropy": 6.136866521835327,
      "epoch": 3.3762261973456433,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038960889776753756,
      "loss": 5.739,
      "mean_token_accuracy": 0.19141697138547897,
      "num_tokens": 74153049.0,
      "step": 29255
    },
    {
      "entropy": 6.2237043380737305,
      "epoch": 3.376803231390652,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003895733859709565,
      "loss": 5.7883,
      "mean_token_accuracy": 0.18805213123559952,
      "num_tokens": 74166701.0,
      "step": 29260
    },
    {
      "entropy": 6.154474544525146,
      "epoch": 3.3773802654356606,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003895378703207566,
      "loss": 5.6574,
      "mean_token_accuracy": 0.18983679711818696,
      "num_tokens": 74178995.0,
      "step": 29265
    },
    {
      "entropy": 6.068514966964722,
      "epoch": 3.3779572994806695,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003895023508181323,
      "loss": 5.601,
      "mean_token_accuracy": 0.20221629589796067,
      "num_tokens": 74191579.0,
      "step": 29270
    },
    {
      "entropy": 6.06135630607605,
      "epoch": 3.378534333525678,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038946682746427824,
      "loss": 5.629,
      "mean_token_accuracy": 0.2009536638855934,
      "num_tokens": 74206241.0,
      "step": 29275
    },
    {
      "entropy": 6.209331512451172,
      "epoch": 3.3791113675706868,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00038943130026038934,
      "loss": 5.7095,
      "mean_token_accuracy": 0.1880050778388977,
      "num_tokens": 74218380.0,
      "step": 29280
    },
    {
      "entropy": 6.094493770599366,
      "epoch": 3.379688401615695,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003893957692076605,
      "loss": 5.5648,
      "mean_token_accuracy": 0.205158269405365,
      "num_tokens": 74230352.0,
      "step": 29285
    },
    {
      "entropy": 5.9807212352752686,
      "epoch": 3.380265435660704,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00038936023430728684,
      "loss": 5.5468,
      "mean_token_accuracy": 0.2043629989027977,
      "num_tokens": 74242817.0,
      "step": 29290
    },
    {
      "entropy": 6.1047436714172365,
      "epoch": 3.3808424697057124,
      "grad_norm": 1.0,
      "learning_rate": 0.0003893246955604635,
      "loss": 5.6448,
      "mean_token_accuracy": 0.1972112163901329,
      "num_tokens": 74255937.0,
      "step": 29295
    },
    {
      "entropy": 6.0725905418396,
      "epoch": 3.3814195037507213,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003892891529683857,
      "loss": 5.6046,
      "mean_token_accuracy": 0.2048858106136322,
      "num_tokens": 74269179.0,
      "step": 29300
    },
    {
      "entropy": 6.035662889480591,
      "epoch": 3.3819965377957297,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038925360653224906,
      "loss": 5.6314,
      "mean_token_accuracy": 0.20917874425649643,
      "num_tokens": 74281623.0,
      "step": 29305
    },
    {
      "entropy": 6.09220814704895,
      "epoch": 3.3825735718407386,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003892180562532491,
      "loss": 5.6006,
      "mean_token_accuracy": 0.20383199900388718,
      "num_tokens": 74293079.0,
      "step": 29310
    },
    {
      "entropy": 6.18207106590271,
      "epoch": 3.3831506058857475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003891825021325817,
      "loss": 5.7448,
      "mean_token_accuracy": 0.18867139369249344,
      "num_tokens": 74306034.0,
      "step": 29315
    },
    {
      "entropy": 6.109046602249146,
      "epoch": 3.383727639930756,
      "grad_norm": 1.265625,
      "learning_rate": 0.00038914694417144235,
      "loss": 5.6382,
      "mean_token_accuracy": 0.20215804725885392,
      "num_tokens": 74318576.0,
      "step": 29320
    },
    {
      "entropy": 6.099431085586548,
      "epoch": 3.3843046739757647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038911138237102735,
      "loss": 5.6562,
      "mean_token_accuracy": 0.18943738639354707,
      "num_tokens": 74330856.0,
      "step": 29325
    },
    {
      "entropy": 6.1361926078796385,
      "epoch": 3.384881708020773,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003890758167325327,
      "loss": 5.6913,
      "mean_token_accuracy": 0.19551837146282197,
      "num_tokens": 74343488.0,
      "step": 29330
    },
    {
      "entropy": 6.176275968551636,
      "epoch": 3.385458742065782,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038904024725715453,
      "loss": 5.8028,
      "mean_token_accuracy": 0.1887669861316681,
      "num_tokens": 74356592.0,
      "step": 29335
    },
    {
      "entropy": 6.186302185058594,
      "epoch": 3.3860357761107904,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038900467394608934,
      "loss": 5.6769,
      "mean_token_accuracy": 0.19546194225549698,
      "num_tokens": 74369987.0,
      "step": 29340
    },
    {
      "entropy": 6.161084699630737,
      "epoch": 3.3866128101557993,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003889690968005336,
      "loss": 5.6633,
      "mean_token_accuracy": 0.2024553120136261,
      "num_tokens": 74382186.0,
      "step": 29345
    },
    {
      "entropy": 6.038913774490356,
      "epoch": 3.3871898442008077,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038893351582168387,
      "loss": 5.6172,
      "mean_token_accuracy": 0.20843385756015778,
      "num_tokens": 74395486.0,
      "step": 29350
    },
    {
      "entropy": 6.074850130081177,
      "epoch": 3.3877668782458166,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038889793101073696,
      "loss": 5.6685,
      "mean_token_accuracy": 0.1960757628083229,
      "num_tokens": 74407641.0,
      "step": 29355
    },
    {
      "entropy": 6.1654150009155275,
      "epoch": 3.388343912290825,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038886234236888966,
      "loss": 5.6902,
      "mean_token_accuracy": 0.19943959563970565,
      "num_tokens": 74420036.0,
      "step": 29360
    },
    {
      "entropy": 5.984434509277344,
      "epoch": 3.388920946335834,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003888267498973391,
      "loss": 5.5644,
      "mean_token_accuracy": 0.2148554190993309,
      "num_tokens": 74433682.0,
      "step": 29365
    },
    {
      "entropy": 6.1506205081939695,
      "epoch": 3.3894979803808427,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003887911535972823,
      "loss": 5.7739,
      "mean_token_accuracy": 0.19183535128831863,
      "num_tokens": 74445127.0,
      "step": 29370
    },
    {
      "entropy": 6.102530288696289,
      "epoch": 3.390075014425851,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003887555534699166,
      "loss": 5.5839,
      "mean_token_accuracy": 0.2062132939696312,
      "num_tokens": 74458445.0,
      "step": 29375
    },
    {
      "entropy": 6.091544198989868,
      "epoch": 3.3906520484708595,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003887199495164393,
      "loss": 5.6401,
      "mean_token_accuracy": 0.19999001622200013,
      "num_tokens": 74472025.0,
      "step": 29380
    },
    {
      "entropy": 6.178681373596191,
      "epoch": 3.3912290825158684,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000388684341738048,
      "loss": 5.7175,
      "mean_token_accuracy": 0.19466122090816498,
      "num_tokens": 74485016.0,
      "step": 29385
    },
    {
      "entropy": 6.1808967113494875,
      "epoch": 3.3918061165608773,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038864873013594043,
      "loss": 5.7361,
      "mean_token_accuracy": 0.19160934090614318,
      "num_tokens": 74498762.0,
      "step": 29390
    },
    {
      "entropy": 6.145447015762329,
      "epoch": 3.3923831506058857,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003886131147113143,
      "loss": 5.7773,
      "mean_token_accuracy": 0.18902845233678817,
      "num_tokens": 74511864.0,
      "step": 29395
    },
    {
      "entropy": 6.1231317043304445,
      "epoch": 3.3929601846508946,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003885774954653673,
      "loss": 5.679,
      "mean_token_accuracy": 0.1956196904182434,
      "num_tokens": 74524887.0,
      "step": 29400
    },
    {
      "entropy": 6.140392446517945,
      "epoch": 3.393537218695903,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003885418723992977,
      "loss": 5.6012,
      "mean_token_accuracy": 0.2015588954091072,
      "num_tokens": 74537490.0,
      "step": 29405
    },
    {
      "entropy": 6.019869327545166,
      "epoch": 3.394114252740912,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003885062455143038,
      "loss": 5.5171,
      "mean_token_accuracy": 0.21081402599811555,
      "num_tokens": 74550027.0,
      "step": 29410
    },
    {
      "entropy": 6.007146310806275,
      "epoch": 3.3946912867859202,
      "grad_norm": 1.0,
      "learning_rate": 0.00038847061481158356,
      "loss": 5.5995,
      "mean_token_accuracy": 0.2112478122115135,
      "num_tokens": 74562318.0,
      "step": 29415
    },
    {
      "entropy": 6.156685495376587,
      "epoch": 3.395268320830929,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038843498029233553,
      "loss": 5.7231,
      "mean_token_accuracy": 0.19226054549217225,
      "num_tokens": 74575316.0,
      "step": 29420
    },
    {
      "entropy": 6.137768363952636,
      "epoch": 3.3958453548759375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003883993419577584,
      "loss": 5.6658,
      "mean_token_accuracy": 0.20140229314565658,
      "num_tokens": 74589043.0,
      "step": 29425
    },
    {
      "entropy": 5.976077508926392,
      "epoch": 3.3964223889209464,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003883636998090507,
      "loss": 5.5387,
      "mean_token_accuracy": 0.2075169414281845,
      "num_tokens": 74602796.0,
      "step": 29430
    },
    {
      "entropy": 6.112657165527343,
      "epoch": 3.396999422965955,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038832805384741124,
      "loss": 5.6386,
      "mean_token_accuracy": 0.20106805860996246,
      "num_tokens": 74615987.0,
      "step": 29435
    },
    {
      "entropy": 6.195538330078125,
      "epoch": 3.3975764570109637,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003882924040740389,
      "loss": 5.7369,
      "mean_token_accuracy": 0.18946260064840317,
      "num_tokens": 74627940.0,
      "step": 29440
    },
    {
      "entropy": 6.184897804260254,
      "epoch": 3.3981534910559725,
      "grad_norm": 0.890625,
      "learning_rate": 0.000388256750490133,
      "loss": 5.7017,
      "mean_token_accuracy": 0.1875414788722992,
      "num_tokens": 74641977.0,
      "step": 29445
    },
    {
      "entropy": 6.149797439575195,
      "epoch": 3.398730525100981,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003882210930968924,
      "loss": 5.6634,
      "mean_token_accuracy": 0.19859969913959502,
      "num_tokens": 74655119.0,
      "step": 29450
    },
    {
      "entropy": 6.124272012710572,
      "epoch": 3.3993075591459894,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003881854318955167,
      "loss": 5.7287,
      "mean_token_accuracy": 0.19595022797584533,
      "num_tokens": 74667548.0,
      "step": 29455
    },
    {
      "entropy": 6.10913667678833,
      "epoch": 3.3998845931909982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038814976688720523,
      "loss": 5.6325,
      "mean_token_accuracy": 0.20080644637346268,
      "num_tokens": 74679365.0,
      "step": 29460
    },
    {
      "entropy": 6.120306587219238,
      "epoch": 3.400461627236007,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038811409807315746,
      "loss": 5.684,
      "mean_token_accuracy": 0.19400031715631486,
      "num_tokens": 74690713.0,
      "step": 29465
    },
    {
      "entropy": 6.083793354034424,
      "epoch": 3.4010386612810155,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003880784254545732,
      "loss": 5.5889,
      "mean_token_accuracy": 0.20644036084413528,
      "num_tokens": 74703106.0,
      "step": 29470
    },
    {
      "entropy": 6.129297351837158,
      "epoch": 3.4016156953260244,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038804274903265226,
      "loss": 5.7176,
      "mean_token_accuracy": 0.19394056349992753,
      "num_tokens": 74715602.0,
      "step": 29475
    },
    {
      "entropy": 6.147923469543457,
      "epoch": 3.402192729371033,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003880070688085947,
      "loss": 5.6516,
      "mean_token_accuracy": 0.19438060671091079,
      "num_tokens": 74727842.0,
      "step": 29480
    },
    {
      "entropy": 6.100360679626465,
      "epoch": 3.4027697634160416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003879713847836004,
      "loss": 5.635,
      "mean_token_accuracy": 0.2019089177250862,
      "num_tokens": 74740480.0,
      "step": 29485
    },
    {
      "entropy": 5.960062074661255,
      "epoch": 3.40334679746105,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038793569695886964,
      "loss": 5.5248,
      "mean_token_accuracy": 0.20709266662597656,
      "num_tokens": 74753939.0,
      "step": 29490
    },
    {
      "entropy": 6.057261514663696,
      "epoch": 3.403923831506059,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000387900005335603,
      "loss": 5.5106,
      "mean_token_accuracy": 0.20501282662153245,
      "num_tokens": 74767055.0,
      "step": 29495
    },
    {
      "entropy": 6.056798315048217,
      "epoch": 3.4045008655510673,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038786430991500064,
      "loss": 5.5768,
      "mean_token_accuracy": 0.2034056305885315,
      "num_tokens": 74779927.0,
      "step": 29500
    },
    {
      "entropy": 6.104568958282471,
      "epoch": 3.405077899596076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038782861069826323,
      "loss": 5.7033,
      "mean_token_accuracy": 0.19332396537065505,
      "num_tokens": 74792845.0,
      "step": 29505
    },
    {
      "entropy": 6.098080444335937,
      "epoch": 3.4056549336410846,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003877929076865917,
      "loss": 5.6164,
      "mean_token_accuracy": 0.2020713523030281,
      "num_tokens": 74807288.0,
      "step": 29510
    },
    {
      "entropy": 6.210349178314209,
      "epoch": 3.4062319676860935,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038775720088118665,
      "loss": 5.7326,
      "mean_token_accuracy": 0.19356990307569505,
      "num_tokens": 74819537.0,
      "step": 29515
    },
    {
      "entropy": 6.173000812530518,
      "epoch": 3.4068090017311023,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038772149028324916,
      "loss": 5.7539,
      "mean_token_accuracy": 0.1912834197282791,
      "num_tokens": 74830890.0,
      "step": 29520
    },
    {
      "entropy": 6.109610748291016,
      "epoch": 3.4073860357761108,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038768577589398033,
      "loss": 5.6735,
      "mean_token_accuracy": 0.204718779027462,
      "num_tokens": 74843286.0,
      "step": 29525
    },
    {
      "entropy": 6.098098182678223,
      "epoch": 3.4079630698211196,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038765005771458145,
      "loss": 5.6208,
      "mean_token_accuracy": 0.20144327878952026,
      "num_tokens": 74854743.0,
      "step": 29530
    },
    {
      "entropy": 6.1022031784057615,
      "epoch": 3.408540103866128,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003876143357462538,
      "loss": 5.651,
      "mean_token_accuracy": 0.19274944067001343,
      "num_tokens": 74867031.0,
      "step": 29535
    },
    {
      "entropy": 6.085686588287354,
      "epoch": 3.409117137911137,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003875786099901989,
      "loss": 5.5849,
      "mean_token_accuracy": 0.20769606679677963,
      "num_tokens": 74880045.0,
      "step": 29540
    },
    {
      "entropy": 6.096614694595337,
      "epoch": 3.4096941719561453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038754288044761835,
      "loss": 5.7055,
      "mean_token_accuracy": 0.1931267186999321,
      "num_tokens": 74893671.0,
      "step": 29545
    },
    {
      "entropy": 6.17376971244812,
      "epoch": 3.410271206001154,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038750714711971397,
      "loss": 5.7209,
      "mean_token_accuracy": 0.1895170271396637,
      "num_tokens": 74906580.0,
      "step": 29550
    },
    {
      "entropy": 6.1414618492126465,
      "epoch": 3.4108482400461626,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00038747141000768754,
      "loss": 5.7186,
      "mean_token_accuracy": 0.1929113507270813,
      "num_tokens": 74919540.0,
      "step": 29555
    },
    {
      "entropy": 6.073526620864868,
      "epoch": 3.4114252740911715,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038743566911274117,
      "loss": 5.6117,
      "mean_token_accuracy": 0.19729551821947097,
      "num_tokens": 74930931.0,
      "step": 29560
    },
    {
      "entropy": 6.047491931915284,
      "epoch": 3.41200230813618,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038739992443607686,
      "loss": 5.6043,
      "mean_token_accuracy": 0.20607941150665282,
      "num_tokens": 74943531.0,
      "step": 29565
    },
    {
      "entropy": 6.127089881896973,
      "epoch": 3.4125793421811887,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038736417597889694,
      "loss": 5.7204,
      "mean_token_accuracy": 0.19538949579000472,
      "num_tokens": 74955808.0,
      "step": 29570
    },
    {
      "entropy": 6.150683736801147,
      "epoch": 3.413156376226197,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003873284237424038,
      "loss": 5.6547,
      "mean_token_accuracy": 0.2048019826412201,
      "num_tokens": 74967412.0,
      "step": 29575
    },
    {
      "entropy": 6.09517731666565,
      "epoch": 3.413733410271206,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038729266772779994,
      "loss": 5.6447,
      "mean_token_accuracy": 0.19893274009227752,
      "num_tokens": 74979889.0,
      "step": 29580
    },
    {
      "entropy": 6.106925916671753,
      "epoch": 3.4143104443162144,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038725690793628794,
      "loss": 5.6048,
      "mean_token_accuracy": 0.20259840935468673,
      "num_tokens": 74992550.0,
      "step": 29585
    },
    {
      "entropy": 6.088739490509033,
      "epoch": 3.4148874783612233,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038722114436907056,
      "loss": 5.6795,
      "mean_token_accuracy": 0.19715912640094757,
      "num_tokens": 75005940.0,
      "step": 29590
    },
    {
      "entropy": 6.081855726242066,
      "epoch": 3.415464512406232,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003871853770273508,
      "loss": 5.6012,
      "mean_token_accuracy": 0.198858542740345,
      "num_tokens": 75018102.0,
      "step": 29595
    },
    {
      "entropy": 6.134600687026977,
      "epoch": 3.4160415464512406,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003871496059123316,
      "loss": 5.6571,
      "mean_token_accuracy": 0.19920790195465088,
      "num_tokens": 75031192.0,
      "step": 29600
    },
    {
      "entropy": 6.148057126998902,
      "epoch": 3.4166185804962494,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003871138310252161,
      "loss": 5.7188,
      "mean_token_accuracy": 0.20085345953702927,
      "num_tokens": 75042791.0,
      "step": 29605
    },
    {
      "entropy": 6.135023784637451,
      "epoch": 3.417195614541258,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003870780523672075,
      "loss": 5.6842,
      "mean_token_accuracy": 0.19695188999176025,
      "num_tokens": 75055649.0,
      "step": 29610
    },
    {
      "entropy": 6.0502818584442135,
      "epoch": 3.4177726485862667,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003870422699395094,
      "loss": 5.6054,
      "mean_token_accuracy": 0.20833633989095687,
      "num_tokens": 75068209.0,
      "step": 29615
    },
    {
      "entropy": 6.117885112762451,
      "epoch": 3.418349682631275,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038700648374332514,
      "loss": 5.6231,
      "mean_token_accuracy": 0.2067483514547348,
      "num_tokens": 75080606.0,
      "step": 29620
    },
    {
      "entropy": 6.151182460784912,
      "epoch": 3.418926716676284,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003869706937798585,
      "loss": 5.7085,
      "mean_token_accuracy": 0.19779286980628968,
      "num_tokens": 75094585.0,
      "step": 29625
    },
    {
      "entropy": 6.175467395782471,
      "epoch": 3.4195037507212924,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038693490005031316,
      "loss": 5.765,
      "mean_token_accuracy": 0.19431446641683578,
      "num_tokens": 75106953.0,
      "step": 29630
    },
    {
      "entropy": 6.1615156650543215,
      "epoch": 3.4200807847663013,
      "grad_norm": 0.875,
      "learning_rate": 0.00038689910255589304,
      "loss": 5.7342,
      "mean_token_accuracy": 0.18868091106414794,
      "num_tokens": 75119506.0,
      "step": 29635
    },
    {
      "entropy": 6.085178852081299,
      "epoch": 3.4206578188113097,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038686330129780223,
      "loss": 5.5538,
      "mean_token_accuracy": 0.20220239609479904,
      "num_tokens": 75132219.0,
      "step": 29640
    },
    {
      "entropy": 6.153129911422729,
      "epoch": 3.4212348528563186,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038682749627724485,
      "loss": 5.7022,
      "mean_token_accuracy": 0.19907985478639603,
      "num_tokens": 75143880.0,
      "step": 29645
    },
    {
      "entropy": 6.052488231658936,
      "epoch": 3.4218118869013274,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003867916874954251,
      "loss": 5.5956,
      "mean_token_accuracy": 0.20786201804876328,
      "num_tokens": 75156673.0,
      "step": 29650
    },
    {
      "entropy": 6.104920434951782,
      "epoch": 3.422388920946336,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003867558749535475,
      "loss": 5.6067,
      "mean_token_accuracy": 0.19818609654903413,
      "num_tokens": 75168846.0,
      "step": 29655
    },
    {
      "entropy": 6.153359460830688,
      "epoch": 3.4229659549913443,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003867200586528166,
      "loss": 5.6826,
      "mean_token_accuracy": 0.2006438657641411,
      "num_tokens": 75181046.0,
      "step": 29660
    },
    {
      "entropy": 6.0886023998260494,
      "epoch": 3.423542989036353,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038668423859443705,
      "loss": 5.622,
      "mean_token_accuracy": 0.20443181544542313,
      "num_tokens": 75193266.0,
      "step": 29665
    },
    {
      "entropy": 6.017653083801269,
      "epoch": 3.424120023081362,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003866484147796136,
      "loss": 5.5372,
      "mean_token_accuracy": 0.20884893089532852,
      "num_tokens": 75204267.0,
      "step": 29670
    },
    {
      "entropy": 6.119679880142212,
      "epoch": 3.4246970571263704,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038661258720955105,
      "loss": 5.7099,
      "mean_token_accuracy": 0.19929271340370178,
      "num_tokens": 75216932.0,
      "step": 29675
    },
    {
      "entropy": 6.08361177444458,
      "epoch": 3.4252740911713793,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038657675588545467,
      "loss": 5.5936,
      "mean_token_accuracy": 0.20028530210256576,
      "num_tokens": 75229084.0,
      "step": 29680
    },
    {
      "entropy": 6.047318172454834,
      "epoch": 3.4258511252163877,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003865409208085295,
      "loss": 5.6856,
      "mean_token_accuracy": 0.197913059592247,
      "num_tokens": 75241356.0,
      "step": 29685
    },
    {
      "entropy": 6.052640295028686,
      "epoch": 3.4264281592613965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038650508197998095,
      "loss": 5.6026,
      "mean_token_accuracy": 0.204889939725399,
      "num_tokens": 75252768.0,
      "step": 29690
    },
    {
      "entropy": 6.075566387176513,
      "epoch": 3.427005193306405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003864692394010143,
      "loss": 5.6342,
      "mean_token_accuracy": 0.20133931785821915,
      "num_tokens": 75265517.0,
      "step": 29695
    },
    {
      "entropy": 6.2092978954315186,
      "epoch": 3.427582227351414,
      "grad_norm": 1.0,
      "learning_rate": 0.00038643339307283507,
      "loss": 5.8072,
      "mean_token_accuracy": 0.18519458025693894,
      "num_tokens": 75278191.0,
      "step": 29700
    },
    {
      "entropy": 6.098193979263305,
      "epoch": 3.4281592613964222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038639754299664913,
      "loss": 5.5716,
      "mean_token_accuracy": 0.2043537065386772,
      "num_tokens": 75290794.0,
      "step": 29705
    },
    {
      "entropy": 6.1307285785675045,
      "epoch": 3.428736295441431,
      "grad_norm": 0.90625,
      "learning_rate": 0.00038636168917366214,
      "loss": 5.729,
      "mean_token_accuracy": 0.19462745487689972,
      "num_tokens": 75303438.0,
      "step": 29710
    },
    {
      "entropy": 6.06499490737915,
      "epoch": 3.4293133294864395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038632583160508,
      "loss": 5.6315,
      "mean_token_accuracy": 0.19745396375656127,
      "num_tokens": 75317336.0,
      "step": 29715
    },
    {
      "entropy": 6.164554691314697,
      "epoch": 3.4298903635314484,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038628997029210887,
      "loss": 5.6883,
      "mean_token_accuracy": 0.19707987010478972,
      "num_tokens": 75328952.0,
      "step": 29720
    },
    {
      "entropy": 6.189921903610229,
      "epoch": 3.4304673975764572,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038625410523595487,
      "loss": 5.7309,
      "mean_token_accuracy": 0.1925668701529503,
      "num_tokens": 75341116.0,
      "step": 29725
    },
    {
      "entropy": 6.115351819992066,
      "epoch": 3.4310444316214657,
      "grad_norm": 0.875,
      "learning_rate": 0.00038621823643782426,
      "loss": 5.715,
      "mean_token_accuracy": 0.19497351348400116,
      "num_tokens": 75354150.0,
      "step": 29730
    },
    {
      "entropy": 6.127152633666992,
      "epoch": 3.431621465666474,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003861823638989235,
      "loss": 5.6333,
      "mean_token_accuracy": 0.20510401725769042,
      "num_tokens": 75367809.0,
      "step": 29735
    },
    {
      "entropy": 5.998108196258545,
      "epoch": 3.432198499711483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038614648762045923,
      "loss": 5.5162,
      "mean_token_accuracy": 0.21514474004507064,
      "num_tokens": 75380536.0,
      "step": 29740
    },
    {
      "entropy": 6.008195495605468,
      "epoch": 3.432775533756492,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000386110607603638,
      "loss": 5.5644,
      "mean_token_accuracy": 0.2020410791039467,
      "num_tokens": 75392994.0,
      "step": 29745
    },
    {
      "entropy": 6.11928014755249,
      "epoch": 3.4333525678015,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003860747238496667,
      "loss": 5.6601,
      "mean_token_accuracy": 0.19896188378334045,
      "num_tokens": 75405344.0,
      "step": 29750
    },
    {
      "entropy": 6.146852827072143,
      "epoch": 3.433929601846509,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038603883635975224,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1926712363958359,
      "num_tokens": 75418059.0,
      "step": 29755
    },
    {
      "entropy": 6.156238651275634,
      "epoch": 3.4345066358915175,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003860029451351016,
      "loss": 5.7103,
      "mean_token_accuracy": 0.19851579368114472,
      "num_tokens": 75431316.0,
      "step": 29760
    },
    {
      "entropy": 6.127622365951538,
      "epoch": 3.4350836699365264,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038596705017692216,
      "loss": 5.6404,
      "mean_token_accuracy": 0.200221648812294,
      "num_tokens": 75443542.0,
      "step": 29765
    },
    {
      "entropy": 6.101916790008545,
      "epoch": 3.435660703981535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038593115148642107,
      "loss": 5.6967,
      "mean_token_accuracy": 0.19485777020454406,
      "num_tokens": 75455422.0,
      "step": 29770
    },
    {
      "entropy": 6.04071307182312,
      "epoch": 3.4362377380265436,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038589524906480576,
      "loss": 5.6359,
      "mean_token_accuracy": 0.20377057790756226,
      "num_tokens": 75468702.0,
      "step": 29775
    },
    {
      "entropy": 6.078157997131347,
      "epoch": 3.436814772071552,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003858593429132838,
      "loss": 5.5813,
      "mean_token_accuracy": 0.2084820792078972,
      "num_tokens": 75480217.0,
      "step": 29780
    },
    {
      "entropy": 6.1541829109191895,
      "epoch": 3.437391806116561,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00038582343303306306,
      "loss": 5.6918,
      "mean_token_accuracy": 0.1958676680922508,
      "num_tokens": 75492826.0,
      "step": 29785
    },
    {
      "entropy": 6.1325288772583,
      "epoch": 3.4379688401615693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003857875194253511,
      "loss": 5.7224,
      "mean_token_accuracy": 0.19690243005752564,
      "num_tokens": 75504398.0,
      "step": 29790
    },
    {
      "entropy": 6.113526630401611,
      "epoch": 3.438545874206578,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003857516020913559,
      "loss": 5.6434,
      "mean_token_accuracy": 0.19955738335847856,
      "num_tokens": 75516028.0,
      "step": 29795
    },
    {
      "entropy": 6.1810108661651615,
      "epoch": 3.439122908251587,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00038571568103228565,
      "loss": 5.6899,
      "mean_token_accuracy": 0.20068355053663253,
      "num_tokens": 75528811.0,
      "step": 29800
    },
    {
      "entropy": 6.084721946716309,
      "epoch": 3.4396999422965955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003856797562493485,
      "loss": 5.579,
      "mean_token_accuracy": 0.2105577364563942,
      "num_tokens": 75542784.0,
      "step": 29805
    },
    {
      "entropy": 6.183552837371826,
      "epoch": 3.4402769763416043,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003856438277437527,
      "loss": 5.7185,
      "mean_token_accuracy": 0.1952583149075508,
      "num_tokens": 75555295.0,
      "step": 29810
    },
    {
      "entropy": 6.1390893936157225,
      "epoch": 3.4408540103866128,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038560789551670663,
      "loss": 5.7117,
      "mean_token_accuracy": 0.1996374875307083,
      "num_tokens": 75568234.0,
      "step": 29815
    },
    {
      "entropy": 6.114632940292358,
      "epoch": 3.4414310444316216,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038557195956941896,
      "loss": 5.6728,
      "mean_token_accuracy": 0.19620812833309173,
      "num_tokens": 75580865.0,
      "step": 29820
    },
    {
      "entropy": 6.187064981460571,
      "epoch": 3.44200807847663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038553601990309845,
      "loss": 5.7315,
      "mean_token_accuracy": 0.18760953694581986,
      "num_tokens": 75593679.0,
      "step": 29825
    },
    {
      "entropy": 6.2334929466247555,
      "epoch": 3.442585112521639,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038550007651895376,
      "loss": 5.7552,
      "mean_token_accuracy": 0.19637635201215745,
      "num_tokens": 75607061.0,
      "step": 29830
    },
    {
      "entropy": 6.1914153575897215,
      "epoch": 3.4431621465666473,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038546412941819385,
      "loss": 5.6557,
      "mean_token_accuracy": 0.19760273694992064,
      "num_tokens": 75619729.0,
      "step": 29835
    },
    {
      "entropy": 6.212173795700073,
      "epoch": 3.443739180611656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003854281786020279,
      "loss": 5.7283,
      "mean_token_accuracy": 0.18926496505737306,
      "num_tokens": 75631942.0,
      "step": 29840
    },
    {
      "entropy": 6.124419403076172,
      "epoch": 3.4443162146566646,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038539222407166495,
      "loss": 5.6556,
      "mean_token_accuracy": 0.2042503371834755,
      "num_tokens": 75646181.0,
      "step": 29845
    },
    {
      "entropy": 6.0467602729797365,
      "epoch": 3.4448932487016735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003853562658283144,
      "loss": 5.5389,
      "mean_token_accuracy": 0.2085070088505745,
      "num_tokens": 75657808.0,
      "step": 29850
    },
    {
      "entropy": 6.101883792877198,
      "epoch": 3.445470282746682,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038532030387318565,
      "loss": 5.6297,
      "mean_token_accuracy": 0.2070928171277046,
      "num_tokens": 75670166.0,
      "step": 29855
    },
    {
      "entropy": 6.158042001724243,
      "epoch": 3.4460473167916907,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038528433820748833,
      "loss": 5.7725,
      "mean_token_accuracy": 0.19122541099786758,
      "num_tokens": 75682511.0,
      "step": 29860
    },
    {
      "entropy": 6.241187047958374,
      "epoch": 3.446624350836699,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038524836883243197,
      "loss": 5.7343,
      "mean_token_accuracy": 0.19922872483730317,
      "num_tokens": 75694785.0,
      "step": 29865
    },
    {
      "entropy": 6.171201515197754,
      "epoch": 3.447201384881708,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038521239574922664,
      "loss": 5.7396,
      "mean_token_accuracy": 0.19373102933168412,
      "num_tokens": 75707908.0,
      "step": 29870
    },
    {
      "entropy": 6.118019390106201,
      "epoch": 3.447778418926717,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000385176418959082,
      "loss": 5.6314,
      "mean_token_accuracy": 0.1992066130042076,
      "num_tokens": 75720733.0,
      "step": 29875
    },
    {
      "entropy": 6.058538627624512,
      "epoch": 3.4483554529717253,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038514043846320825,
      "loss": 5.5394,
      "mean_token_accuracy": 0.20247066915035247,
      "num_tokens": 75732511.0,
      "step": 29880
    },
    {
      "entropy": 6.078581619262695,
      "epoch": 3.448932487016734,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038510445426281556,
      "loss": 5.6344,
      "mean_token_accuracy": 0.2075391560792923,
      "num_tokens": 75744949.0,
      "step": 29885
    },
    {
      "entropy": 6.052122259140015,
      "epoch": 3.4495095210617426,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003850684663591143,
      "loss": 5.6593,
      "mean_token_accuracy": 0.20081855803728105,
      "num_tokens": 75758049.0,
      "step": 29890
    },
    {
      "entropy": 6.11500358581543,
      "epoch": 3.4500865551067514,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003850324747533148,
      "loss": 5.5933,
      "mean_token_accuracy": 0.20215218067169188,
      "num_tokens": 75772242.0,
      "step": 29895
    },
    {
      "entropy": 6.0574428081512455,
      "epoch": 3.45066358915176,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003849964794466277,
      "loss": 5.6042,
      "mean_token_accuracy": 0.20257695019245148,
      "num_tokens": 75784711.0,
      "step": 29900
    },
    {
      "entropy": 6.184817266464234,
      "epoch": 3.4512406231967687,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003849604804402636,
      "loss": 5.7421,
      "mean_token_accuracy": 0.18540485352277755,
      "num_tokens": 75796757.0,
      "step": 29905
    },
    {
      "entropy": 6.119535970687866,
      "epoch": 3.451817657241777,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038492447773543344,
      "loss": 5.7091,
      "mean_token_accuracy": 0.19726979732513428,
      "num_tokens": 75810195.0,
      "step": 29910
    },
    {
      "entropy": 6.145711421966553,
      "epoch": 3.452394691286786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000384888471333348,
      "loss": 5.6758,
      "mean_token_accuracy": 0.20566664785146713,
      "num_tokens": 75822645.0,
      "step": 29915
    },
    {
      "entropy": 6.099514055252075,
      "epoch": 3.4529717253317944,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003848524612352184,
      "loss": 5.6856,
      "mean_token_accuracy": 0.1920801430940628,
      "num_tokens": 75834277.0,
      "step": 29920
    },
    {
      "entropy": 6.133346652984619,
      "epoch": 3.4535487593768033,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003848164474422559,
      "loss": 5.6784,
      "mean_token_accuracy": 0.19923868924379348,
      "num_tokens": 75846111.0,
      "step": 29925
    },
    {
      "entropy": 6.224630117416382,
      "epoch": 3.4541257934218117,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038478042995567175,
      "loss": 5.7037,
      "mean_token_accuracy": 0.19568544030189514,
      "num_tokens": 75858453.0,
      "step": 29930
    },
    {
      "entropy": 6.14057502746582,
      "epoch": 3.4547028274668206,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003847444087766773,
      "loss": 5.7376,
      "mean_token_accuracy": 0.19105585813522338,
      "num_tokens": 75870127.0,
      "step": 29935
    },
    {
      "entropy": 6.026944828033447,
      "epoch": 3.455279861511829,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038470838390648423,
      "loss": 5.5559,
      "mean_token_accuracy": 0.20011749863624573,
      "num_tokens": 75883320.0,
      "step": 29940
    },
    {
      "entropy": 6.069692182540893,
      "epoch": 3.455856895556838,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038467235534630405,
      "loss": 5.6433,
      "mean_token_accuracy": 0.1909273758530617,
      "num_tokens": 75897433.0,
      "step": 29945
    },
    {
      "entropy": 6.154717588424683,
      "epoch": 3.4564339296018467,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003846363230973488,
      "loss": 5.6346,
      "mean_token_accuracy": 0.1993203043937683,
      "num_tokens": 75909919.0,
      "step": 29950
    },
    {
      "entropy": 6.145422649383545,
      "epoch": 3.457010963646855,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038460028716083024,
      "loss": 5.7148,
      "mean_token_accuracy": 0.198944990336895,
      "num_tokens": 75923101.0,
      "step": 29955
    },
    {
      "entropy": 6.022760200500488,
      "epoch": 3.457587997691864,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003845642475379605,
      "loss": 5.6163,
      "mean_token_accuracy": 0.19617946296930314,
      "num_tokens": 75936331.0,
      "step": 29960
    },
    {
      "entropy": 6.14332947731018,
      "epoch": 3.4581650317368724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003845282042299516,
      "loss": 5.5985,
      "mean_token_accuracy": 0.20384318083524705,
      "num_tokens": 75947921.0,
      "step": 29965
    },
    {
      "entropy": 6.136355543136597,
      "epoch": 3.4587420657818813,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003844921572380161,
      "loss": 5.62,
      "mean_token_accuracy": 0.1979285404086113,
      "num_tokens": 75960454.0,
      "step": 29970
    },
    {
      "entropy": 6.107445621490479,
      "epoch": 3.4593190998268897,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003844561065633662,
      "loss": 5.7218,
      "mean_token_accuracy": 0.19018979072570802,
      "num_tokens": 75972909.0,
      "step": 29975
    },
    {
      "entropy": 6.128451633453369,
      "epoch": 3.4598961338718985,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003844200522072145,
      "loss": 5.6193,
      "mean_token_accuracy": 0.1998048573732376,
      "num_tokens": 75985360.0,
      "step": 29980
    },
    {
      "entropy": 6.137365913391113,
      "epoch": 3.460473167916907,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003843839941707738,
      "loss": 5.5964,
      "mean_token_accuracy": 0.2057602271437645,
      "num_tokens": 75996434.0,
      "step": 29985
    },
    {
      "entropy": 6.055975151062012,
      "epoch": 3.461050201961916,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003843479324552567,
      "loss": 5.6117,
      "mean_token_accuracy": 0.19530674815177917,
      "num_tokens": 76008721.0,
      "step": 29990
    },
    {
      "entropy": 6.133390188217163,
      "epoch": 3.4616272360069242,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038431186706187626,
      "loss": 5.6959,
      "mean_token_accuracy": 0.19397947639226915,
      "num_tokens": 76021775.0,
      "step": 29995
    },
    {
      "entropy": 6.104957056045532,
      "epoch": 3.462204270051933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038427579799184546,
      "loss": 5.6331,
      "mean_token_accuracy": 0.20255368947982788,
      "num_tokens": 76034607.0,
      "step": 30000
    },
    {
      "epoch": 3.462204270051933,
      "eval_entropy": 5.971964979700518,
      "eval_loss": 5.795499801635742,
      "eval_mean_token_accuracy": 0.2002593917851086,
      "eval_num_tokens": 76034607.0,
      "eval_runtime": 22.8734,
      "eval_samples_per_second": 1230.076,
      "eval_steps_per_second": 153.76,
      "step": 30000
    },
    {
      "entropy": 6.116971206665039,
      "epoch": 3.462781304096942,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003842397252463774,
      "loss": 5.6857,
      "mean_token_accuracy": 0.19487107545137405,
      "num_tokens": 76048694.0,
      "step": 30005
    },
    {
      "entropy": 6.163836860656739,
      "epoch": 3.4633583381419504,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003842036488266856,
      "loss": 5.63,
      "mean_token_accuracy": 0.19729094505310057,
      "num_tokens": 76061130.0,
      "step": 30010
    },
    {
      "entropy": 6.082635068893433,
      "epoch": 3.463935372186959,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003841675687339832,
      "loss": 5.5723,
      "mean_token_accuracy": 0.19619133323431015,
      "num_tokens": 76072933.0,
      "step": 30015
    },
    {
      "entropy": 6.1245582580566404,
      "epoch": 3.4645124062319677,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003841314849694839,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1994563892483711,
      "num_tokens": 76085380.0,
      "step": 30020
    },
    {
      "entropy": 6.074115467071533,
      "epoch": 3.4650894402769765,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003840953975344012,
      "loss": 5.6277,
      "mean_token_accuracy": 0.20541253238916396,
      "num_tokens": 76097948.0,
      "step": 30025
    },
    {
      "entropy": 6.185827398300171,
      "epoch": 3.465666474321985,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038405930642994914,
      "loss": 5.6916,
      "mean_token_accuracy": 0.19627868086099626,
      "num_tokens": 76111665.0,
      "step": 30030
    },
    {
      "entropy": 6.105507659912109,
      "epoch": 3.466243508366994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003840232116573414,
      "loss": 5.6577,
      "mean_token_accuracy": 0.19467235803604127,
      "num_tokens": 76124060.0,
      "step": 30035
    },
    {
      "entropy": 6.182118988037109,
      "epoch": 3.466820542412002,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003839871132177921,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1960706204175949,
      "num_tokens": 76136865.0,
      "step": 30040
    },
    {
      "entropy": 6.088036346435547,
      "epoch": 3.467397576457011,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003839510111125155,
      "loss": 5.6312,
      "mean_token_accuracy": 0.195370677113533,
      "num_tokens": 76151068.0,
      "step": 30045
    },
    {
      "entropy": 6.060251951217651,
      "epoch": 3.4679746105020195,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003839149053427256,
      "loss": 5.5311,
      "mean_token_accuracy": 0.21270291805267333,
      "num_tokens": 76163481.0,
      "step": 30050
    },
    {
      "entropy": 6.109107255935669,
      "epoch": 3.4685516445470284,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038387879590963696,
      "loss": 5.6341,
      "mean_token_accuracy": 0.2059398427605629,
      "num_tokens": 76175673.0,
      "step": 30055
    },
    {
      "entropy": 6.082851934432983,
      "epoch": 3.4691286785920368,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003838426828144642,
      "loss": 5.6161,
      "mean_token_accuracy": 0.19848414957523347,
      "num_tokens": 76188113.0,
      "step": 30060
    },
    {
      "entropy": 6.155608320236206,
      "epoch": 3.4697057126370456,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038380656605842186,
      "loss": 5.681,
      "mean_token_accuracy": 0.18838610202074052,
      "num_tokens": 76200318.0,
      "step": 30065
    },
    {
      "entropy": 6.06912145614624,
      "epoch": 3.470282746682054,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003837704456427247,
      "loss": 5.6579,
      "mean_token_accuracy": 0.20329781472682953,
      "num_tokens": 76212135.0,
      "step": 30070
    },
    {
      "entropy": 6.104427242279053,
      "epoch": 3.470859780727063,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038373432156858756,
      "loss": 5.6381,
      "mean_token_accuracy": 0.20531237572431565,
      "num_tokens": 76225023.0,
      "step": 30075
    },
    {
      "entropy": 6.190938282012939,
      "epoch": 3.4714368147720718,
      "grad_norm": 1.0,
      "learning_rate": 0.0003836981938372256,
      "loss": 5.7401,
      "mean_token_accuracy": 0.18945249617099763,
      "num_tokens": 76238196.0,
      "step": 30080
    },
    {
      "entropy": 6.169899749755859,
      "epoch": 3.47201384881708,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038366206244985383,
      "loss": 5.7319,
      "mean_token_accuracy": 0.194173426926136,
      "num_tokens": 76251634.0,
      "step": 30085
    },
    {
      "entropy": 6.173556566238403,
      "epoch": 3.472590882862089,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038362592740768766,
      "loss": 5.6606,
      "mean_token_accuracy": 0.19777562469244003,
      "num_tokens": 76265378.0,
      "step": 30090
    },
    {
      "entropy": 6.07429256439209,
      "epoch": 3.4731679169070975,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038358978871194226,
      "loss": 5.6711,
      "mean_token_accuracy": 0.19805676937103273,
      "num_tokens": 76277889.0,
      "step": 30095
    },
    {
      "entropy": 6.139761304855346,
      "epoch": 3.4737449509521063,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003835536463638334,
      "loss": 5.6663,
      "mean_token_accuracy": 0.1934621140360832,
      "num_tokens": 76289891.0,
      "step": 30100
    },
    {
      "entropy": 6.138318586349487,
      "epoch": 3.4743219849971148,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003835175003645764,
      "loss": 5.6807,
      "mean_token_accuracy": 0.19737306386232376,
      "num_tokens": 76301362.0,
      "step": 30105
    },
    {
      "entropy": 6.135520076751709,
      "epoch": 3.4748990190421236,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003834813507153873,
      "loss": 5.6234,
      "mean_token_accuracy": 0.1988933190703392,
      "num_tokens": 76313243.0,
      "step": 30110
    },
    {
      "entropy": 6.063269090652466,
      "epoch": 3.475476053087132,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003834451974174818,
      "loss": 5.6281,
      "mean_token_accuracy": 0.2100749433040619,
      "num_tokens": 76325777.0,
      "step": 30115
    },
    {
      "entropy": 6.143790531158447,
      "epoch": 3.476053087132141,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038340904047207597,
      "loss": 5.6811,
      "mean_token_accuracy": 0.19428813308477402,
      "num_tokens": 76338783.0,
      "step": 30120
    },
    {
      "entropy": 6.120352268218994,
      "epoch": 3.4766301211771493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038337287988038593,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19149962812662125,
      "num_tokens": 76350626.0,
      "step": 30125
    },
    {
      "entropy": 6.11890606880188,
      "epoch": 3.477207155222158,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038333671564362787,
      "loss": 5.6325,
      "mean_token_accuracy": 0.19931449145078659,
      "num_tokens": 76363819.0,
      "step": 30130
    },
    {
      "entropy": 6.125185537338257,
      "epoch": 3.4777841892671666,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003833005477630182,
      "loss": 5.5743,
      "mean_token_accuracy": 0.20325966626405717,
      "num_tokens": 76375119.0,
      "step": 30135
    },
    {
      "entropy": 6.152893304824829,
      "epoch": 3.4783612233121755,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003832643762397734,
      "loss": 5.7171,
      "mean_token_accuracy": 0.19343860000371932,
      "num_tokens": 76387762.0,
      "step": 30140
    },
    {
      "entropy": 6.050218629837036,
      "epoch": 3.478938257357184,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0003832282010751101,
      "loss": 5.6516,
      "mean_token_accuracy": 0.2044332057237625,
      "num_tokens": 76401449.0,
      "step": 30145
    },
    {
      "entropy": 6.126082992553711,
      "epoch": 3.4795152914021927,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038319202227024503,
      "loss": 5.6309,
      "mean_token_accuracy": 0.2095692291855812,
      "num_tokens": 76414591.0,
      "step": 30150
    },
    {
      "entropy": 6.1600096225738525,
      "epoch": 3.4800923254472016,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000383155839826395,
      "loss": 5.6315,
      "mean_token_accuracy": 0.20340606570243835,
      "num_tokens": 76426987.0,
      "step": 30155
    },
    {
      "entropy": 6.141730880737304,
      "epoch": 3.48066935949221,
      "grad_norm": 1.0,
      "learning_rate": 0.000383119653744777,
      "loss": 5.6727,
      "mean_token_accuracy": 0.1953875780105591,
      "num_tokens": 76439733.0,
      "step": 30160
    },
    {
      "entropy": 6.180585813522339,
      "epoch": 3.481246393537219,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038308346402660825,
      "loss": 5.7049,
      "mean_token_accuracy": 0.19781098365783692,
      "num_tokens": 76451836.0,
      "step": 30165
    },
    {
      "entropy": 6.148232173919678,
      "epoch": 3.4818234275822273,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003830472706731058,
      "loss": 5.7443,
      "mean_token_accuracy": 0.19583188742399216,
      "num_tokens": 76465036.0,
      "step": 30170
    },
    {
      "entropy": 6.067029333114624,
      "epoch": 3.482400461627236,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003830110736854871,
      "loss": 5.5874,
      "mean_token_accuracy": 0.20140080600976945,
      "num_tokens": 76477523.0,
      "step": 30175
    },
    {
      "entropy": 6.002858829498291,
      "epoch": 3.4829774956722446,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003829748730649695,
      "loss": 5.5125,
      "mean_token_accuracy": 0.20988383144140244,
      "num_tokens": 76490456.0,
      "step": 30180
    },
    {
      "entropy": 6.095826196670532,
      "epoch": 3.4835545297172534,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038293866881277083,
      "loss": 5.662,
      "mean_token_accuracy": 0.19799549132585526,
      "num_tokens": 76503371.0,
      "step": 30185
    },
    {
      "entropy": 6.054685735702515,
      "epoch": 3.484131563762262,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038290246093010855,
      "loss": 5.6571,
      "mean_token_accuracy": 0.1949399471282959,
      "num_tokens": 76516434.0,
      "step": 30190
    },
    {
      "entropy": 6.16289005279541,
      "epoch": 3.4847085978072707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038286624941820066,
      "loss": 5.6912,
      "mean_token_accuracy": 0.20036305636167526,
      "num_tokens": 76530147.0,
      "step": 30195
    },
    {
      "entropy": 6.17379322052002,
      "epoch": 3.485285631852279,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00038283003427826503,
      "loss": 5.768,
      "mean_token_accuracy": 0.19312623292207717,
      "num_tokens": 76543563.0,
      "step": 30200
    },
    {
      "entropy": 6.110541105270386,
      "epoch": 3.485862665897288,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038279381551151973,
      "loss": 5.5952,
      "mean_token_accuracy": 0.20653214156627656,
      "num_tokens": 76556488.0,
      "step": 30205
    },
    {
      "entropy": 6.157381010055542,
      "epoch": 3.4864396999422964,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003827575931191829,
      "loss": 5.6688,
      "mean_token_accuracy": 0.2037475049495697,
      "num_tokens": 76569800.0,
      "step": 30210
    },
    {
      "entropy": 6.135750150680542,
      "epoch": 3.4870167339873053,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003827213671024731,
      "loss": 5.7277,
      "mean_token_accuracy": 0.19277291148900985,
      "num_tokens": 76582468.0,
      "step": 30215
    },
    {
      "entropy": 6.155987405776978,
      "epoch": 3.4875937680323137,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003826851374626086,
      "loss": 5.7058,
      "mean_token_accuracy": 0.1915629878640175,
      "num_tokens": 76595207.0,
      "step": 30220
    },
    {
      "entropy": 6.20315523147583,
      "epoch": 3.4881708020773226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038264890420080795,
      "loss": 5.7562,
      "mean_token_accuracy": 0.19180708676576613,
      "num_tokens": 76608020.0,
      "step": 30225
    },
    {
      "entropy": 6.128471803665161,
      "epoch": 3.4887478361223314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003826126673182899,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1984275385737419,
      "num_tokens": 76620671.0,
      "step": 30230
    },
    {
      "entropy": 6.203733348846436,
      "epoch": 3.48932487016734,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003825764268162733,
      "loss": 5.7742,
      "mean_token_accuracy": 0.18989548236131668,
      "num_tokens": 76632723.0,
      "step": 30235
    },
    {
      "entropy": 6.123423957824707,
      "epoch": 3.4899019042123487,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003825401826959769,
      "loss": 5.7142,
      "mean_token_accuracy": 0.19327777475118638,
      "num_tokens": 76644811.0,
      "step": 30240
    },
    {
      "entropy": 6.074578714370728,
      "epoch": 3.490478938257357,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003825039349586199,
      "loss": 5.6406,
      "mean_token_accuracy": 0.20226256251335145,
      "num_tokens": 76656674.0,
      "step": 30245
    },
    {
      "entropy": 6.2125060081481935,
      "epoch": 3.491055972302366,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003824676836054215,
      "loss": 5.7074,
      "mean_token_accuracy": 0.1934037834405899,
      "num_tokens": 76669421.0,
      "step": 30250
    },
    {
      "entropy": 6.183376789093018,
      "epoch": 3.4916330063473744,
      "grad_norm": 0.890625,
      "learning_rate": 0.00038243142863760086,
      "loss": 5.7032,
      "mean_token_accuracy": 0.19767117649316787,
      "num_tokens": 76681273.0,
      "step": 30255
    },
    {
      "entropy": 6.094620180130005,
      "epoch": 3.4922100403923833,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003823951700563776,
      "loss": 5.6076,
      "mean_token_accuracy": 0.20499284714460372,
      "num_tokens": 76692949.0,
      "step": 30260
    },
    {
      "entropy": 6.083004999160766,
      "epoch": 3.4927870744373917,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003823589078629711,
      "loss": 5.6975,
      "mean_token_accuracy": 0.19626200646162034,
      "num_tokens": 76705650.0,
      "step": 30265
    },
    {
      "entropy": 6.184039115905762,
      "epoch": 3.4933641084824005,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038232264205860105,
      "loss": 5.6651,
      "mean_token_accuracy": 0.20441165417432786,
      "num_tokens": 76718756.0,
      "step": 30270
    },
    {
      "entropy": 6.2080934047698975,
      "epoch": 3.493941142527409,
      "grad_norm": 0.875,
      "learning_rate": 0.00038228637264448717,
      "loss": 5.7025,
      "mean_token_accuracy": 0.19622491002082826,
      "num_tokens": 76731434.0,
      "step": 30275
    },
    {
      "entropy": 6.174611043930054,
      "epoch": 3.494518176572418,
      "grad_norm": 1.0,
      "learning_rate": 0.00038225009962184956,
      "loss": 5.6761,
      "mean_token_accuracy": 0.19993743002414704,
      "num_tokens": 76743413.0,
      "step": 30280
    },
    {
      "entropy": 6.121385097503662,
      "epoch": 3.4950952106174267,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038221382299190803,
      "loss": 5.6897,
      "mean_token_accuracy": 0.19764068871736526,
      "num_tokens": 76756545.0,
      "step": 30285
    },
    {
      "entropy": 6.055549049377442,
      "epoch": 3.495672244662435,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003821775427558829,
      "loss": 5.577,
      "mean_token_accuracy": 0.2102785140275955,
      "num_tokens": 76769795.0,
      "step": 30290
    },
    {
      "entropy": 6.132511615753174,
      "epoch": 3.4962492787074435,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003821412589149943,
      "loss": 5.6917,
      "mean_token_accuracy": 0.1965790271759033,
      "num_tokens": 76782713.0,
      "step": 30295
    },
    {
      "entropy": 6.094518184661865,
      "epoch": 3.4968263127524524,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003821049714704627,
      "loss": 5.6439,
      "mean_token_accuracy": 0.19943341761827468,
      "num_tokens": 76794588.0,
      "step": 30300
    },
    {
      "entropy": 6.061864376068115,
      "epoch": 3.4974033467974612,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038206868042350855,
      "loss": 5.5752,
      "mean_token_accuracy": 0.20061683654785156,
      "num_tokens": 76807358.0,
      "step": 30305
    },
    {
      "entropy": 6.072805500030517,
      "epoch": 3.4979803808424696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003820323857753526,
      "loss": 5.6811,
      "mean_token_accuracy": 0.19788562804460524,
      "num_tokens": 76821334.0,
      "step": 30310
    },
    {
      "entropy": 6.065698909759521,
      "epoch": 3.4985574148874785,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038199608752721535,
      "loss": 5.5677,
      "mean_token_accuracy": 0.2045666053891182,
      "num_tokens": 76835691.0,
      "step": 30315
    },
    {
      "entropy": 6.179615449905396,
      "epoch": 3.499134448932487,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000381959785680318,
      "loss": 5.6928,
      "mean_token_accuracy": 0.19413203746080399,
      "num_tokens": 76847277.0,
      "step": 30320
    },
    {
      "entropy": 6.167409896850586,
      "epoch": 3.499711482977496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00038192348023588144,
      "loss": 5.7129,
      "mean_token_accuracy": 0.193221053481102,
      "num_tokens": 76860063.0,
      "step": 30325
    },
    {
      "entropy": 6.1212207794189455,
      "epoch": 3.500288517022504,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003818871711951266,
      "loss": 5.7027,
      "mean_token_accuracy": 0.19126492738723755,
      "num_tokens": 76871748.0,
      "step": 30330
    },
    {
      "entropy": 6.200066375732422,
      "epoch": 3.500865551067513,
      "grad_norm": 0.875,
      "learning_rate": 0.0003818508585592749,
      "loss": 5.7544,
      "mean_token_accuracy": 0.19638368487358093,
      "num_tokens": 76884958.0,
      "step": 30335
    },
    {
      "entropy": 6.160165119171142,
      "epoch": 3.5014425851125215,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003818145423295476,
      "loss": 5.6893,
      "mean_token_accuracy": 0.19635801166296005,
      "num_tokens": 76897924.0,
      "step": 30340
    },
    {
      "entropy": 6.127073764801025,
      "epoch": 3.5020196191575304,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003817782225071663,
      "loss": 5.7466,
      "mean_token_accuracy": 0.19972751587629317,
      "num_tokens": 76909374.0,
      "step": 30345
    },
    {
      "entropy": 6.132970476150513,
      "epoch": 3.5025966532025388,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003817418990933525,
      "loss": 5.6498,
      "mean_token_accuracy": 0.2034968763589859,
      "num_tokens": 76920629.0,
      "step": 30350
    },
    {
      "entropy": 6.156960153579712,
      "epoch": 3.5031736872475476,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038170557208932793,
      "loss": 5.7445,
      "mean_token_accuracy": 0.19333231151103974,
      "num_tokens": 76932539.0,
      "step": 30355
    },
    {
      "entropy": 6.190456914901733,
      "epoch": 3.5037507212925565,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038166924149631444,
      "loss": 5.7344,
      "mean_token_accuracy": 0.19323522001504898,
      "num_tokens": 76945210.0,
      "step": 30360
    },
    {
      "entropy": 6.135421657562256,
      "epoch": 3.504327755337565,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038163290731553403,
      "loss": 5.7077,
      "mean_token_accuracy": 0.19858040064573287,
      "num_tokens": 76958465.0,
      "step": 30365
    },
    {
      "entropy": 6.0406403064727785,
      "epoch": 3.5049047893825733,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003815965695482088,
      "loss": 5.5159,
      "mean_token_accuracy": 0.20895894169807433,
      "num_tokens": 76970563.0,
      "step": 30370
    },
    {
      "entropy": 6.128117465972901,
      "epoch": 3.505481823427582,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003815602281955608,
      "loss": 5.7071,
      "mean_token_accuracy": 0.19536690413951874,
      "num_tokens": 76983215.0,
      "step": 30375
    },
    {
      "entropy": 6.135241174697876,
      "epoch": 3.506058857472591,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00038152388325881243,
      "loss": 5.651,
      "mean_token_accuracy": 0.19385750740766525,
      "num_tokens": 76995433.0,
      "step": 30380
    },
    {
      "entropy": 6.131248664855957,
      "epoch": 3.5066358915175995,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003814875347391862,
      "loss": 5.6456,
      "mean_token_accuracy": 0.20077474415302277,
      "num_tokens": 77007427.0,
      "step": 30385
    },
    {
      "entropy": 6.129200267791748,
      "epoch": 3.5072129255626083,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038145118263790474,
      "loss": 5.627,
      "mean_token_accuracy": 0.2046310856938362,
      "num_tokens": 77019916.0,
      "step": 30390
    },
    {
      "entropy": 6.117960453033447,
      "epoch": 3.5077899596076167,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003814148269561905,
      "loss": 5.6825,
      "mean_token_accuracy": 0.2021052598953247,
      "num_tokens": 77032385.0,
      "step": 30395
    },
    {
      "entropy": 6.131684923171997,
      "epoch": 3.5083669936526256,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038137846769526645,
      "loss": 5.6988,
      "mean_token_accuracy": 0.1952743798494339,
      "num_tokens": 77045459.0,
      "step": 30400
    },
    {
      "entropy": 6.205504083633423,
      "epoch": 3.508944027697634,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003813421048563555,
      "loss": 5.6866,
      "mean_token_accuracy": 0.19790969043970108,
      "num_tokens": 77058386.0,
      "step": 30405
    },
    {
      "entropy": 6.10381851196289,
      "epoch": 3.509521061742643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003813057384406807,
      "loss": 5.6482,
      "mean_token_accuracy": 0.20617625415325164,
      "num_tokens": 77071085.0,
      "step": 30410
    },
    {
      "entropy": 6.115856122970581,
      "epoch": 3.5100980957876513,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003812693684494652,
      "loss": 5.6986,
      "mean_token_accuracy": 0.1923053815960884,
      "num_tokens": 77084619.0,
      "step": 30415
    },
    {
      "entropy": 6.1303667545318605,
      "epoch": 3.51067512983266,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038123299488393225,
      "loss": 5.6267,
      "mean_token_accuracy": 0.20061812847852706,
      "num_tokens": 77097144.0,
      "step": 30420
    },
    {
      "entropy": 6.002131605148316,
      "epoch": 3.5112521638776686,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003811966177453052,
      "loss": 5.4938,
      "mean_token_accuracy": 0.2166547879576683,
      "num_tokens": 77109875.0,
      "step": 30425
    },
    {
      "entropy": 6.120227909088134,
      "epoch": 3.5118291979226774,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003811602370348078,
      "loss": 5.6678,
      "mean_token_accuracy": 0.20702530443668365,
      "num_tokens": 77123022.0,
      "step": 30430
    },
    {
      "entropy": 6.107196950912476,
      "epoch": 3.5124062319676863,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038112385275366345,
      "loss": 5.6153,
      "mean_token_accuracy": 0.20772140920162202,
      "num_tokens": 77134845.0,
      "step": 30435
    },
    {
      "entropy": 6.093294954299926,
      "epoch": 3.5129832660126947,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003810874649030961,
      "loss": 5.5404,
      "mean_token_accuracy": 0.21414842903614045,
      "num_tokens": 77147204.0,
      "step": 30440
    },
    {
      "entropy": 6.12237696647644,
      "epoch": 3.513560300057703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038105107348432955,
      "loss": 5.6181,
      "mean_token_accuracy": 0.21166556179523469,
      "num_tokens": 77160374.0,
      "step": 30445
    },
    {
      "entropy": 6.041955614089966,
      "epoch": 3.514137334102712,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038101467849858777,
      "loss": 5.635,
      "mean_token_accuracy": 0.2026488408446312,
      "num_tokens": 77172558.0,
      "step": 30450
    },
    {
      "entropy": 6.167855834960937,
      "epoch": 3.514714368147721,
      "grad_norm": 0.96875,
      "learning_rate": 0.000380978279947095,
      "loss": 5.6769,
      "mean_token_accuracy": 0.19385907500982286,
      "num_tokens": 77185838.0,
      "step": 30455
    },
    {
      "entropy": 6.014473724365234,
      "epoch": 3.5152914021927293,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038094187783107545,
      "loss": 5.5321,
      "mean_token_accuracy": 0.2116029068827629,
      "num_tokens": 77197335.0,
      "step": 30460
    },
    {
      "entropy": 6.128593063354492,
      "epoch": 3.515868436237738,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038090547215175336,
      "loss": 5.6881,
      "mean_token_accuracy": 0.19996896833181382,
      "num_tokens": 77209180.0,
      "step": 30465
    },
    {
      "entropy": 6.026110696792602,
      "epoch": 3.5164454702827466,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003808690629103534,
      "loss": 5.4983,
      "mean_token_accuracy": 0.2137421876192093,
      "num_tokens": 77221393.0,
      "step": 30470
    },
    {
      "entropy": 5.930377435684204,
      "epoch": 3.5170225043277554,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003808326501081001,
      "loss": 5.5155,
      "mean_token_accuracy": 0.21995847374200822,
      "num_tokens": 77234152.0,
      "step": 30475
    },
    {
      "entropy": 6.126913022994995,
      "epoch": 3.517599538372764,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00038079623374621816,
      "loss": 5.672,
      "mean_token_accuracy": 0.19826942384243013,
      "num_tokens": 77248056.0,
      "step": 30480
    },
    {
      "entropy": 6.110919618606568,
      "epoch": 3.5181765724177727,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038075981382593236,
      "loss": 5.6768,
      "mean_token_accuracy": 0.19786618053913116,
      "num_tokens": 77262006.0,
      "step": 30485
    },
    {
      "entropy": 6.119785499572754,
      "epoch": 3.5187536064627816,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003807233903484679,
      "loss": 5.7,
      "mean_token_accuracy": 0.19304803162813186,
      "num_tokens": 77275419.0,
      "step": 30490
    },
    {
      "entropy": 6.162968254089355,
      "epoch": 3.51933064050779,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003806869633150497,
      "loss": 5.6819,
      "mean_token_accuracy": 0.19188643097877503,
      "num_tokens": 77287960.0,
      "step": 30495
    },
    {
      "entropy": 6.056926250457764,
      "epoch": 3.5199076745527984,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038065053272690297,
      "loss": 5.5824,
      "mean_token_accuracy": 0.2100779339671135,
      "num_tokens": 77300174.0,
      "step": 30500
    },
    {
      "entropy": 6.079039287567139,
      "epoch": 3.5204847085978073,
      "grad_norm": 0.9375,
      "learning_rate": 0.000380614098585253,
      "loss": 5.5571,
      "mean_token_accuracy": 0.2080882743000984,
      "num_tokens": 77312736.0,
      "step": 30505
    },
    {
      "entropy": 6.083523559570312,
      "epoch": 3.521061742642816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038057766089132533,
      "loss": 5.6459,
      "mean_token_accuracy": 0.20180234909057618,
      "num_tokens": 77325344.0,
      "step": 30510
    },
    {
      "entropy": 6.1154541015625,
      "epoch": 3.5216387766878245,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038054121964634547,
      "loss": 5.6107,
      "mean_token_accuracy": 0.20732833594083785,
      "num_tokens": 77337917.0,
      "step": 30515
    },
    {
      "entropy": 6.06594033241272,
      "epoch": 3.522215810732833,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003805047748515391,
      "loss": 5.6897,
      "mean_token_accuracy": 0.2076156795024872,
      "num_tokens": 77350654.0,
      "step": 30520
    },
    {
      "entropy": 6.010657501220703,
      "epoch": 3.522792844777842,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000380468326508132,
      "loss": 5.6097,
      "mean_token_accuracy": 0.20525875836610794,
      "num_tokens": 77363969.0,
      "step": 30525
    },
    {
      "entropy": 6.073437023162842,
      "epoch": 3.5233698788228507,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038043187461735025,
      "loss": 5.5762,
      "mean_token_accuracy": 0.2024112582206726,
      "num_tokens": 77377354.0,
      "step": 30530
    },
    {
      "entropy": 6.205715703964233,
      "epoch": 3.523946912867859,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003803954191804197,
      "loss": 5.6985,
      "mean_token_accuracy": 0.1945109859108925,
      "num_tokens": 77389990.0,
      "step": 30535
    },
    {
      "entropy": 6.10283899307251,
      "epoch": 3.524523946912868,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038035896019856656,
      "loss": 5.5798,
      "mean_token_accuracy": 0.20296572744846345,
      "num_tokens": 77401809.0,
      "step": 30540
    },
    {
      "entropy": 5.96373233795166,
      "epoch": 3.5251009809578764,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038032249767301716,
      "loss": 5.5204,
      "mean_token_accuracy": 0.2058837056159973,
      "num_tokens": 77413763.0,
      "step": 30545
    },
    {
      "entropy": 6.066835355758667,
      "epoch": 3.5256780150028852,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038028603160499783,
      "loss": 5.6438,
      "mean_token_accuracy": 0.19838371574878694,
      "num_tokens": 77427132.0,
      "step": 30550
    },
    {
      "entropy": 6.115768003463745,
      "epoch": 3.5262550490478937,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038024956199573514,
      "loss": 5.6428,
      "mean_token_accuracy": 0.2061450108885765,
      "num_tokens": 77439074.0,
      "step": 30555
    },
    {
      "entropy": 6.158787202835083,
      "epoch": 3.5268320830929025,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003802130888464557,
      "loss": 5.6896,
      "mean_token_accuracy": 0.19760521948337556,
      "num_tokens": 77452159.0,
      "step": 30560
    },
    {
      "entropy": 6.136129570007324,
      "epoch": 3.5274091171379114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003801766121583863,
      "loss": 5.6553,
      "mean_token_accuracy": 0.19396378546953202,
      "num_tokens": 77465139.0,
      "step": 30565
    },
    {
      "entropy": 6.06098313331604,
      "epoch": 3.52798615118292,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003801401319327538,
      "loss": 5.6353,
      "mean_token_accuracy": 0.2043432429432869,
      "num_tokens": 77478577.0,
      "step": 30570
    },
    {
      "entropy": 6.1209249019622805,
      "epoch": 3.5285631852279282,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038010364817078513,
      "loss": 5.6132,
      "mean_token_accuracy": 0.2048458054661751,
      "num_tokens": 77491352.0,
      "step": 30575
    },
    {
      "entropy": 6.040082406997681,
      "epoch": 3.529140219272937,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003800671608737076,
      "loss": 5.5735,
      "mean_token_accuracy": 0.2121508464217186,
      "num_tokens": 77503465.0,
      "step": 30580
    },
    {
      "entropy": 6.199835252761841,
      "epoch": 3.529717253317946,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003800306700427481,
      "loss": 5.7001,
      "mean_token_accuracy": 0.19505725353956221,
      "num_tokens": 77516558.0,
      "step": 30585
    },
    {
      "entropy": 6.047958326339722,
      "epoch": 3.5302942873629544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037999417567913436,
      "loss": 5.6151,
      "mean_token_accuracy": 0.19728896617889405,
      "num_tokens": 77530335.0,
      "step": 30590
    },
    {
      "entropy": 6.104905891418457,
      "epoch": 3.5308713214079632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003799576777840936,
      "loss": 5.6137,
      "mean_token_accuracy": 0.21162358224391936,
      "num_tokens": 77543996.0,
      "step": 30595
    },
    {
      "entropy": 6.1977808475494385,
      "epoch": 3.5314483554529716,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037992117635885354,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19997528791427613,
      "num_tokens": 77556585.0,
      "step": 30600
    },
    {
      "entropy": 6.133581829071045,
      "epoch": 3.5320253894979805,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003798846714046418,
      "loss": 5.6366,
      "mean_token_accuracy": 0.20247101336717604,
      "num_tokens": 77569254.0,
      "step": 30605
    },
    {
      "entropy": 6.115343332290649,
      "epoch": 3.532602423542989,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003798481629226862,
      "loss": 5.656,
      "mean_token_accuracy": 0.2043052703142166,
      "num_tokens": 77582308.0,
      "step": 30610
    },
    {
      "entropy": 6.081429147720337,
      "epoch": 3.533179457587998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003798116509142147,
      "loss": 5.6248,
      "mean_token_accuracy": 0.2031274288892746,
      "num_tokens": 77596161.0,
      "step": 30615
    },
    {
      "entropy": 6.148749732971192,
      "epoch": 3.533756491633006,
      "grad_norm": 0.875,
      "learning_rate": 0.00037977513538045544,
      "loss": 5.7282,
      "mean_token_accuracy": 0.19472491145133972,
      "num_tokens": 77608387.0,
      "step": 30620
    },
    {
      "entropy": 6.197894144058227,
      "epoch": 3.534333525678015,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037973861632263655,
      "loss": 5.668,
      "mean_token_accuracy": 0.20166428834199907,
      "num_tokens": 77620502.0,
      "step": 30625
    },
    {
      "entropy": 6.129067325592041,
      "epoch": 3.5349105597230235,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003797020937419863,
      "loss": 5.6979,
      "mean_token_accuracy": 0.1904497981071472,
      "num_tokens": 77633106.0,
      "step": 30630
    },
    {
      "entropy": 6.018396043777466,
      "epoch": 3.5354875937680323,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037966556763973313,
      "loss": 5.6115,
      "mean_token_accuracy": 0.2036408379673958,
      "num_tokens": 77644938.0,
      "step": 30635
    },
    {
      "entropy": 6.127925968170166,
      "epoch": 3.536064627813041,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003796290380171056,
      "loss": 5.6214,
      "mean_token_accuracy": 0.20307651460170745,
      "num_tokens": 77656192.0,
      "step": 30640
    },
    {
      "entropy": 6.057373189926148,
      "epoch": 3.5366416618580496,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003795925048753324,
      "loss": 5.5721,
      "mean_token_accuracy": 0.21015395671129228,
      "num_tokens": 77667845.0,
      "step": 30645
    },
    {
      "entropy": 6.098120546340942,
      "epoch": 3.537218695903058,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037955596821564216,
      "loss": 5.6624,
      "mean_token_accuracy": 0.2039669007062912,
      "num_tokens": 77679401.0,
      "step": 30650
    },
    {
      "entropy": 6.151260232925415,
      "epoch": 3.537795729948067,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003795194280392639,
      "loss": 5.6745,
      "mean_token_accuracy": 0.19547478407621383,
      "num_tokens": 77692223.0,
      "step": 30655
    },
    {
      "entropy": 6.0903863430023195,
      "epoch": 3.5383727639930758,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003794828843474266,
      "loss": 5.6113,
      "mean_token_accuracy": 0.19964486509561538,
      "num_tokens": 77703955.0,
      "step": 30660
    },
    {
      "entropy": 6.066116666793823,
      "epoch": 3.538949798038084,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003794463371413594,
      "loss": 5.6664,
      "mean_token_accuracy": 0.19972672015428544,
      "num_tokens": 77716861.0,
      "step": 30665
    },
    {
      "entropy": 6.104123973846436,
      "epoch": 3.539526832083093,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037940978642229153,
      "loss": 5.6402,
      "mean_token_accuracy": 0.20932407677173615,
      "num_tokens": 77728650.0,
      "step": 30670
    },
    {
      "entropy": 6.11107873916626,
      "epoch": 3.5401038661281015,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037937323219145233,
      "loss": 5.6551,
      "mean_token_accuracy": 0.1996922791004181,
      "num_tokens": 77740676.0,
      "step": 30675
    },
    {
      "entropy": 6.154971170425415,
      "epoch": 3.5406809001731103,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037933667445007123,
      "loss": 5.6433,
      "mean_token_accuracy": 0.19993229508399962,
      "num_tokens": 77753126.0,
      "step": 30680
    },
    {
      "entropy": 5.966343212127685,
      "epoch": 3.5412579342181187,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037930011319937815,
      "loss": 5.5073,
      "mean_token_accuracy": 0.21381933689117433,
      "num_tokens": 77766089.0,
      "step": 30685
    },
    {
      "entropy": 6.14495325088501,
      "epoch": 3.5418349682631276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037926354844060236,
      "loss": 5.7256,
      "mean_token_accuracy": 0.19190252125263213,
      "num_tokens": 77779171.0,
      "step": 30690
    },
    {
      "entropy": 6.198380994796753,
      "epoch": 3.542412002308136,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037922698017497393,
      "loss": 5.7077,
      "mean_token_accuracy": 0.19388759583234788,
      "num_tokens": 77792172.0,
      "step": 30695
    },
    {
      "entropy": 6.151600933074951,
      "epoch": 3.542989036353145,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003791904084037229,
      "loss": 5.7357,
      "mean_token_accuracy": 0.18923739343881607,
      "num_tokens": 77804629.0,
      "step": 30700
    },
    {
      "entropy": 6.086050176620484,
      "epoch": 3.5435660703981533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003791538331280792,
      "loss": 5.6417,
      "mean_token_accuracy": 0.2014468416571617,
      "num_tokens": 77816954.0,
      "step": 30705
    },
    {
      "entropy": 6.116134786605835,
      "epoch": 3.544143104443162,
      "grad_norm": 0.96875,
      "learning_rate": 0.000379117254349273,
      "loss": 5.6296,
      "mean_token_accuracy": 0.21035561561584473,
      "num_tokens": 77829192.0,
      "step": 30710
    },
    {
      "entropy": 6.170983219146729,
      "epoch": 3.544720138488171,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003790806720685347,
      "loss": 5.694,
      "mean_token_accuracy": 0.20258939564228057,
      "num_tokens": 77842529.0,
      "step": 30715
    },
    {
      "entropy": 6.049699831008911,
      "epoch": 3.5452971725331794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003790440862870947,
      "loss": 5.5656,
      "mean_token_accuracy": 0.20837958604097367,
      "num_tokens": 77855910.0,
      "step": 30720
    },
    {
      "entropy": 6.140993928909301,
      "epoch": 3.545874206578188,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037900749700618355,
      "loss": 5.5867,
      "mean_token_accuracy": 0.1982567086815834,
      "num_tokens": 77867321.0,
      "step": 30725
    },
    {
      "entropy": 6.04323182106018,
      "epoch": 3.5464512406231967,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037897090422703186,
      "loss": 5.5682,
      "mean_token_accuracy": 0.20740952342748642,
      "num_tokens": 77880022.0,
      "step": 30730
    },
    {
      "entropy": 6.0994984149932865,
      "epoch": 3.5470282746682056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037893430795087047,
      "loss": 5.6565,
      "mean_token_accuracy": 0.20117796957492828,
      "num_tokens": 77892903.0,
      "step": 30735
    },
    {
      "entropy": 6.085923957824707,
      "epoch": 3.547605308713214,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003788977081789302,
      "loss": 5.6054,
      "mean_token_accuracy": 0.2040314793586731,
      "num_tokens": 77905860.0,
      "step": 30740
    },
    {
      "entropy": 6.081289148330688,
      "epoch": 3.548182342758223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003788611049124422,
      "loss": 5.6399,
      "mean_token_accuracy": 0.1989624246954918,
      "num_tokens": 77918747.0,
      "step": 30745
    },
    {
      "entropy": 6.067815256118775,
      "epoch": 3.5487593768032313,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037882449815263757,
      "loss": 5.657,
      "mean_token_accuracy": 0.19605450332164764,
      "num_tokens": 77931575.0,
      "step": 30750
    },
    {
      "entropy": 6.14686131477356,
      "epoch": 3.54933641084824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003787878879007474,
      "loss": 5.7459,
      "mean_token_accuracy": 0.19474599808454512,
      "num_tokens": 77944281.0,
      "step": 30755
    },
    {
      "entropy": 6.19490385055542,
      "epoch": 3.5499134448932486,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037875127415800324,
      "loss": 5.6551,
      "mean_token_accuracy": 0.1992063343524933,
      "num_tokens": 77956416.0,
      "step": 30760
    },
    {
      "entropy": 6.08354115486145,
      "epoch": 3.5504904789382574,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037871465692563655,
      "loss": 5.6009,
      "mean_token_accuracy": 0.2053515985608101,
      "num_tokens": 77970420.0,
      "step": 30765
    },
    {
      "entropy": 6.011940383911133,
      "epoch": 3.5510675129832663,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037867803620487885,
      "loss": 5.5667,
      "mean_token_accuracy": 0.20466463565826415,
      "num_tokens": 77982752.0,
      "step": 30770
    },
    {
      "entropy": 6.1181739330291744,
      "epoch": 3.5516445470282747,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003786414119969619,
      "loss": 5.7109,
      "mean_token_accuracy": 0.19602808505296707,
      "num_tokens": 77995106.0,
      "step": 30775
    },
    {
      "entropy": 6.192669486999511,
      "epoch": 3.552221581073283,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003786047843031175,
      "loss": 5.689,
      "mean_token_accuracy": 0.1950067847967148,
      "num_tokens": 78007387.0,
      "step": 30780
    },
    {
      "entropy": 6.170553970336914,
      "epoch": 3.552798615118292,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00037856815312457777,
      "loss": 5.709,
      "mean_token_accuracy": 0.19518875032663346,
      "num_tokens": 78020301.0,
      "step": 30785
    },
    {
      "entropy": 6.0381241798400875,
      "epoch": 3.553375649163301,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037853151846257456,
      "loss": 5.5402,
      "mean_token_accuracy": 0.2079501673579216,
      "num_tokens": 78032909.0,
      "step": 30790
    },
    {
      "entropy": 6.11513876914978,
      "epoch": 3.5539526832083093,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037849488031834014,
      "loss": 5.705,
      "mean_token_accuracy": 0.19704343229532242,
      "num_tokens": 78045123.0,
      "step": 30795
    },
    {
      "entropy": 6.1120264530181885,
      "epoch": 3.5545297172533177,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037845823869310687,
      "loss": 5.6536,
      "mean_token_accuracy": 0.19778777062892913,
      "num_tokens": 78057399.0,
      "step": 30800
    },
    {
      "entropy": 6.171220541000366,
      "epoch": 3.5551067512983265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003784215935881071,
      "loss": 5.6992,
      "mean_token_accuracy": 0.19713576287031173,
      "num_tokens": 78069439.0,
      "step": 30805
    },
    {
      "entropy": 6.183669328689575,
      "epoch": 3.5556837853433354,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037838494500457346,
      "loss": 5.7341,
      "mean_token_accuracy": 0.19349171221256256,
      "num_tokens": 78081882.0,
      "step": 30810
    },
    {
      "entropy": 6.102444124221802,
      "epoch": 3.556260819388344,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037834829294373847,
      "loss": 5.6589,
      "mean_token_accuracy": 0.19830785393714906,
      "num_tokens": 78094914.0,
      "step": 30815
    },
    {
      "entropy": 6.135072708129883,
      "epoch": 3.5568378534333527,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000378311637406835,
      "loss": 5.7151,
      "mean_token_accuracy": 0.19131042510271073,
      "num_tokens": 78106700.0,
      "step": 30820
    },
    {
      "entropy": 6.095702028274536,
      "epoch": 3.557414887478361,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037827497839509596,
      "loss": 5.6731,
      "mean_token_accuracy": 0.20272047370672225,
      "num_tokens": 78120616.0,
      "step": 30825
    },
    {
      "entropy": 6.230276012420655,
      "epoch": 3.55799192152337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037823831590975434,
      "loss": 5.7252,
      "mean_token_accuracy": 0.19898563772439956,
      "num_tokens": 78133068.0,
      "step": 30830
    },
    {
      "entropy": 6.039736795425415,
      "epoch": 3.5585689555683784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003782016499520432,
      "loss": 5.606,
      "mean_token_accuracy": 0.20099491328001023,
      "num_tokens": 78146283.0,
      "step": 30835
    },
    {
      "entropy": 6.146065425872803,
      "epoch": 3.5591459896133872,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037816498052319574,
      "loss": 5.6758,
      "mean_token_accuracy": 0.19377929866313934,
      "num_tokens": 78159014.0,
      "step": 30840
    },
    {
      "entropy": 6.210922622680664,
      "epoch": 3.559723023658396,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037812830762444554,
      "loss": 5.6838,
      "mean_token_accuracy": 0.19790124148130417,
      "num_tokens": 78171710.0,
      "step": 30845
    },
    {
      "entropy": 6.07824296951294,
      "epoch": 3.5603000577034045,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003780916312570259,
      "loss": 5.6375,
      "mean_token_accuracy": 0.19316561967134477,
      "num_tokens": 78185193.0,
      "step": 30850
    },
    {
      "entropy": 6.027272033691406,
      "epoch": 3.560877091748413,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003780549514221705,
      "loss": 5.5578,
      "mean_token_accuracy": 0.2072715863585472,
      "num_tokens": 78198026.0,
      "step": 30855
    },
    {
      "entropy": 6.073961353302002,
      "epoch": 3.561454125793422,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037801826812111297,
      "loss": 5.6121,
      "mean_token_accuracy": 0.1993874117732048,
      "num_tokens": 78210301.0,
      "step": 30860
    },
    {
      "entropy": 6.105393362045288,
      "epoch": 3.5620311598384307,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003779815813550871,
      "loss": 5.7344,
      "mean_token_accuracy": 0.18953127712011336,
      "num_tokens": 78223778.0,
      "step": 30865
    },
    {
      "entropy": 6.124974966049194,
      "epoch": 3.562608193883439,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00037794489112532697,
      "loss": 5.6508,
      "mean_token_accuracy": 0.20686106383800507,
      "num_tokens": 78236262.0,
      "step": 30870
    },
    {
      "entropy": 6.092332649230957,
      "epoch": 3.563185227928448,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003779081974330665,
      "loss": 5.6781,
      "mean_token_accuracy": 0.19358438104391099,
      "num_tokens": 78249010.0,
      "step": 30875
    },
    {
      "entropy": 6.0862384796142575,
      "epoch": 3.5637622619734564,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00037787150027953996,
      "loss": 5.5807,
      "mean_token_accuracy": 0.2047438457608223,
      "num_tokens": 78260914.0,
      "step": 30880
    },
    {
      "entropy": 6.111524534225464,
      "epoch": 3.564339296018465,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003778347996659816,
      "loss": 5.7261,
      "mean_token_accuracy": 0.1966250240802765,
      "num_tokens": 78273266.0,
      "step": 30885
    },
    {
      "entropy": 6.180128479003907,
      "epoch": 3.5649163300634736,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003777980955936259,
      "loss": 5.6688,
      "mean_token_accuracy": 0.19904460459947587,
      "num_tokens": 78284816.0,
      "step": 30890
    },
    {
      "entropy": 6.0757588863372805,
      "epoch": 3.5654933641084825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037776138806370726,
      "loss": 5.6405,
      "mean_token_accuracy": 0.20568201690912247,
      "num_tokens": 78298126.0,
      "step": 30895
    },
    {
      "entropy": 6.093035888671875,
      "epoch": 3.566070398153491,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037772467707746036,
      "loss": 5.6245,
      "mean_token_accuracy": 0.2015541449189186,
      "num_tokens": 78309578.0,
      "step": 30900
    },
    {
      "entropy": 6.153381252288819,
      "epoch": 3.5666474321985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037768796263612,
      "loss": 5.6829,
      "mean_token_accuracy": 0.2004859685897827,
      "num_tokens": 78321920.0,
      "step": 30905
    },
    {
      "entropy": 6.10942440032959,
      "epoch": 3.567224466243508,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037765124474092095,
      "loss": 5.6633,
      "mean_token_accuracy": 0.2028588369488716,
      "num_tokens": 78335786.0,
      "step": 30910
    },
    {
      "entropy": 6.015158939361572,
      "epoch": 3.567801500288517,
      "grad_norm": 0.828125,
      "learning_rate": 0.00037761452339309836,
      "loss": 5.5211,
      "mean_token_accuracy": 0.21581538170576095,
      "num_tokens": 78349247.0,
      "step": 30915
    },
    {
      "entropy": 6.094883584976197,
      "epoch": 3.568378534333526,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003775777985938872,
      "loss": 5.6072,
      "mean_token_accuracy": 0.2047165647149086,
      "num_tokens": 78361942.0,
      "step": 30920
    },
    {
      "entropy": 6.116112995147705,
      "epoch": 3.5689555683785343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003775410703445228,
      "loss": 5.6696,
      "mean_token_accuracy": 0.19665875732898713,
      "num_tokens": 78374667.0,
      "step": 30925
    },
    {
      "entropy": 6.136280059814453,
      "epoch": 3.5695326024235428,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003775043386462404,
      "loss": 5.7255,
      "mean_token_accuracy": 0.19536905139684677,
      "num_tokens": 78387174.0,
      "step": 30930
    },
    {
      "entropy": 6.0984354496002195,
      "epoch": 3.5701096364685516,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003774676035002755,
      "loss": 5.6695,
      "mean_token_accuracy": 0.19262166172266007,
      "num_tokens": 78401032.0,
      "step": 30935
    },
    {
      "entropy": 6.1314390182495115,
      "epoch": 3.5706866705135605,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003774308649078636,
      "loss": 5.7591,
      "mean_token_accuracy": 0.19342711716890335,
      "num_tokens": 78413626.0,
      "step": 30940
    },
    {
      "entropy": 6.1283214569091795,
      "epoch": 3.571263704558569,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003773941228702405,
      "loss": 5.6279,
      "mean_token_accuracy": 0.20067551583051682,
      "num_tokens": 78426073.0,
      "step": 30945
    },
    {
      "entropy": 6.116126918792725,
      "epoch": 3.5718407386035778,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003773573773886419,
      "loss": 5.6541,
      "mean_token_accuracy": 0.19624051451683044,
      "num_tokens": 78439139.0,
      "step": 30950
    },
    {
      "entropy": 6.138286781311035,
      "epoch": 3.572417772648586,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00037732062846430376,
      "loss": 5.681,
      "mean_token_accuracy": 0.202020500600338,
      "num_tokens": 78452266.0,
      "step": 30955
    },
    {
      "entropy": 6.053314065933227,
      "epoch": 3.572994806693595,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003772838760984621,
      "loss": 5.5844,
      "mean_token_accuracy": 0.2030759036540985,
      "num_tokens": 78465828.0,
      "step": 30960
    },
    {
      "entropy": 6.166993618011475,
      "epoch": 3.5735718407386035,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003772471202923531,
      "loss": 5.7308,
      "mean_token_accuracy": 0.196023128926754,
      "num_tokens": 78478648.0,
      "step": 30965
    },
    {
      "entropy": 6.140810585021972,
      "epoch": 3.5741488747836123,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037721036104721296,
      "loss": 5.6955,
      "mean_token_accuracy": 0.19484125524759294,
      "num_tokens": 78490364.0,
      "step": 30970
    },
    {
      "entropy": 6.074029159545899,
      "epoch": 3.5747259088286207,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00037717359836427817,
      "loss": 5.6084,
      "mean_token_accuracy": 0.19940755367279053,
      "num_tokens": 78502375.0,
      "step": 30975
    },
    {
      "entropy": 6.125685834884644,
      "epoch": 3.5753029428736296,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003771368322447851,
      "loss": 5.6987,
      "mean_token_accuracy": 0.19677154272794722,
      "num_tokens": 78515138.0,
      "step": 30980
    },
    {
      "entropy": 6.094970655441284,
      "epoch": 3.575879976918638,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003771000626899704,
      "loss": 5.5526,
      "mean_token_accuracy": 0.2114349529147148,
      "num_tokens": 78526204.0,
      "step": 30985
    },
    {
      "entropy": 6.097543573379516,
      "epoch": 3.576457010963647,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037706328970107077,
      "loss": 5.6693,
      "mean_token_accuracy": 0.19401396363973616,
      "num_tokens": 78539014.0,
      "step": 30990
    },
    {
      "entropy": 6.110267591476441,
      "epoch": 3.5770340450086557,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037702651327932326,
      "loss": 5.6504,
      "mean_token_accuracy": 0.20237064808607103,
      "num_tokens": 78551404.0,
      "step": 30995
    },
    {
      "entropy": 6.13531904220581,
      "epoch": 3.577611079053664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003769897334259645,
      "loss": 5.6423,
      "mean_token_accuracy": 0.19947001785039903,
      "num_tokens": 78563934.0,
      "step": 31000
    },
    {
      "entropy": 6.088009357452393,
      "epoch": 3.5781881130986726,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003769529501422317,
      "loss": 5.6951,
      "mean_token_accuracy": 0.19454893469810486,
      "num_tokens": 78576869.0,
      "step": 31005
    },
    {
      "entropy": 6.166034984588623,
      "epoch": 3.5787651471436814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037691616342936215,
      "loss": 5.6157,
      "mean_token_accuracy": 0.20636184364557267,
      "num_tokens": 78589569.0,
      "step": 31010
    },
    {
      "entropy": 6.109597206115723,
      "epoch": 3.5793421811886903,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00037687937328859306,
      "loss": 5.6355,
      "mean_token_accuracy": 0.20117154270410537,
      "num_tokens": 78602241.0,
      "step": 31015
    },
    {
      "entropy": 5.994266605377197,
      "epoch": 3.5799192152336987,
      "grad_norm": 0.875,
      "learning_rate": 0.00037684257972116187,
      "loss": 5.5466,
      "mean_token_accuracy": 0.2096637010574341,
      "num_tokens": 78615508.0,
      "step": 31020
    },
    {
      "entropy": 6.063689804077148,
      "epoch": 3.5804962492787076,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000376805782728306,
      "loss": 5.5702,
      "mean_token_accuracy": 0.20972448140382766,
      "num_tokens": 78627825.0,
      "step": 31025
    },
    {
      "entropy": 6.10751519203186,
      "epoch": 3.581073283323716,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003767689823112633,
      "loss": 5.6622,
      "mean_token_accuracy": 0.19690069705247878,
      "num_tokens": 78640934.0,
      "step": 31030
    },
    {
      "entropy": 6.151818943023682,
      "epoch": 3.581650317368725,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003767321784712714,
      "loss": 5.7085,
      "mean_token_accuracy": 0.19243402034044266,
      "num_tokens": 78653189.0,
      "step": 31035
    },
    {
      "entropy": 6.146539926528931,
      "epoch": 3.5822273514137333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003766953712095682,
      "loss": 5.6837,
      "mean_token_accuracy": 0.19898529797792436,
      "num_tokens": 78665520.0,
      "step": 31040
    },
    {
      "entropy": 6.073436117172241,
      "epoch": 3.582804385458742,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003766585605273917,
      "loss": 5.5918,
      "mean_token_accuracy": 0.20371832549571992,
      "num_tokens": 78678617.0,
      "step": 31045
    },
    {
      "entropy": 6.134043169021607,
      "epoch": 3.583381419503751,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037662174642598,
      "loss": 5.6918,
      "mean_token_accuracy": 0.19946915209293364,
      "num_tokens": 78690267.0,
      "step": 31050
    },
    {
      "entropy": 6.096961545944214,
      "epoch": 3.5839584535487594,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037658492890657136,
      "loss": 5.6132,
      "mean_token_accuracy": 0.206666000187397,
      "num_tokens": 78702885.0,
      "step": 31055
    },
    {
      "entropy": 6.092486429214477,
      "epoch": 3.584535487593768,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037654810797040414,
      "loss": 5.5558,
      "mean_token_accuracy": 0.2064187854528427,
      "num_tokens": 78715511.0,
      "step": 31060
    },
    {
      "entropy": 6.14454836845398,
      "epoch": 3.5851125216387767,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003765112836187166,
      "loss": 5.6911,
      "mean_token_accuracy": 0.1931006669998169,
      "num_tokens": 78729020.0,
      "step": 31065
    },
    {
      "entropy": 6.101726770401001,
      "epoch": 3.5856895556837856,
      "grad_norm": 2.328125,
      "learning_rate": 0.0003764744558527476,
      "loss": 5.5702,
      "mean_token_accuracy": 0.2070610851049423,
      "num_tokens": 78742049.0,
      "step": 31070
    },
    {
      "entropy": 6.214226198196411,
      "epoch": 3.586266589728794,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037643762467373567,
      "loss": 5.763,
      "mean_token_accuracy": 0.19577471017837525,
      "num_tokens": 78753444.0,
      "step": 31075
    },
    {
      "entropy": 6.126230955123901,
      "epoch": 3.5868436237738024,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003764007900829196,
      "loss": 5.6398,
      "mean_token_accuracy": 0.19948194026947022,
      "num_tokens": 78765793.0,
      "step": 31080
    },
    {
      "entropy": 6.112858676910401,
      "epoch": 3.5874206578188113,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003763639520815383,
      "loss": 5.6104,
      "mean_token_accuracy": 0.20604192614555358,
      "num_tokens": 78777896.0,
      "step": 31085
    },
    {
      "entropy": 6.1368941307067875,
      "epoch": 3.58799769186382,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003763271106708309,
      "loss": 5.7312,
      "mean_token_accuracy": 0.19502220749855043,
      "num_tokens": 78789836.0,
      "step": 31090
    },
    {
      "entropy": 6.172865200042724,
      "epoch": 3.5885747259088285,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037629026585203634,
      "loss": 5.7335,
      "mean_token_accuracy": 0.1922583431005478,
      "num_tokens": 78803440.0,
      "step": 31095
    },
    {
      "entropy": 6.145515012741089,
      "epoch": 3.5891517599538374,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003762534176263941,
      "loss": 5.6746,
      "mean_token_accuracy": 0.19972286969423295,
      "num_tokens": 78815627.0,
      "step": 31100
    },
    {
      "entropy": 6.136574029922485,
      "epoch": 3.589728793998846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037621656599514335,
      "loss": 5.6348,
      "mean_token_accuracy": 0.20076872706413268,
      "num_tokens": 78828676.0,
      "step": 31105
    },
    {
      "entropy": 6.201172494888306,
      "epoch": 3.5903058280438547,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003761797109595238,
      "loss": 5.7091,
      "mean_token_accuracy": 0.1937038704752922,
      "num_tokens": 78841551.0,
      "step": 31110
    },
    {
      "entropy": 6.126094150543213,
      "epoch": 3.590882862088863,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003761428525207749,
      "loss": 5.6651,
      "mean_token_accuracy": 0.20048174262046814,
      "num_tokens": 78853855.0,
      "step": 31115
    },
    {
      "entropy": 6.168135690689087,
      "epoch": 3.591459896133872,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037610599068013635,
      "loss": 5.7135,
      "mean_token_accuracy": 0.19472835659980775,
      "num_tokens": 78865747.0,
      "step": 31120
    },
    {
      "entropy": 6.063004207611084,
      "epoch": 3.592036930178881,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003760691254388481,
      "loss": 5.5392,
      "mean_token_accuracy": 0.2051069974899292,
      "num_tokens": 78877965.0,
      "step": 31125
    },
    {
      "entropy": 6.178427696228027,
      "epoch": 3.5926139642238892,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003760322567981499,
      "loss": 5.7009,
      "mean_token_accuracy": 0.19913222789764404,
      "num_tokens": 78890152.0,
      "step": 31130
    },
    {
      "entropy": 6.078564071655274,
      "epoch": 3.5931909982688977,
      "grad_norm": 1.0,
      "learning_rate": 0.000375995384759282,
      "loss": 5.6685,
      "mean_token_accuracy": 0.20232728868722916,
      "num_tokens": 78903777.0,
      "step": 31135
    },
    {
      "entropy": 6.074064588546753,
      "epoch": 3.5937680323139065,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003759585093234845,
      "loss": 5.6579,
      "mean_token_accuracy": 0.19905036985874175,
      "num_tokens": 78916452.0,
      "step": 31140
    },
    {
      "entropy": 6.202279949188233,
      "epoch": 3.5943450663589154,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003759216304919977,
      "loss": 5.6549,
      "mean_token_accuracy": 0.2038138896226883,
      "num_tokens": 78928990.0,
      "step": 31145
    },
    {
      "entropy": 6.20055456161499,
      "epoch": 3.594922100403924,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037588474826606195,
      "loss": 5.8073,
      "mean_token_accuracy": 0.18546199500560762,
      "num_tokens": 78943068.0,
      "step": 31150
    },
    {
      "entropy": 6.096027088165283,
      "epoch": 3.595499134448932,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037584786264691783,
      "loss": 5.61,
      "mean_token_accuracy": 0.20543482005596161,
      "num_tokens": 78956572.0,
      "step": 31155
    },
    {
      "entropy": 6.1550483226776125,
      "epoch": 3.596076168493941,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037581097363580595,
      "loss": 5.7362,
      "mean_token_accuracy": 0.19378693252801896,
      "num_tokens": 78969339.0,
      "step": 31160
    },
    {
      "entropy": 6.17869086265564,
      "epoch": 3.59665320253895,
      "grad_norm": 1.0,
      "learning_rate": 0.00037577408123396703,
      "loss": 5.6656,
      "mean_token_accuracy": 0.2036919116973877,
      "num_tokens": 78981860.0,
      "step": 31165
    },
    {
      "entropy": 6.159994745254517,
      "epoch": 3.5972302365839584,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00037573718544264194,
      "loss": 5.6571,
      "mean_token_accuracy": 0.19783925414085388,
      "num_tokens": 78993184.0,
      "step": 31170
    },
    {
      "entropy": 6.07230396270752,
      "epoch": 3.597807270628967,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003757002862630717,
      "loss": 5.6408,
      "mean_token_accuracy": 0.198989836871624,
      "num_tokens": 79004836.0,
      "step": 31175
    },
    {
      "entropy": 6.12829852104187,
      "epoch": 3.5983843046739756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003756633836964974,
      "loss": 5.6089,
      "mean_token_accuracy": 0.20449527949094773,
      "num_tokens": 79016593.0,
      "step": 31180
    },
    {
      "entropy": 6.113247394561768,
      "epoch": 3.5989613387189845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003756264777441601,
      "loss": 5.6588,
      "mean_token_accuracy": 0.20088752508163452,
      "num_tokens": 79030150.0,
      "step": 31185
    },
    {
      "entropy": 6.154535484313965,
      "epoch": 3.599538372763993,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003755895684073012,
      "loss": 5.6677,
      "mean_token_accuracy": 0.19760429710149766,
      "num_tokens": 79042407.0,
      "step": 31190
    },
    {
      "entropy": 6.067593050003052,
      "epoch": 3.6001154068090018,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037555265568716214,
      "loss": 5.6681,
      "mean_token_accuracy": 0.2039575919508934,
      "num_tokens": 79055169.0,
      "step": 31195
    },
    {
      "entropy": 6.104705572128296,
      "epoch": 3.6006924408540106,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003755157395849845,
      "loss": 5.6724,
      "mean_token_accuracy": 0.20768189430236816,
      "num_tokens": 79069907.0,
      "step": 31200
    },
    {
      "entropy": 6.1781505107879635,
      "epoch": 3.601269474899019,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037547882010200985,
      "loss": 5.7283,
      "mean_token_accuracy": 0.2009938195347786,
      "num_tokens": 79082270.0,
      "step": 31205
    },
    {
      "entropy": 6.166623640060425,
      "epoch": 3.6018465089440275,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037544189723948005,
      "loss": 5.6681,
      "mean_token_accuracy": 0.20290228575468064,
      "num_tokens": 79094648.0,
      "step": 31210
    },
    {
      "entropy": 6.038410568237305,
      "epoch": 3.6024235429890363,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003754049709986368,
      "loss": 5.5687,
      "mean_token_accuracy": 0.2051565796136856,
      "num_tokens": 79106360.0,
      "step": 31215
    },
    {
      "entropy": 6.184214544296265,
      "epoch": 3.603000577034045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003753680413807223,
      "loss": 5.7577,
      "mean_token_accuracy": 0.19263286888599396,
      "num_tokens": 79120030.0,
      "step": 31220
    },
    {
      "entropy": 6.142088794708252,
      "epoch": 3.6035776110790536,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003753311083869786,
      "loss": 5.6498,
      "mean_token_accuracy": 0.20029440373182297,
      "num_tokens": 79132200.0,
      "step": 31225
    },
    {
      "entropy": 6.084656143188477,
      "epoch": 3.6041546451240625,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003752941720186479,
      "loss": 5.6411,
      "mean_token_accuracy": 0.2016658678650856,
      "num_tokens": 79144520.0,
      "step": 31230
    },
    {
      "entropy": 6.113428068161011,
      "epoch": 3.604731679169071,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037525723227697255,
      "loss": 5.5905,
      "mean_token_accuracy": 0.20646243691444396,
      "num_tokens": 79159203.0,
      "step": 31235
    },
    {
      "entropy": 6.121013355255127,
      "epoch": 3.6053087132140798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000375220289163195,
      "loss": 5.7295,
      "mean_token_accuracy": 0.19636738002300264,
      "num_tokens": 79171490.0,
      "step": 31240
    },
    {
      "entropy": 6.109932565689087,
      "epoch": 3.605885747259088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003751833426785578,
      "loss": 5.6564,
      "mean_token_accuracy": 0.2032413586974144,
      "num_tokens": 79184835.0,
      "step": 31245
    },
    {
      "entropy": 6.104979991912842,
      "epoch": 3.606462781304097,
      "grad_norm": 1.0,
      "learning_rate": 0.00037514639282430355,
      "loss": 5.6169,
      "mean_token_accuracy": 0.20019723773002623,
      "num_tokens": 79197554.0,
      "step": 31250
    },
    {
      "entropy": 6.168771028518677,
      "epoch": 3.6070398153491054,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003751094396016752,
      "loss": 5.6995,
      "mean_token_accuracy": 0.20213110893964767,
      "num_tokens": 79212034.0,
      "step": 31255
    },
    {
      "entropy": 6.129372692108154,
      "epoch": 3.6076168493941143,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003750724830119156,
      "loss": 5.6832,
      "mean_token_accuracy": 0.19807273298501968,
      "num_tokens": 79225180.0,
      "step": 31260
    },
    {
      "entropy": 6.145452785491943,
      "epoch": 3.6081938834391227,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003750355230562677,
      "loss": 5.7132,
      "mean_token_accuracy": 0.19505878388881684,
      "num_tokens": 79238447.0,
      "step": 31265
    },
    {
      "entropy": 6.029267930984497,
      "epoch": 3.6087709174841316,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00037499855973597467,
      "loss": 5.5644,
      "mean_token_accuracy": 0.21064301282167436,
      "num_tokens": 79251636.0,
      "step": 31270
    },
    {
      "entropy": 6.031064510345459,
      "epoch": 3.6093479515291405,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003749615930522798,
      "loss": 5.5616,
      "mean_token_accuracy": 0.21446867138147355,
      "num_tokens": 79264555.0,
      "step": 31275
    },
    {
      "entropy": 6.08279275894165,
      "epoch": 3.609924985574149,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003749246230064264,
      "loss": 5.6043,
      "mean_token_accuracy": 0.20554226785898208,
      "num_tokens": 79277224.0,
      "step": 31280
    },
    {
      "entropy": 6.154872608184815,
      "epoch": 3.6105020196191573,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003748876495996579,
      "loss": 5.6278,
      "mean_token_accuracy": 0.19912316352128984,
      "num_tokens": 79289411.0,
      "step": 31285
    },
    {
      "entropy": 6.084700632095337,
      "epoch": 3.611079053664166,
      "grad_norm": 0.90625,
      "learning_rate": 0.00037485067283321794,
      "loss": 5.6332,
      "mean_token_accuracy": 0.20061580538749696,
      "num_tokens": 79302692.0,
      "step": 31290
    },
    {
      "entropy": 5.981883335113525,
      "epoch": 3.611656087709175,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037481369270835014,
      "loss": 5.5458,
      "mean_token_accuracy": 0.21054453253746033,
      "num_tokens": 79316309.0,
      "step": 31295
    },
    {
      "entropy": 6.198971366882324,
      "epoch": 3.6122331217541834,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00037477670922629846,
      "loss": 5.7164,
      "mean_token_accuracy": 0.19782219529151918,
      "num_tokens": 79330595.0,
      "step": 31300
    },
    {
      "entropy": 6.192024183273316,
      "epoch": 3.6128101557991923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037473972238830674,
      "loss": 5.7029,
      "mean_token_accuracy": 0.19738591760396956,
      "num_tokens": 79342539.0,
      "step": 31305
    },
    {
      "entropy": 5.916602087020874,
      "epoch": 3.6133871898442007,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00037470273219561895,
      "loss": 5.4534,
      "mean_token_accuracy": 0.21594885140657424,
      "num_tokens": 79355566.0,
      "step": 31310
    },
    {
      "entropy": 6.106811761856079,
      "epoch": 3.6139642238892096,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037466573864947925,
      "loss": 5.6467,
      "mean_token_accuracy": 0.19952652752399444,
      "num_tokens": 79368500.0,
      "step": 31315
    },
    {
      "entropy": 6.128374910354614,
      "epoch": 3.614541257934218,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037462874175113193,
      "loss": 5.646,
      "mean_token_accuracy": 0.19766153693199157,
      "num_tokens": 79380558.0,
      "step": 31320
    },
    {
      "entropy": 6.0616683006286625,
      "epoch": 3.615118291979227,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003745917415018214,
      "loss": 5.6455,
      "mean_token_accuracy": 0.20524810403585433,
      "num_tokens": 79393590.0,
      "step": 31325
    },
    {
      "entropy": 6.130316066741943,
      "epoch": 3.6156953260242353,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00037455473790279223,
      "loss": 5.6335,
      "mean_token_accuracy": 0.1942039117217064,
      "num_tokens": 79405403.0,
      "step": 31330
    },
    {
      "entropy": 6.099060583114624,
      "epoch": 3.616272360069244,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003745177309552887,
      "loss": 5.6684,
      "mean_token_accuracy": 0.19182250499725342,
      "num_tokens": 79417867.0,
      "step": 31335
    },
    {
      "entropy": 5.975255060195923,
      "epoch": 3.6168493941142525,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003744807206605558,
      "loss": 5.4463,
      "mean_token_accuracy": 0.2217454880475998,
      "num_tokens": 79431355.0,
      "step": 31340
    },
    {
      "entropy": 6.097650623321533,
      "epoch": 3.6174264281592614,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003744437070198383,
      "loss": 5.6256,
      "mean_token_accuracy": 0.19868793189525605,
      "num_tokens": 79443355.0,
      "step": 31345
    },
    {
      "entropy": 6.163055086135865,
      "epoch": 3.6180034622042703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003744066900343812,
      "loss": 5.7217,
      "mean_token_accuracy": 0.19222093373537064,
      "num_tokens": 79457082.0,
      "step": 31350
    },
    {
      "entropy": 6.184404182434082,
      "epoch": 3.6185804962492787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003743696697054293,
      "loss": 5.6926,
      "mean_token_accuracy": 0.20261137932538986,
      "num_tokens": 79469529.0,
      "step": 31355
    },
    {
      "entropy": 6.102014017105103,
      "epoch": 3.619157530294287,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000374332646034228,
      "loss": 5.617,
      "mean_token_accuracy": 0.20069116950035096,
      "num_tokens": 79481994.0,
      "step": 31360
    },
    {
      "entropy": 6.175708436965943,
      "epoch": 3.619734564339296,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037429561902202246,
      "loss": 5.7504,
      "mean_token_accuracy": 0.18803782314062117,
      "num_tokens": 79496813.0,
      "step": 31365
    },
    {
      "entropy": 6.144582653045655,
      "epoch": 3.620311598384305,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037425858867005815,
      "loss": 5.6686,
      "mean_token_accuracy": 0.1948112055659294,
      "num_tokens": 79508898.0,
      "step": 31370
    },
    {
      "entropy": 6.175226783752441,
      "epoch": 3.6208886324293132,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003742215549795805,
      "loss": 5.6721,
      "mean_token_accuracy": 0.20321782380342485,
      "num_tokens": 79521026.0,
      "step": 31375
    },
    {
      "entropy": 6.166785764694214,
      "epoch": 3.621465666474322,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000374184517951835,
      "loss": 5.6464,
      "mean_token_accuracy": 0.20286843925714493,
      "num_tokens": 79532434.0,
      "step": 31380
    },
    {
      "entropy": 6.119183158874511,
      "epoch": 3.6220427005193305,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037414747758806764,
      "loss": 5.6938,
      "mean_token_accuracy": 0.19908654689788818,
      "num_tokens": 79545144.0,
      "step": 31385
    },
    {
      "entropy": 6.1015989780426025,
      "epoch": 3.6226197345643394,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037411043388952417,
      "loss": 5.6543,
      "mean_token_accuracy": 0.19497503638267516,
      "num_tokens": 79557266.0,
      "step": 31390
    },
    {
      "entropy": 6.11690731048584,
      "epoch": 3.623196768609348,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003740733868574504,
      "loss": 5.6307,
      "mean_token_accuracy": 0.19708458185195923,
      "num_tokens": 79571091.0,
      "step": 31395
    },
    {
      "entropy": 6.059296703338623,
      "epoch": 3.6237738026543567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003740363364930925,
      "loss": 5.5911,
      "mean_token_accuracy": 0.20651842206716536,
      "num_tokens": 79584151.0,
      "step": 31400
    },
    {
      "entropy": 6.0477519035339355,
      "epoch": 3.6243508366993655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003739992827976966,
      "loss": 5.6074,
      "mean_token_accuracy": 0.20339070856571198,
      "num_tokens": 79596110.0,
      "step": 31405
    },
    {
      "entropy": 6.184017133712769,
      "epoch": 3.624927870744374,
      "grad_norm": 0.953125,
      "learning_rate": 0.000373962225772509,
      "loss": 5.6691,
      "mean_token_accuracy": 0.1967795044183731,
      "num_tokens": 79608559.0,
      "step": 31410
    },
    {
      "entropy": 6.211363792419434,
      "epoch": 3.6255049047893824,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037392516541877606,
      "loss": 5.6728,
      "mean_token_accuracy": 0.2042370244860649,
      "num_tokens": 79621267.0,
      "step": 31415
    },
    {
      "entropy": 6.085593748092651,
      "epoch": 3.6260819388343912,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003738881017377444,
      "loss": 5.6292,
      "mean_token_accuracy": 0.19884878695011138,
      "num_tokens": 79633282.0,
      "step": 31420
    },
    {
      "entropy": 6.102064180374145,
      "epoch": 3.6266589728794,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003738510347306605,
      "loss": 5.651,
      "mean_token_accuracy": 0.20146681368350983,
      "num_tokens": 79644705.0,
      "step": 31425
    },
    {
      "entropy": 6.099654722213745,
      "epoch": 3.6272360069244085,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037381396439877115,
      "loss": 5.5985,
      "mean_token_accuracy": 0.2030884489417076,
      "num_tokens": 79657846.0,
      "step": 31430
    },
    {
      "entropy": 6.141608333587646,
      "epoch": 3.627813040969417,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003737768907433232,
      "loss": 5.6223,
      "mean_token_accuracy": 0.20401741117238997,
      "num_tokens": 79669753.0,
      "step": 31435
    },
    {
      "entropy": 6.084135389328003,
      "epoch": 3.628390075014426,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003737398137655636,
      "loss": 5.611,
      "mean_token_accuracy": 0.2058655336499214,
      "num_tokens": 79683812.0,
      "step": 31440
    },
    {
      "entropy": 6.0418041229248045,
      "epoch": 3.6289671090594346,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037370273346673945,
      "loss": 5.6166,
      "mean_token_accuracy": 0.20544273257255555,
      "num_tokens": 79697162.0,
      "step": 31445
    },
    {
      "entropy": 6.100403261184693,
      "epoch": 3.629544143104443,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003736656498480978,
      "loss": 5.6244,
      "mean_token_accuracy": 0.2104505330324173,
      "num_tokens": 79710195.0,
      "step": 31450
    },
    {
      "entropy": 6.04806227684021,
      "epoch": 3.630121177149452,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003736285629108861,
      "loss": 5.583,
      "mean_token_accuracy": 0.2008063316345215,
      "num_tokens": 79722127.0,
      "step": 31455
    },
    {
      "entropy": 6.110556697845459,
      "epoch": 3.6306982111944603,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037359147265635166,
      "loss": 5.6567,
      "mean_token_accuracy": 0.19124801456928253,
      "num_tokens": 79733360.0,
      "step": 31460
    },
    {
      "entropy": 6.197495937347412,
      "epoch": 3.631275245239469,
      "grad_norm": 0.8515625,
      "learning_rate": 0.000373554379085742,
      "loss": 5.7007,
      "mean_token_accuracy": 0.20026664882898332,
      "num_tokens": 79746948.0,
      "step": 31465
    },
    {
      "entropy": 6.004641771316528,
      "epoch": 3.6318522792844776,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003735172822003048,
      "loss": 5.551,
      "mean_token_accuracy": 0.21951310932636262,
      "num_tokens": 79759816.0,
      "step": 31470
    },
    {
      "entropy": 6.110378551483154,
      "epoch": 3.6324293133294865,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003734801820012878,
      "loss": 5.6079,
      "mean_token_accuracy": 0.19847784638404847,
      "num_tokens": 79772205.0,
      "step": 31475
    },
    {
      "entropy": 6.080461645126343,
      "epoch": 3.6330063473744953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037344307848993856,
      "loss": 5.6062,
      "mean_token_accuracy": 0.20924705266952515,
      "num_tokens": 79785886.0,
      "step": 31480
    },
    {
      "entropy": 6.084271717071533,
      "epoch": 3.6335833814195038,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037340597166750546,
      "loss": 5.6346,
      "mean_token_accuracy": 0.20435042530298234,
      "num_tokens": 79798223.0,
      "step": 31485
    },
    {
      "entropy": 6.123413944244385,
      "epoch": 3.634160415464512,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037336886153523633,
      "loss": 5.7318,
      "mean_token_accuracy": 0.20027197301387786,
      "num_tokens": 79812374.0,
      "step": 31490
    },
    {
      "entropy": 6.14485993385315,
      "epoch": 3.634737449509521,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003733317480943795,
      "loss": 5.6858,
      "mean_token_accuracy": 0.1963226854801178,
      "num_tokens": 79825406.0,
      "step": 31495
    },
    {
      "entropy": 6.153828144073486,
      "epoch": 3.63531448355453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000373294631346183,
      "loss": 5.659,
      "mean_token_accuracy": 0.2052447095513344,
      "num_tokens": 79839684.0,
      "step": 31500
    },
    {
      "entropy": 6.101085996627807,
      "epoch": 3.6358915175995383,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003732575112918955,
      "loss": 5.6744,
      "mean_token_accuracy": 0.20450015962123871,
      "num_tokens": 79851673.0,
      "step": 31505
    },
    {
      "entropy": 6.074370670318603,
      "epoch": 3.636468551644547,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003732203879327654,
      "loss": 5.5735,
      "mean_token_accuracy": 0.20676430612802504,
      "num_tokens": 79864996.0,
      "step": 31510
    },
    {
      "entropy": 6.201430368423462,
      "epoch": 3.6370455856895556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003731832612700413,
      "loss": 5.7188,
      "mean_token_accuracy": 0.19358039051294326,
      "num_tokens": 79877067.0,
      "step": 31515
    },
    {
      "entropy": 6.09938383102417,
      "epoch": 3.6376226197345645,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037314613130497203,
      "loss": 5.6407,
      "mean_token_accuracy": 0.2018402874469757,
      "num_tokens": 79889246.0,
      "step": 31520
    },
    {
      "entropy": 6.1251811504364015,
      "epoch": 3.638199653779573,
      "grad_norm": 1.203125,
      "learning_rate": 0.00037310899803880636,
      "loss": 5.7036,
      "mean_token_accuracy": 0.1957951843738556,
      "num_tokens": 79903020.0,
      "step": 31525
    },
    {
      "entropy": 6.089523792266846,
      "epoch": 3.6387766878245817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003730718614727934,
      "loss": 5.6061,
      "mean_token_accuracy": 0.2040889173746109,
      "num_tokens": 79914688.0,
      "step": 31530
    },
    {
      "entropy": 6.084633922576904,
      "epoch": 3.63935372186959,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037303472160818196,
      "loss": 5.5667,
      "mean_token_accuracy": 0.19837968945503234,
      "num_tokens": 79927489.0,
      "step": 31535
    },
    {
      "entropy": 6.089060115814209,
      "epoch": 3.639930755914599,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037299757844622134,
      "loss": 5.669,
      "mean_token_accuracy": 0.20063736885786057,
      "num_tokens": 79939760.0,
      "step": 31540
    },
    {
      "entropy": 6.131798601150512,
      "epoch": 3.6405077899596074,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003729604319881608,
      "loss": 5.6852,
      "mean_token_accuracy": 0.20525304675102235,
      "num_tokens": 79952651.0,
      "step": 31545
    },
    {
      "entropy": 6.07037672996521,
      "epoch": 3.6410848240046163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003729232822352499,
      "loss": 5.5902,
      "mean_token_accuracy": 0.20292763859033586,
      "num_tokens": 79965058.0,
      "step": 31550
    },
    {
      "entropy": 6.131277751922608,
      "epoch": 3.641661858049625,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037288612918873806,
      "loss": 5.6853,
      "mean_token_accuracy": 0.1946774885058403,
      "num_tokens": 79978167.0,
      "step": 31555
    },
    {
      "entropy": 6.111083507537842,
      "epoch": 3.6422388920946336,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00037284897284987483,
      "loss": 5.5665,
      "mean_token_accuracy": 0.21005533039569854,
      "num_tokens": 79991170.0,
      "step": 31560
    },
    {
      "entropy": 6.084380674362182,
      "epoch": 3.642815926139642,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037281181321991005,
      "loss": 5.6336,
      "mean_token_accuracy": 0.20630927234888077,
      "num_tokens": 80003220.0,
      "step": 31565
    },
    {
      "entropy": 5.968671894073486,
      "epoch": 3.643392960184651,
      "grad_norm": 1.0,
      "learning_rate": 0.00037277465030009343,
      "loss": 5.5643,
      "mean_token_accuracy": 0.2134526029229164,
      "num_tokens": 80015928.0,
      "step": 31570
    },
    {
      "entropy": 6.100400924682617,
      "epoch": 3.6439699942296597,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003727374840916751,
      "loss": 5.5705,
      "mean_token_accuracy": 0.21390258967876435,
      "num_tokens": 80028314.0,
      "step": 31575
    },
    {
      "entropy": 6.099554061889648,
      "epoch": 3.644547028274668,
      "grad_norm": 1.015625,
      "learning_rate": 0.000372700314595905,
      "loss": 5.5845,
      "mean_token_accuracy": 0.20813449323177338,
      "num_tokens": 80040581.0,
      "step": 31580
    },
    {
      "entropy": 6.104198884963989,
      "epoch": 3.645124062319677,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037266314181403336,
      "loss": 5.6477,
      "mean_token_accuracy": 0.20370962470769882,
      "num_tokens": 80052501.0,
      "step": 31585
    },
    {
      "entropy": 6.0481285572052,
      "epoch": 3.6457010963646854,
      "grad_norm": 1.15625,
      "learning_rate": 0.00037262596574731043,
      "loss": 5.4381,
      "mean_token_accuracy": 0.21739273220300676,
      "num_tokens": 80065692.0,
      "step": 31590
    },
    {
      "entropy": 6.01230754852295,
      "epoch": 3.6462781304096943,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037258878639698673,
      "loss": 5.5625,
      "mean_token_accuracy": 0.21352598071098328,
      "num_tokens": 80079033.0,
      "step": 31595
    },
    {
      "entropy": 6.015496683120728,
      "epoch": 3.6468551644547027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00037255160376431257,
      "loss": 5.6469,
      "mean_token_accuracy": 0.20038847327232362,
      "num_tokens": 80092362.0,
      "step": 31600
    },
    {
      "entropy": 6.121754503250122,
      "epoch": 3.6474321984997116,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037251441785053873,
      "loss": 5.6576,
      "mean_token_accuracy": 0.2037595272064209,
      "num_tokens": 80106094.0,
      "step": 31605
    },
    {
      "entropy": 6.077968549728394,
      "epoch": 3.64800923254472,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037247722865691585,
      "loss": 5.5859,
      "mean_token_accuracy": 0.21136993616819383,
      "num_tokens": 80120000.0,
      "step": 31610
    },
    {
      "entropy": 6.1014386177062985,
      "epoch": 3.648586266589729,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003724400361846948,
      "loss": 5.6064,
      "mean_token_accuracy": 0.20282262116670607,
      "num_tokens": 80131769.0,
      "step": 31615
    },
    {
      "entropy": 6.049617958068848,
      "epoch": 3.6491633006347373,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037240284043512663,
      "loss": 5.5564,
      "mean_token_accuracy": 0.21292811930179595,
      "num_tokens": 80143870.0,
      "step": 31620
    },
    {
      "entropy": 6.060082960128784,
      "epoch": 3.649740334679746,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003723656414094622,
      "loss": 5.5724,
      "mean_token_accuracy": 0.20353516042232514,
      "num_tokens": 80157753.0,
      "step": 31625
    },
    {
      "entropy": 6.075562238693237,
      "epoch": 3.650317368724755,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037232843910895276,
      "loss": 5.5855,
      "mean_token_accuracy": 0.20195764750242234,
      "num_tokens": 80170432.0,
      "step": 31630
    },
    {
      "entropy": 6.0409571647644045,
      "epoch": 3.6508944027697634,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003722912335348497,
      "loss": 5.6243,
      "mean_token_accuracy": 0.20404012352228165,
      "num_tokens": 80182855.0,
      "step": 31635
    },
    {
      "entropy": 6.110857391357422,
      "epoch": 3.651471436814772,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003722540246884043,
      "loss": 5.6566,
      "mean_token_accuracy": 0.2052586182951927,
      "num_tokens": 80195509.0,
      "step": 31640
    },
    {
      "entropy": 6.084488201141357,
      "epoch": 3.6520484708597807,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003722168125708681,
      "loss": 5.6234,
      "mean_token_accuracy": 0.20165701061487198,
      "num_tokens": 80208055.0,
      "step": 31645
    },
    {
      "entropy": 6.10513744354248,
      "epoch": 3.6526255049047895,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037217959718349264,
      "loss": 5.623,
      "mean_token_accuracy": 0.20423029214143754,
      "num_tokens": 80219653.0,
      "step": 31650
    },
    {
      "entropy": 6.163469839096069,
      "epoch": 3.653202538949798,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003721423785275298,
      "loss": 5.6164,
      "mean_token_accuracy": 0.20854012668132782,
      "num_tokens": 80232793.0,
      "step": 31655
    },
    {
      "entropy": 6.124817609786987,
      "epoch": 3.653779572994807,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003721051566042313,
      "loss": 5.6171,
      "mean_token_accuracy": 0.20256602317094802,
      "num_tokens": 80245237.0,
      "step": 31660
    },
    {
      "entropy": 6.075761890411377,
      "epoch": 3.6543566070398152,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000372067931414849,
      "loss": 5.5633,
      "mean_token_accuracy": 0.20456094443798065,
      "num_tokens": 80258934.0,
      "step": 31665
    },
    {
      "entropy": 6.164009284973145,
      "epoch": 3.654933641084824,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003720307029606351,
      "loss": 5.7022,
      "mean_token_accuracy": 0.19464025795459747,
      "num_tokens": 80271605.0,
      "step": 31670
    },
    {
      "entropy": 6.068563556671142,
      "epoch": 3.6555106751298325,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037199347124284167,
      "loss": 5.5203,
      "mean_token_accuracy": 0.2149454951286316,
      "num_tokens": 80284332.0,
      "step": 31675
    },
    {
      "entropy": 6.1339404582977295,
      "epoch": 3.6560877091748414,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0003719562362627211,
      "loss": 5.6466,
      "mean_token_accuracy": 0.1995929554104805,
      "num_tokens": 80297277.0,
      "step": 31680
    },
    {
      "entropy": 6.114528608322144,
      "epoch": 3.6566647432198502,
      "grad_norm": 0.921875,
      "learning_rate": 0.00037191899802152556,
      "loss": 5.7024,
      "mean_token_accuracy": 0.1973910540342331,
      "num_tokens": 80310915.0,
      "step": 31685
    },
    {
      "entropy": 6.118069076538086,
      "epoch": 3.6572417772648587,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037188175652050776,
      "loss": 5.6699,
      "mean_token_accuracy": 0.19662147015333176,
      "num_tokens": 80323041.0,
      "step": 31690
    },
    {
      "entropy": 6.114025974273682,
      "epoch": 3.657818811309867,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003718445117609202,
      "loss": 5.6346,
      "mean_token_accuracy": 0.2040783792734146,
      "num_tokens": 80335543.0,
      "step": 31695
    },
    {
      "entropy": 6.098688316345215,
      "epoch": 3.658395845354876,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037180726374401555,
      "loss": 5.6487,
      "mean_token_accuracy": 0.20411901772022248,
      "num_tokens": 80348362.0,
      "step": 31700
    },
    {
      "entropy": 6.075296688079834,
      "epoch": 3.658972879399885,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003717700124710467,
      "loss": 5.5031,
      "mean_token_accuracy": 0.21381102949380876,
      "num_tokens": 80361189.0,
      "step": 31705
    },
    {
      "entropy": 6.168242454528809,
      "epoch": 3.659549913444893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003717327579432665,
      "loss": 5.7004,
      "mean_token_accuracy": 0.20038793683052064,
      "num_tokens": 80373732.0,
      "step": 31710
    },
    {
      "entropy": 5.927784585952759,
      "epoch": 3.6601269474899016,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037169550016192805,
      "loss": 5.5107,
      "mean_token_accuracy": 0.21772821247577667,
      "num_tokens": 80388140.0,
      "step": 31715
    },
    {
      "entropy": 5.994295883178711,
      "epoch": 3.6607039815349105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037165823912828453,
      "loss": 5.5563,
      "mean_token_accuracy": 0.21009753793478012,
      "num_tokens": 80402085.0,
      "step": 31720
    },
    {
      "entropy": 6.139581871032715,
      "epoch": 3.6612810155799194,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003716209748435891,
      "loss": 5.6694,
      "mean_token_accuracy": 0.20428819954395294,
      "num_tokens": 80414433.0,
      "step": 31725
    },
    {
      "entropy": 6.054508543014526,
      "epoch": 3.661858049624928,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003715837073090952,
      "loss": 5.6241,
      "mean_token_accuracy": 0.20364027470350266,
      "num_tokens": 80427414.0,
      "step": 31730
    },
    {
      "entropy": 6.183839225769043,
      "epoch": 3.6624350836699366,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003715464365260563,
      "loss": 5.7128,
      "mean_token_accuracy": 0.20573733747005463,
      "num_tokens": 80441525.0,
      "step": 31735
    },
    {
      "entropy": 6.212820720672608,
      "epoch": 3.663012117714945,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003715091624957259,
      "loss": 5.7024,
      "mean_token_accuracy": 0.19779739081859588,
      "num_tokens": 80453555.0,
      "step": 31740
    },
    {
      "entropy": 6.091965055465698,
      "epoch": 3.663589151759954,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037147188521935784,
      "loss": 5.6264,
      "mean_token_accuracy": 0.20326377153396608,
      "num_tokens": 80466817.0,
      "step": 31745
    },
    {
      "entropy": 6.129201173782349,
      "epoch": 3.6641661858049623,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003714346046982058,
      "loss": 5.6564,
      "mean_token_accuracy": 0.20087906271219252,
      "num_tokens": 80477792.0,
      "step": 31750
    },
    {
      "entropy": 6.096215391159058,
      "epoch": 3.664743219849971,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003713973209335237,
      "loss": 5.5781,
      "mean_token_accuracy": 0.20512400269508363,
      "num_tokens": 80489913.0,
      "step": 31755
    },
    {
      "entropy": 6.001148509979248,
      "epoch": 3.66532025389498,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037136003392656565,
      "loss": 5.527,
      "mean_token_accuracy": 0.20822664499282836,
      "num_tokens": 80502361.0,
      "step": 31760
    },
    {
      "entropy": 6.0439900875091555,
      "epoch": 3.6658972879399885,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003713227436785858,
      "loss": 5.5834,
      "mean_token_accuracy": 0.21159652918577193,
      "num_tokens": 80515117.0,
      "step": 31765
    },
    {
      "entropy": 6.117327785491943,
      "epoch": 3.666474321984997,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003712854501908382,
      "loss": 5.6715,
      "mean_token_accuracy": 0.20520202070474625,
      "num_tokens": 80527726.0,
      "step": 31770
    },
    {
      "entropy": 6.110523128509522,
      "epoch": 3.6670513560300058,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003712481534645772,
      "loss": 5.6218,
      "mean_token_accuracy": 0.2014847904443741,
      "num_tokens": 80540863.0,
      "step": 31775
    },
    {
      "entropy": 5.997666931152343,
      "epoch": 3.6676283900750146,
      "grad_norm": 1.296875,
      "learning_rate": 0.00037121085350105755,
      "loss": 5.5299,
      "mean_token_accuracy": 0.21280371248722077,
      "num_tokens": 80553597.0,
      "step": 31780
    },
    {
      "entropy": 6.07890362739563,
      "epoch": 3.668205424120023,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003711735503015336,
      "loss": 5.6014,
      "mean_token_accuracy": 0.20825276523828506,
      "num_tokens": 80566644.0,
      "step": 31785
    },
    {
      "entropy": 6.170917558670044,
      "epoch": 3.668782458165032,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037113624386726004,
      "loss": 5.7298,
      "mean_token_accuracy": 0.19499072879552842,
      "num_tokens": 80579696.0,
      "step": 31790
    },
    {
      "entropy": 6.083804702758789,
      "epoch": 3.6693594922100403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003710989341994916,
      "loss": 5.5953,
      "mean_token_accuracy": 0.2050243452191353,
      "num_tokens": 80592794.0,
      "step": 31795
    },
    {
      "entropy": 6.066362810134888,
      "epoch": 3.669936526255049,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037106162129948333,
      "loss": 5.6143,
      "mean_token_accuracy": 0.19946397244930267,
      "num_tokens": 80606587.0,
      "step": 31800
    },
    {
      "entropy": 6.152547931671142,
      "epoch": 3.6705135603000576,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037102430516849016,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1996324509382248,
      "num_tokens": 80618912.0,
      "step": 31805
    },
    {
      "entropy": 6.0614057064056395,
      "epoch": 3.6710905943450665,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037098698580776714,
      "loss": 5.5849,
      "mean_token_accuracy": 0.2072437062859535,
      "num_tokens": 80631352.0,
      "step": 31810
    },
    {
      "entropy": 6.115705585479736,
      "epoch": 3.671667628390075,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003709496632185695,
      "loss": 5.6446,
      "mean_token_accuracy": 0.19653357863426207,
      "num_tokens": 80643430.0,
      "step": 31815
    },
    {
      "entropy": 6.082391166687012,
      "epoch": 3.6722446624350837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003709123374021527,
      "loss": 5.6185,
      "mean_token_accuracy": 0.21651341617107392,
      "num_tokens": 80657786.0,
      "step": 31820
    },
    {
      "entropy": 6.124112892150879,
      "epoch": 3.672821696480092,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000370875008359772,
      "loss": 5.6309,
      "mean_token_accuracy": 0.20375102907419204,
      "num_tokens": 80671131.0,
      "step": 31825
    },
    {
      "entropy": 6.10866150856018,
      "epoch": 3.673398730525101,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000370837676092683,
      "loss": 5.6525,
      "mean_token_accuracy": 0.20393079817295073,
      "num_tokens": 80686579.0,
      "step": 31830
    },
    {
      "entropy": 6.0974045753479,
      "epoch": 3.67397576457011,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037080034060214134,
      "loss": 5.6047,
      "mean_token_accuracy": 0.20704250931739807,
      "num_tokens": 80698280.0,
      "step": 31835
    },
    {
      "entropy": 6.072479867935181,
      "epoch": 3.6745527986151183,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037076300188940294,
      "loss": 5.5711,
      "mean_token_accuracy": 0.19920947402715683,
      "num_tokens": 80711057.0,
      "step": 31840
    },
    {
      "entropy": 6.137829828262329,
      "epoch": 3.6751298326601267,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037072565995572345,
      "loss": 5.6934,
      "mean_token_accuracy": 0.1974475786089897,
      "num_tokens": 80724085.0,
      "step": 31845
    },
    {
      "entropy": 6.129719829559326,
      "epoch": 3.6757068667051356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037068831480235897,
      "loss": 5.6965,
      "mean_token_accuracy": 0.20455853939056395,
      "num_tokens": 80736705.0,
      "step": 31850
    },
    {
      "entropy": 6.080997133255005,
      "epoch": 3.6762839007501444,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037065096643056544,
      "loss": 5.5986,
      "mean_token_accuracy": 0.19893429726362227,
      "num_tokens": 80749172.0,
      "step": 31855
    },
    {
      "entropy": 6.10428261756897,
      "epoch": 3.676860934795153,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037061361484159934,
      "loss": 5.607,
      "mean_token_accuracy": 0.2059983417391777,
      "num_tokens": 80762569.0,
      "step": 31860
    },
    {
      "entropy": 6.160522174835205,
      "epoch": 3.6774379688401617,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037057626003671667,
      "loss": 5.6538,
      "mean_token_accuracy": 0.20354978740215302,
      "num_tokens": 80775372.0,
      "step": 31865
    },
    {
      "entropy": 6.107165908813476,
      "epoch": 3.67801500288517,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037053890201717407,
      "loss": 5.6494,
      "mean_token_accuracy": 0.19876522868871688,
      "num_tokens": 80787026.0,
      "step": 31870
    },
    {
      "entropy": 6.018663787841797,
      "epoch": 3.678592036930179,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037050154078422786,
      "loss": 5.5878,
      "mean_token_accuracy": 0.20987798869609833,
      "num_tokens": 80799720.0,
      "step": 31875
    },
    {
      "entropy": 6.196094036102295,
      "epoch": 3.6791690709751874,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003704641763391348,
      "loss": 5.71,
      "mean_token_accuracy": 0.19274294525384902,
      "num_tokens": 80812487.0,
      "step": 31880
    },
    {
      "entropy": 6.091052913665772,
      "epoch": 3.6797461050201963,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003704268086831516,
      "loss": 5.5394,
      "mean_token_accuracy": 0.20878402888774872,
      "num_tokens": 80825001.0,
      "step": 31885
    },
    {
      "entropy": 6.052421951293946,
      "epoch": 3.6803231390652047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037038943781753515,
      "loss": 5.6115,
      "mean_token_accuracy": 0.2030135825276375,
      "num_tokens": 80838068.0,
      "step": 31890
    },
    {
      "entropy": 6.073292636871338,
      "epoch": 3.6809001731102136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037035206374354224,
      "loss": 5.6319,
      "mean_token_accuracy": 0.20013385266065598,
      "num_tokens": 80849474.0,
      "step": 31895
    },
    {
      "entropy": 6.156754493713379,
      "epoch": 3.681477207155222,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037031468646243004,
      "loss": 5.7455,
      "mean_token_accuracy": 0.19427785277366638,
      "num_tokens": 80863217.0,
      "step": 31900
    },
    {
      "entropy": 6.136698055267334,
      "epoch": 3.682054241200231,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037027730597545577,
      "loss": 5.6765,
      "mean_token_accuracy": 0.19656821191310883,
      "num_tokens": 80876706.0,
      "step": 31905
    },
    {
      "entropy": 6.2177637100219725,
      "epoch": 3.6826312752452397,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003702399222838766,
      "loss": 5.6494,
      "mean_token_accuracy": 0.1916962280869484,
      "num_tokens": 80888853.0,
      "step": 31910
    },
    {
      "entropy": 6.158915233612061,
      "epoch": 3.683208309290248,
      "grad_norm": 1.0,
      "learning_rate": 0.00037020253538895,
      "loss": 5.6672,
      "mean_token_accuracy": 0.20122171938419342,
      "num_tokens": 80900411.0,
      "step": 31915
    },
    {
      "entropy": 6.146502542495727,
      "epoch": 3.6837853433352565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037016514529193337,
      "loss": 5.6888,
      "mean_token_accuracy": 0.1991058737039566,
      "num_tokens": 80914093.0,
      "step": 31920
    },
    {
      "entropy": 6.09154109954834,
      "epoch": 3.6843623773802654,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00037012775199408434,
      "loss": 5.5246,
      "mean_token_accuracy": 0.21300326734781266,
      "num_tokens": 80926078.0,
      "step": 31925
    },
    {
      "entropy": 6.078769826889038,
      "epoch": 3.6849394114252743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00037009035549666065,
      "loss": 5.6248,
      "mean_token_accuracy": 0.20796998590230942,
      "num_tokens": 80939343.0,
      "step": 31930
    },
    {
      "entropy": 6.053972387313843,
      "epoch": 3.6855164454702827,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003700529558009201,
      "loss": 5.5209,
      "mean_token_accuracy": 0.21146876066923143,
      "num_tokens": 80951264.0,
      "step": 31935
    },
    {
      "entropy": 5.973317956924438,
      "epoch": 3.6860934795152915,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003700155529081205,
      "loss": 5.4417,
      "mean_token_accuracy": 0.21474826484918594,
      "num_tokens": 80964847.0,
      "step": 31940
    },
    {
      "entropy": 6.107014989852905,
      "epoch": 3.6866705135603,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036997814681952,
      "loss": 5.7559,
      "mean_token_accuracy": 0.1906362697482109,
      "num_tokens": 80976743.0,
      "step": 31945
    },
    {
      "entropy": 6.125425052642822,
      "epoch": 3.687247547605309,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003699407375363768,
      "loss": 5.6344,
      "mean_token_accuracy": 0.20296656787395478,
      "num_tokens": 80989116.0,
      "step": 31950
    },
    {
      "entropy": 6.127856349945068,
      "epoch": 3.6878245816503172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000369903325059949,
      "loss": 5.6371,
      "mean_token_accuracy": 0.20229786932468413,
      "num_tokens": 81001984.0,
      "step": 31955
    },
    {
      "entropy": 6.125501203536987,
      "epoch": 3.688401615695326,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036986590939149503,
      "loss": 5.628,
      "mean_token_accuracy": 0.20030225664377213,
      "num_tokens": 81013816.0,
      "step": 31960
    },
    {
      "entropy": 6.056718015670777,
      "epoch": 3.6889786497403345,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003698284905322732,
      "loss": 5.5715,
      "mean_token_accuracy": 0.21383861750364302,
      "num_tokens": 81027688.0,
      "step": 31965
    },
    {
      "entropy": 6.088495826721191,
      "epoch": 3.6895556837853434,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036979106848354236,
      "loss": 5.6118,
      "mean_token_accuracy": 0.20607208758592604,
      "num_tokens": 81039963.0,
      "step": 31970
    },
    {
      "entropy": 6.148860073089599,
      "epoch": 3.690132717830352,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036975364324656095,
      "loss": 5.659,
      "mean_token_accuracy": 0.19596734493970872,
      "num_tokens": 81053193.0,
      "step": 31975
    },
    {
      "entropy": 6.10787615776062,
      "epoch": 3.6907097518753607,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003697162148225878,
      "loss": 5.6756,
      "mean_token_accuracy": 0.20186924189329147,
      "num_tokens": 81065271.0,
      "step": 31980
    },
    {
      "entropy": 6.144801950454712,
      "epoch": 3.6912867859203695,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036967878321288183,
      "loss": 5.73,
      "mean_token_accuracy": 0.19592429101467132,
      "num_tokens": 81078183.0,
      "step": 31985
    },
    {
      "entropy": 6.148482847213745,
      "epoch": 3.691863819965378,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036964134841870207,
      "loss": 5.6445,
      "mean_token_accuracy": 0.2033967360854149,
      "num_tokens": 81090665.0,
      "step": 31990
    },
    {
      "entropy": 6.2074870586395265,
      "epoch": 3.6924408540103864,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036960391044130755,
      "loss": 5.7612,
      "mean_token_accuracy": 0.18606630861759185,
      "num_tokens": 81102834.0,
      "step": 31995
    },
    {
      "entropy": 6.060525369644165,
      "epoch": 3.693017888055395,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003695664692819574,
      "loss": 5.5456,
      "mean_token_accuracy": 0.2055351734161377,
      "num_tokens": 81116380.0,
      "step": 32000
    },
    {
      "entropy": 6.051888370513916,
      "epoch": 3.693594922100404,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036952902494191106,
      "loss": 5.5608,
      "mean_token_accuracy": 0.20717490017414092,
      "num_tokens": 81128304.0,
      "step": 32005
    },
    {
      "entropy": 5.976141023635864,
      "epoch": 3.6941719561454125,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036949157742242796,
      "loss": 5.5333,
      "mean_token_accuracy": 0.20976489931344985,
      "num_tokens": 81141162.0,
      "step": 32010
    },
    {
      "entropy": 6.136828327178955,
      "epoch": 3.6947489901904214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036945412672476763,
      "loss": 5.662,
      "mean_token_accuracy": 0.19366756081581116,
      "num_tokens": 81152690.0,
      "step": 32015
    },
    {
      "entropy": 6.094586896896362,
      "epoch": 3.6953260242354298,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003694166728501896,
      "loss": 5.6671,
      "mean_token_accuracy": 0.20372935235500336,
      "num_tokens": 81166778.0,
      "step": 32020
    },
    {
      "entropy": 6.108336639404297,
      "epoch": 3.6959030582804386,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003693792157999537,
      "loss": 5.5932,
      "mean_token_accuracy": 0.20055043399333955,
      "num_tokens": 81178551.0,
      "step": 32025
    },
    {
      "entropy": 6.094335746765137,
      "epoch": 3.696480092325447,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003693417555753197,
      "loss": 5.5842,
      "mean_token_accuracy": 0.21113443970680237,
      "num_tokens": 81191555.0,
      "step": 32030
    },
    {
      "entropy": 6.049908781051636,
      "epoch": 3.697057126370456,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036930429217754766,
      "loss": 5.5589,
      "mean_token_accuracy": 0.20537756830453874,
      "num_tokens": 81204181.0,
      "step": 32035
    },
    {
      "entropy": 6.191805458068847,
      "epoch": 3.6976341604154648,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003692668256078976,
      "loss": 5.7766,
      "mean_token_accuracy": 0.19429351389408112,
      "num_tokens": 81216307.0,
      "step": 32040
    },
    {
      "entropy": 6.16627459526062,
      "epoch": 3.698211194460473,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003692293558676297,
      "loss": 5.6404,
      "mean_token_accuracy": 0.2060305267572403,
      "num_tokens": 81228985.0,
      "step": 32045
    },
    {
      "entropy": 6.083999395370483,
      "epoch": 3.6987882285054816,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003691918829580042,
      "loss": 5.6404,
      "mean_token_accuracy": 0.20266406685113908,
      "num_tokens": 81242650.0,
      "step": 32050
    },
    {
      "entropy": 6.062539100646973,
      "epoch": 3.6993652625504905,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036915440688028154,
      "loss": 5.6173,
      "mean_token_accuracy": 0.20183532536029816,
      "num_tokens": 81255574.0,
      "step": 32055
    },
    {
      "entropy": 6.068630647659302,
      "epoch": 3.6999422965954993,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036911692763572215,
      "loss": 5.5375,
      "mean_token_accuracy": 0.20998241156339645,
      "num_tokens": 81268774.0,
      "step": 32060
    },
    {
      "entropy": 6.131548833847046,
      "epoch": 3.7005193306405078,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003690794452255866,
      "loss": 5.6378,
      "mean_token_accuracy": 0.20096139162778853,
      "num_tokens": 81281717.0,
      "step": 32065
    },
    {
      "entropy": 6.029264831542969,
      "epoch": 3.701096364685516,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003690419596511357,
      "loss": 5.6085,
      "mean_token_accuracy": 0.20901645570993424,
      "num_tokens": 81293697.0,
      "step": 32070
    },
    {
      "entropy": 6.166714859008789,
      "epoch": 3.701673398730525,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036900447091363017,
      "loss": 5.653,
      "mean_token_accuracy": 0.19878632575273514,
      "num_tokens": 81306728.0,
      "step": 32075
    },
    {
      "entropy": 6.086012887954712,
      "epoch": 3.702250432775534,
      "grad_norm": 0.984375,
      "learning_rate": 0.000368966979014331,
      "loss": 5.5343,
      "mean_token_accuracy": 0.20614220649003984,
      "num_tokens": 81317864.0,
      "step": 32080
    },
    {
      "entropy": 6.100617170333862,
      "epoch": 3.7028274668205423,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003689294839544991,
      "loss": 5.6902,
      "mean_token_accuracy": 0.19468800574541092,
      "num_tokens": 81332027.0,
      "step": 32085
    },
    {
      "entropy": 6.152664375305176,
      "epoch": 3.703404500865551,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036889198573539575,
      "loss": 5.6194,
      "mean_token_accuracy": 0.20684202909469604,
      "num_tokens": 81343472.0,
      "step": 32090
    },
    {
      "entropy": 6.142629384994507,
      "epoch": 3.7039815349105596,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036885448435828205,
      "loss": 5.6542,
      "mean_token_accuracy": 0.2024760514497757,
      "num_tokens": 81356023.0,
      "step": 32095
    },
    {
      "entropy": 5.9660240650177006,
      "epoch": 3.7045585689555685,
      "grad_norm": 0.875,
      "learning_rate": 0.0003688169798244194,
      "loss": 5.4978,
      "mean_token_accuracy": 0.22084505409002303,
      "num_tokens": 81369327.0,
      "step": 32100
    },
    {
      "entropy": 6.178499603271485,
      "epoch": 3.705135603000577,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003687794721350692,
      "loss": 5.7253,
      "mean_token_accuracy": 0.19795726239681244,
      "num_tokens": 81381470.0,
      "step": 32105
    },
    {
      "entropy": 6.144916629791259,
      "epoch": 3.7057126370455857,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036874196129149303,
      "loss": 5.5749,
      "mean_token_accuracy": 0.2026337668299675,
      "num_tokens": 81393764.0,
      "step": 32110
    },
    {
      "entropy": 6.1313048839569095,
      "epoch": 3.7062896710905946,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003687044472949526,
      "loss": 5.7123,
      "mean_token_accuracy": 0.19428895860910417,
      "num_tokens": 81405257.0,
      "step": 32115
    },
    {
      "entropy": 6.082985210418701,
      "epoch": 3.706866705135603,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003686669301467097,
      "loss": 5.6037,
      "mean_token_accuracy": 0.21281479597091674,
      "num_tokens": 81417938.0,
      "step": 32120
    },
    {
      "entropy": 6.133819055557251,
      "epoch": 3.7074437391806114,
      "grad_norm": 0.90625,
      "learning_rate": 0.000368629409848026,
      "loss": 5.5802,
      "mean_token_accuracy": 0.2060667559504509,
      "num_tokens": 81430183.0,
      "step": 32125
    },
    {
      "entropy": 6.029097986221314,
      "epoch": 3.7080207732256203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036859188640016365,
      "loss": 5.5732,
      "mean_token_accuracy": 0.20648828744888306,
      "num_tokens": 81441831.0,
      "step": 32130
    },
    {
      "entropy": 6.119962310791015,
      "epoch": 3.708597807270629,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003685543598043847,
      "loss": 5.7341,
      "mean_token_accuracy": 0.19169845879077912,
      "num_tokens": 81453327.0,
      "step": 32135
    },
    {
      "entropy": 6.1532786846160885,
      "epoch": 3.7091748413156376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003685168300619513,
      "loss": 5.6517,
      "mean_token_accuracy": 0.20710574686527253,
      "num_tokens": 81466462.0,
      "step": 32140
    },
    {
      "entropy": 6.135969352722168,
      "epoch": 3.7097518753606464,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036847929717412586,
      "loss": 5.5632,
      "mean_token_accuracy": 0.20706499069929124,
      "num_tokens": 81478768.0,
      "step": 32145
    },
    {
      "entropy": 6.101963472366333,
      "epoch": 3.710328909405655,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003684417611421706,
      "loss": 5.6502,
      "mean_token_accuracy": 0.1990034058690071,
      "num_tokens": 81492942.0,
      "step": 32150
    },
    {
      "entropy": 6.104153394699097,
      "epoch": 3.7109059434506637,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003684042219673481,
      "loss": 5.636,
      "mean_token_accuracy": 0.20014396011829377,
      "num_tokens": 81504010.0,
      "step": 32155
    },
    {
      "entropy": 6.106266021728516,
      "epoch": 3.711482977495672,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036836667965092106,
      "loss": 5.6078,
      "mean_token_accuracy": 0.2027534544467926,
      "num_tokens": 81517327.0,
      "step": 32160
    },
    {
      "entropy": 6.145915746688843,
      "epoch": 3.712060011540681,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036832913419415213,
      "loss": 5.6022,
      "mean_token_accuracy": 0.20093379616737367,
      "num_tokens": 81528712.0,
      "step": 32165
    },
    {
      "entropy": 5.9268005847930905,
      "epoch": 3.7126370455856894,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003682915855983041,
      "loss": 5.411,
      "mean_token_accuracy": 0.2226401910185814,
      "num_tokens": 81541406.0,
      "step": 32170
    },
    {
      "entropy": 6.095519542694092,
      "epoch": 3.7132140796306983,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036825403386464,
      "loss": 5.5704,
      "mean_token_accuracy": 0.20921113193035126,
      "num_tokens": 81554235.0,
      "step": 32175
    },
    {
      "entropy": 6.052953290939331,
      "epoch": 3.7137911136757067,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036821647899442267,
      "loss": 5.6269,
      "mean_token_accuracy": 0.20576815456151962,
      "num_tokens": 81566527.0,
      "step": 32180
    },
    {
      "entropy": 6.028816843032837,
      "epoch": 3.7143681477207156,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003681789209889155,
      "loss": 5.526,
      "mean_token_accuracy": 0.20969487279653548,
      "num_tokens": 81579093.0,
      "step": 32185
    },
    {
      "entropy": 6.089748191833496,
      "epoch": 3.7149451817657244,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036814135984938146,
      "loss": 5.6348,
      "mean_token_accuracy": 0.19932120740413667,
      "num_tokens": 81591907.0,
      "step": 32190
    },
    {
      "entropy": 6.158211755752563,
      "epoch": 3.715522215810733,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003681037955770842,
      "loss": 5.6779,
      "mean_token_accuracy": 0.20113901048898697,
      "num_tokens": 81604744.0,
      "step": 32195
    },
    {
      "entropy": 6.145216941833496,
      "epoch": 3.7160992498557412,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036806622817328697,
      "loss": 5.673,
      "mean_token_accuracy": 0.1975746050477028,
      "num_tokens": 81618122.0,
      "step": 32200
    },
    {
      "entropy": 6.092971324920654,
      "epoch": 3.71667628390075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036802865763925334,
      "loss": 5.658,
      "mean_token_accuracy": 0.19615527838468552,
      "num_tokens": 81630710.0,
      "step": 32205
    },
    {
      "entropy": 6.064430475234985,
      "epoch": 3.717253317945759,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036799108397624704,
      "loss": 5.5793,
      "mean_token_accuracy": 0.2079721987247467,
      "num_tokens": 81643629.0,
      "step": 32210
    },
    {
      "entropy": 6.131985902786255,
      "epoch": 3.7178303519907674,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003679535071855318,
      "loss": 5.7067,
      "mean_token_accuracy": 0.19480953812599183,
      "num_tokens": 81656489.0,
      "step": 32215
    },
    {
      "entropy": 6.185256242752075,
      "epoch": 3.7184073860357763,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003679159272683716,
      "loss": 5.6569,
      "mean_token_accuracy": 0.20028447955846787,
      "num_tokens": 81669767.0,
      "step": 32220
    },
    {
      "entropy": 6.1107412815094,
      "epoch": 3.7189844200807847,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003678783442260303,
      "loss": 5.5728,
      "mean_token_accuracy": 0.21131436973810197,
      "num_tokens": 81682404.0,
      "step": 32225
    },
    {
      "entropy": 6.067472362518311,
      "epoch": 3.7195614541257935,
      "grad_norm": 0.875,
      "learning_rate": 0.00036784075805977194,
      "loss": 5.6601,
      "mean_token_accuracy": 0.20134028494358064,
      "num_tokens": 81694778.0,
      "step": 32230
    },
    {
      "entropy": 6.07591986656189,
      "epoch": 3.720138488170802,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003678031687708609,
      "loss": 5.6338,
      "mean_token_accuracy": 0.2068109005689621,
      "num_tokens": 81708823.0,
      "step": 32235
    },
    {
      "entropy": 5.955157327651977,
      "epoch": 3.720715522215811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003677655763605614,
      "loss": 5.4912,
      "mean_token_accuracy": 0.20995953381061555,
      "num_tokens": 81721558.0,
      "step": 32240
    },
    {
      "entropy": 6.091207122802734,
      "epoch": 3.7212925562608192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036772798083013774,
      "loss": 5.6482,
      "mean_token_accuracy": 0.202237831056118,
      "num_tokens": 81733170.0,
      "step": 32245
    },
    {
      "entropy": 6.07231297492981,
      "epoch": 3.721869590305828,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003676903821808545,
      "loss": 5.5973,
      "mean_token_accuracy": 0.20762179642915726,
      "num_tokens": 81747076.0,
      "step": 32250
    },
    {
      "entropy": 6.134697151184082,
      "epoch": 3.7224466243508365,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00036765278041397633,
      "loss": 5.6568,
      "mean_token_accuracy": 0.20537785291671753,
      "num_tokens": 81758908.0,
      "step": 32255
    },
    {
      "entropy": 6.059366750717163,
      "epoch": 3.7230236583958454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036761517553076794,
      "loss": 5.5451,
      "mean_token_accuracy": 0.21384347677230836,
      "num_tokens": 81772398.0,
      "step": 32260
    },
    {
      "entropy": 6.054136562347412,
      "epoch": 3.7236006924408542,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003675775675324941,
      "loss": 5.6192,
      "mean_token_accuracy": 0.20268476903438568,
      "num_tokens": 81784048.0,
      "step": 32265
    },
    {
      "entropy": 6.1055957794189455,
      "epoch": 3.7241777264858626,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036753995642041976,
      "loss": 5.6109,
      "mean_token_accuracy": 0.20258331000804902,
      "num_tokens": 81796301.0,
      "step": 32270
    },
    {
      "entropy": 6.1669464111328125,
      "epoch": 3.724754760530871,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003675023421958099,
      "loss": 5.7571,
      "mean_token_accuracy": 0.19046144783496857,
      "num_tokens": 81807897.0,
      "step": 32275
    },
    {
      "entropy": 6.1084295272827145,
      "epoch": 3.72533179457588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003674647248599298,
      "loss": 5.6064,
      "mean_token_accuracy": 0.20409170985221864,
      "num_tokens": 81820138.0,
      "step": 32280
    },
    {
      "entropy": 6.14130802154541,
      "epoch": 3.725908828620889,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003674271044140446,
      "loss": 5.6587,
      "mean_token_accuracy": 0.20303584337234498,
      "num_tokens": 81832147.0,
      "step": 32285
    },
    {
      "entropy": 6.120838594436646,
      "epoch": 3.726485862665897,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003673894808594196,
      "loss": 5.638,
      "mean_token_accuracy": 0.2116558387875557,
      "num_tokens": 81844714.0,
      "step": 32290
    },
    {
      "entropy": 6.150388288497925,
      "epoch": 3.727062896710906,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003673518541973203,
      "loss": 5.5673,
      "mean_token_accuracy": 0.2091709017753601,
      "num_tokens": 81857310.0,
      "step": 32295
    },
    {
      "entropy": 6.0021473407745365,
      "epoch": 3.7276399307559145,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036731422442901236,
      "loss": 5.5737,
      "mean_token_accuracy": 0.20878852307796478,
      "num_tokens": 81871441.0,
      "step": 32300
    },
    {
      "entropy": 6.019360828399658,
      "epoch": 3.7282169648009233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003672765915557613,
      "loss": 5.5562,
      "mean_token_accuracy": 0.2149242043495178,
      "num_tokens": 81883523.0,
      "step": 32305
    },
    {
      "entropy": 6.074365425109863,
      "epoch": 3.7287939988459318,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003672389555788328,
      "loss": 5.6582,
      "mean_token_accuracy": 0.20570338666439056,
      "num_tokens": 81896910.0,
      "step": 32310
    },
    {
      "entropy": 6.1055327415466305,
      "epoch": 3.7293710328909406,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036720131649949295,
      "loss": 5.6657,
      "mean_token_accuracy": 0.20025507658720015,
      "num_tokens": 81911205.0,
      "step": 32315
    },
    {
      "entropy": 6.09299144744873,
      "epoch": 3.7299480669359495,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003671636743190076,
      "loss": 5.6204,
      "mean_token_accuracy": 0.19980547279119493,
      "num_tokens": 81923556.0,
      "step": 32320
    },
    {
      "entropy": 6.098953056335449,
      "epoch": 3.730525100980958,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036712602903864286,
      "loss": 5.6434,
      "mean_token_accuracy": 0.20586923360824586,
      "num_tokens": 81935613.0,
      "step": 32325
    },
    {
      "entropy": 6.109227991104126,
      "epoch": 3.7311021350259663,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003670883806596649,
      "loss": 5.5783,
      "mean_token_accuracy": 0.20650579333305358,
      "num_tokens": 81947951.0,
      "step": 32330
    },
    {
      "entropy": 6.105522298812867,
      "epoch": 3.731679169070975,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003670507291833399,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1967855527997017,
      "num_tokens": 81960622.0,
      "step": 32335
    },
    {
      "entropy": 6.052056503295899,
      "epoch": 3.732256203115984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003670130746109344,
      "loss": 5.6375,
      "mean_token_accuracy": 0.20374275594949723,
      "num_tokens": 81974110.0,
      "step": 32340
    },
    {
      "entropy": 6.13268256187439,
      "epoch": 3.7328332371609925,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036697541694371487,
      "loss": 5.6309,
      "mean_token_accuracy": 0.20146786272525788,
      "num_tokens": 81987392.0,
      "step": 32345
    },
    {
      "entropy": 6.132433652877808,
      "epoch": 3.733410271206001,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003669377561829479,
      "loss": 5.6546,
      "mean_token_accuracy": 0.20318444520235063,
      "num_tokens": 81999751.0,
      "step": 32350
    },
    {
      "entropy": 6.148645305633545,
      "epoch": 3.7339873052510097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036690009232990016,
      "loss": 5.6802,
      "mean_token_accuracy": 0.20016422420740126,
      "num_tokens": 82013751.0,
      "step": 32355
    },
    {
      "entropy": 6.070813322067261,
      "epoch": 3.7345643392960186,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036686242538583845,
      "loss": 5.5573,
      "mean_token_accuracy": 0.20743968486785888,
      "num_tokens": 82026681.0,
      "step": 32360
    },
    {
      "entropy": 6.051439189910889,
      "epoch": 3.735141373341027,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003668247553520297,
      "loss": 5.5989,
      "mean_token_accuracy": 0.20208120942115784,
      "num_tokens": 82039624.0,
      "step": 32365
    },
    {
      "entropy": 6.084421634674072,
      "epoch": 3.735718407386036,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036678708222974096,
      "loss": 5.6735,
      "mean_token_accuracy": 0.20162639021873474,
      "num_tokens": 82052106.0,
      "step": 32370
    },
    {
      "entropy": 5.97212700843811,
      "epoch": 3.7362954414310443,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003667494060202392,
      "loss": 5.5307,
      "mean_token_accuracy": 0.207809779047966,
      "num_tokens": 82064647.0,
      "step": 32375
    },
    {
      "entropy": 6.106335926055908,
      "epoch": 3.736872475476053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036671172672479176,
      "loss": 5.6239,
      "mean_token_accuracy": 0.2098295956850052,
      "num_tokens": 82077242.0,
      "step": 32380
    },
    {
      "entropy": 6.081893062591552,
      "epoch": 3.7374495095210616,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000366674044344666,
      "loss": 5.62,
      "mean_token_accuracy": 0.20567358285188675,
      "num_tokens": 82090667.0,
      "step": 32385
    },
    {
      "entropy": 6.1361884593963625,
      "epoch": 3.7380265435660704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036663635888112923,
      "loss": 5.579,
      "mean_token_accuracy": 0.20831230282783508,
      "num_tokens": 82104303.0,
      "step": 32390
    },
    {
      "entropy": 6.086204147338867,
      "epoch": 3.7386035776110793,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036659867033544914,
      "loss": 5.5816,
      "mean_token_accuracy": 0.21155169755220413,
      "num_tokens": 82117398.0,
      "step": 32395
    },
    {
      "entropy": 6.118532514572143,
      "epoch": 3.7391806116560877,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003665609787088932,
      "loss": 5.5997,
      "mean_token_accuracy": 0.1960316300392151,
      "num_tokens": 82132716.0,
      "step": 32400
    },
    {
      "entropy": 6.103175687789917,
      "epoch": 3.739757645701096,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003665232840027292,
      "loss": 5.6407,
      "mean_token_accuracy": 0.20233808755874633,
      "num_tokens": 82145442.0,
      "step": 32405
    },
    {
      "entropy": 6.148085021972657,
      "epoch": 3.740334679746105,
      "grad_norm": 0.875,
      "learning_rate": 0.0003664855862182251,
      "loss": 5.6916,
      "mean_token_accuracy": 0.20155177414417266,
      "num_tokens": 82157802.0,
      "step": 32410
    },
    {
      "entropy": 5.986072540283203,
      "epoch": 3.740911713791114,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036644788535664867,
      "loss": 5.5148,
      "mean_token_accuracy": 0.2107320696115494,
      "num_tokens": 82170587.0,
      "step": 32415
    },
    {
      "entropy": 6.094866514205933,
      "epoch": 3.7414887478361223,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003664101814192681,
      "loss": 5.6424,
      "mean_token_accuracy": 0.2065053954720497,
      "num_tokens": 82183010.0,
      "step": 32420
    },
    {
      "entropy": 6.180253267288208,
      "epoch": 3.742065781881131,
      "grad_norm": 0.890625,
      "learning_rate": 0.00036637247440735146,
      "loss": 5.7195,
      "mean_token_accuracy": 0.19784776717424393,
      "num_tokens": 82195340.0,
      "step": 32425
    },
    {
      "entropy": 6.1490922451019285,
      "epoch": 3.7426428159261396,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000366334764322167,
      "loss": 5.6692,
      "mean_token_accuracy": 0.2010624051094055,
      "num_tokens": 82208489.0,
      "step": 32430
    },
    {
      "entropy": 6.157190895080566,
      "epoch": 3.7432198499711484,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036629705116498313,
      "loss": 5.669,
      "mean_token_accuracy": 0.20122681558132172,
      "num_tokens": 82220456.0,
      "step": 32435
    },
    {
      "entropy": 6.027271842956543,
      "epoch": 3.743796884016157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036625933493706833,
      "loss": 5.5141,
      "mean_token_accuracy": 0.2121756851673126,
      "num_tokens": 82233253.0,
      "step": 32440
    },
    {
      "entropy": 6.059979724884033,
      "epoch": 3.7443739180611657,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003662216156396911,
      "loss": 5.5843,
      "mean_token_accuracy": 0.2064562827348709,
      "num_tokens": 82246006.0,
      "step": 32445
    },
    {
      "entropy": 6.150467824935913,
      "epoch": 3.744950952106174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003661838932741201,
      "loss": 5.7059,
      "mean_token_accuracy": 0.20373200327157975,
      "num_tokens": 82258504.0,
      "step": 32450
    },
    {
      "entropy": 6.008761072158814,
      "epoch": 3.745527986151183,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003661461678416242,
      "loss": 5.4907,
      "mean_token_accuracy": 0.20857551395893098,
      "num_tokens": 82271388.0,
      "step": 32455
    },
    {
      "entropy": 6.131132125854492,
      "epoch": 3.7461050201961914,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003661084393434722,
      "loss": 5.6546,
      "mean_token_accuracy": 0.19991757720708847,
      "num_tokens": 82284092.0,
      "step": 32460
    },
    {
      "entropy": 6.052231788635254,
      "epoch": 3.7466820542412003,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003660707077809331,
      "loss": 5.5109,
      "mean_token_accuracy": 0.22182962745428086,
      "num_tokens": 82296470.0,
      "step": 32465
    },
    {
      "entropy": 6.0498615264892575,
      "epoch": 3.747259088286209,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000366032973155276,
      "loss": 5.664,
      "mean_token_accuracy": 0.20293652415275573,
      "num_tokens": 82310048.0,
      "step": 32470
    },
    {
      "entropy": 6.0782859325408936,
      "epoch": 3.7478361223312175,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003659952354677701,
      "loss": 5.5424,
      "mean_token_accuracy": 0.2100034087896347,
      "num_tokens": 82322618.0,
      "step": 32475
    },
    {
      "entropy": 6.119845294952393,
      "epoch": 3.748413156376226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003659574947196845,
      "loss": 5.609,
      "mean_token_accuracy": 0.19862225651741028,
      "num_tokens": 82335179.0,
      "step": 32480
    },
    {
      "entropy": 6.11025562286377,
      "epoch": 3.748990190421235,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003659197509122889,
      "loss": 5.6331,
      "mean_token_accuracy": 0.20117217004299165,
      "num_tokens": 82346807.0,
      "step": 32485
    },
    {
      "entropy": 6.035577821731567,
      "epoch": 3.7495672244662437,
      "grad_norm": 1.53125,
      "learning_rate": 0.00036588200404685263,
      "loss": 5.5832,
      "mean_token_accuracy": 0.20178569704294205,
      "num_tokens": 82358914.0,
      "step": 32490
    },
    {
      "entropy": 6.1075663566589355,
      "epoch": 3.750144258511252,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036584425412464525,
      "loss": 5.6186,
      "mean_token_accuracy": 0.19736004173755645,
      "num_tokens": 82371417.0,
      "step": 32495
    },
    {
      "entropy": 6.146180009841919,
      "epoch": 3.750721292556261,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036580650114693654,
      "loss": 5.5702,
      "mean_token_accuracy": 0.20541742891073228,
      "num_tokens": 82384013.0,
      "step": 32500
    },
    {
      "entropy": 6.180217361450195,
      "epoch": 3.7512983266012694,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003657687451149962,
      "loss": 5.6528,
      "mean_token_accuracy": 0.20355466306209563,
      "num_tokens": 82398546.0,
      "step": 32505
    },
    {
      "entropy": 6.053690719604492,
      "epoch": 3.7518753606462782,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036573098603009425,
      "loss": 5.5698,
      "mean_token_accuracy": 0.2176451340317726,
      "num_tokens": 82410321.0,
      "step": 32510
    },
    {
      "entropy": 6.042766523361206,
      "epoch": 3.7524523946912867,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003656932238935007,
      "loss": 5.5071,
      "mean_token_accuracy": 0.21227299869060517,
      "num_tokens": 82422463.0,
      "step": 32515
    },
    {
      "entropy": 6.133840274810791,
      "epoch": 3.7530294287362955,
      "grad_norm": 0.921875,
      "learning_rate": 0.00036565545870648555,
      "loss": 5.6877,
      "mean_token_accuracy": 0.19425924271345138,
      "num_tokens": 82434430.0,
      "step": 32520
    },
    {
      "entropy": 6.084829998016358,
      "epoch": 3.753606462781304,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036561769047031904,
      "loss": 5.5517,
      "mean_token_accuracy": 0.20324511528015138,
      "num_tokens": 82445731.0,
      "step": 32525
    },
    {
      "entropy": 6.2098170757293705,
      "epoch": 3.754183496826313,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003655799191862716,
      "loss": 5.7579,
      "mean_token_accuracy": 0.18895328491926194,
      "num_tokens": 82458894.0,
      "step": 32530
    },
    {
      "entropy": 6.085763072967529,
      "epoch": 3.7547605308713212,
      "grad_norm": 0.796875,
      "learning_rate": 0.00036554214485561357,
      "loss": 5.6517,
      "mean_token_accuracy": 0.20021233707666397,
      "num_tokens": 82472043.0,
      "step": 32535
    },
    {
      "entropy": 6.143430900573731,
      "epoch": 3.75533756491633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003655043674796155,
      "loss": 5.5601,
      "mean_token_accuracy": 0.20251185148954393,
      "num_tokens": 82483304.0,
      "step": 32540
    },
    {
      "entropy": 6.16084771156311,
      "epoch": 3.755914598961339,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003654665870595479,
      "loss": 5.6474,
      "mean_token_accuracy": 0.20687719732522963,
      "num_tokens": 82495902.0,
      "step": 32545
    },
    {
      "entropy": 6.153928470611572,
      "epoch": 3.7564916330063474,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00036542880359668155,
      "loss": 5.6609,
      "mean_token_accuracy": 0.2065740406513214,
      "num_tokens": 82508350.0,
      "step": 32550
    },
    {
      "entropy": 6.008171558380127,
      "epoch": 3.757068667051356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036539101709228745,
      "loss": 5.5566,
      "mean_token_accuracy": 0.21697248071432113,
      "num_tokens": 82521992.0,
      "step": 32555
    },
    {
      "entropy": 6.12149863243103,
      "epoch": 3.7576457010963646,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036535322754763634,
      "loss": 5.6703,
      "mean_token_accuracy": 0.19976507723331452,
      "num_tokens": 82534408.0,
      "step": 32560
    },
    {
      "entropy": 6.109060573577881,
      "epoch": 3.7582227351413735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036531543496399923,
      "loss": 5.5734,
      "mean_token_accuracy": 0.21320396661758423,
      "num_tokens": 82547372.0,
      "step": 32565
    },
    {
      "entropy": 5.975702953338623,
      "epoch": 3.758799769186382,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036527763934264735,
      "loss": 5.4823,
      "mean_token_accuracy": 0.22043978422880173,
      "num_tokens": 82559335.0,
      "step": 32570
    },
    {
      "entropy": 6.161577606201172,
      "epoch": 3.759376803231391,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000365239840684852,
      "loss": 5.658,
      "mean_token_accuracy": 0.1953011304140091,
      "num_tokens": 82571420.0,
      "step": 32575
    },
    {
      "entropy": 6.117753601074218,
      "epoch": 3.759953837276399,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003652020389918844,
      "loss": 5.6144,
      "mean_token_accuracy": 0.20099452883005142,
      "num_tokens": 82585105.0,
      "step": 32580
    },
    {
      "entropy": 6.046821117401123,
      "epoch": 3.760530871321408,
      "grad_norm": 1.0,
      "learning_rate": 0.00036516423426501593,
      "loss": 5.5997,
      "mean_token_accuracy": 0.20807549208402634,
      "num_tokens": 82597305.0,
      "step": 32585
    },
    {
      "entropy": 6.084345006942749,
      "epoch": 3.7611079053664165,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003651264265055183,
      "loss": 5.5154,
      "mean_token_accuracy": 0.21751891523599626,
      "num_tokens": 82610779.0,
      "step": 32590
    },
    {
      "entropy": 6.110610914230347,
      "epoch": 3.7616849394114253,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036508861571466313,
      "loss": 5.6102,
      "mean_token_accuracy": 0.20716689229011537,
      "num_tokens": 82624152.0,
      "step": 32595
    },
    {
      "entropy": 6.064979410171508,
      "epoch": 3.762261973456434,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036505080189372206,
      "loss": 5.6117,
      "mean_token_accuracy": 0.2089684158563614,
      "num_tokens": 82637638.0,
      "step": 32600
    },
    {
      "entropy": 6.0444416999816895,
      "epoch": 3.7628390075014426,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000365012985043967,
      "loss": 5.5543,
      "mean_token_accuracy": 0.20645839273929595,
      "num_tokens": 82650665.0,
      "step": 32605
    },
    {
      "entropy": 6.099220132827758,
      "epoch": 3.763416041546451,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003649751651666699,
      "loss": 5.6003,
      "mean_token_accuracy": 0.2077594816684723,
      "num_tokens": 82663317.0,
      "step": 32610
    },
    {
      "entropy": 6.048624086380005,
      "epoch": 3.76399307559146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003649373422631028,
      "loss": 5.575,
      "mean_token_accuracy": 0.21233842968940736,
      "num_tokens": 82676310.0,
      "step": 32615
    },
    {
      "entropy": 6.155058479309082,
      "epoch": 3.7645701096364688,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003648995163345379,
      "loss": 5.6596,
      "mean_token_accuracy": 0.20171172320842742,
      "num_tokens": 82687832.0,
      "step": 32620
    },
    {
      "entropy": 6.196661853790284,
      "epoch": 3.765147143681477,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003648616873822474,
      "loss": 5.731,
      "mean_token_accuracy": 0.19700943529605866,
      "num_tokens": 82700870.0,
      "step": 32625
    },
    {
      "entropy": 6.151900148391723,
      "epoch": 3.7657241777264856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036482385540750374,
      "loss": 5.6878,
      "mean_token_accuracy": 0.2043942928314209,
      "num_tokens": 82713141.0,
      "step": 32630
    },
    {
      "entropy": 6.095987129211426,
      "epoch": 3.7663012117714945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036478602041157925,
      "loss": 5.6258,
      "mean_token_accuracy": 0.20356874614953996,
      "num_tokens": 82725354.0,
      "step": 32635
    },
    {
      "entropy": 6.0637561798095705,
      "epoch": 3.7668782458165033,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036474818239574664,
      "loss": 5.5142,
      "mean_token_accuracy": 0.21121003180742265,
      "num_tokens": 82737728.0,
      "step": 32640
    },
    {
      "entropy": 6.044016695022583,
      "epoch": 3.7674552798615117,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003647103413612784,
      "loss": 5.6443,
      "mean_token_accuracy": 0.20572449564933776,
      "num_tokens": 82749262.0,
      "step": 32645
    },
    {
      "entropy": 6.005218839645385,
      "epoch": 3.7680323139065206,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036467249730944744,
      "loss": 5.5386,
      "mean_token_accuracy": 0.21305883377790452,
      "num_tokens": 82762622.0,
      "step": 32650
    },
    {
      "entropy": 6.13591685295105,
      "epoch": 3.768609347951529,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003646346502415266,
      "loss": 5.7109,
      "mean_token_accuracy": 0.19854780584573745,
      "num_tokens": 82774795.0,
      "step": 32655
    },
    {
      "entropy": 6.185804843902588,
      "epoch": 3.769186381996538,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036459680015878875,
      "loss": 5.6672,
      "mean_token_accuracy": 0.20425862967967987,
      "num_tokens": 82786437.0,
      "step": 32660
    },
    {
      "entropy": 6.187558364868164,
      "epoch": 3.7697634160415463,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003645589470625071,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19404049813747407,
      "num_tokens": 82798813.0,
      "step": 32665
    },
    {
      "entropy": 5.974389934539795,
      "epoch": 3.770340450086555,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003645210909539547,
      "loss": 5.5305,
      "mean_token_accuracy": 0.21113889217376708,
      "num_tokens": 82810765.0,
      "step": 32670
    },
    {
      "entropy": 6.078088092803955,
      "epoch": 3.770917484131564,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003644832318344049,
      "loss": 5.5524,
      "mean_token_accuracy": 0.21098872423171997,
      "num_tokens": 82824046.0,
      "step": 32675
    },
    {
      "entropy": 6.137503910064697,
      "epoch": 3.7714945181765724,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000364445369705131,
      "loss": 5.6642,
      "mean_token_accuracy": 0.20074114799499512,
      "num_tokens": 82837629.0,
      "step": 32680
    },
    {
      "entropy": 5.910543584823609,
      "epoch": 3.772071552221581,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036440750456740665,
      "loss": 5.4883,
      "mean_token_accuracy": 0.2203031063079834,
      "num_tokens": 82850921.0,
      "step": 32685
    },
    {
      "entropy": 6.110146474838257,
      "epoch": 3.7726485862665897,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003643696364225052,
      "loss": 5.6069,
      "mean_token_accuracy": 0.2026948794722557,
      "num_tokens": 82864114.0,
      "step": 32690
    },
    {
      "entropy": 6.149781513214111,
      "epoch": 3.7732256203115986,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003643317652717004,
      "loss": 5.6809,
      "mean_token_accuracy": 0.20298480242490768,
      "num_tokens": 82876384.0,
      "step": 32695
    },
    {
      "entropy": 6.064359569549561,
      "epoch": 3.773802654356607,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036429389111626614,
      "loss": 5.5679,
      "mean_token_accuracy": 0.2105186551809311,
      "num_tokens": 82887890.0,
      "step": 32700
    },
    {
      "entropy": 6.1342222690582275,
      "epoch": 3.7743796884016154,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036425601395747616,
      "loss": 5.6703,
      "mean_token_accuracy": 0.2015447735786438,
      "num_tokens": 82901033.0,
      "step": 32705
    },
    {
      "entropy": 6.057837057113647,
      "epoch": 3.7749567224466243,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036421813379660447,
      "loss": 5.5293,
      "mean_token_accuracy": 0.20858179032802582,
      "num_tokens": 82913562.0,
      "step": 32710
    },
    {
      "entropy": 6.201757526397705,
      "epoch": 3.775533756491633,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003641802506349251,
      "loss": 5.7605,
      "mean_token_accuracy": 0.19189680069684983,
      "num_tokens": 82926175.0,
      "step": 32715
    },
    {
      "entropy": 6.125349712371826,
      "epoch": 3.7761107905366416,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003641423644737124,
      "loss": 5.6589,
      "mean_token_accuracy": 0.2019471138715744,
      "num_tokens": 82937797.0,
      "step": 32720
    },
    {
      "entropy": 6.1520851135253904,
      "epoch": 3.7766878245816504,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036410447531424057,
      "loss": 5.6349,
      "mean_token_accuracy": 0.20505717992782593,
      "num_tokens": 82950554.0,
      "step": 32725
    },
    {
      "entropy": 6.164891242980957,
      "epoch": 3.777264858626659,
      "grad_norm": 1.09375,
      "learning_rate": 0.00036406658315778386,
      "loss": 5.6508,
      "mean_token_accuracy": 0.20125698447227477,
      "num_tokens": 82962378.0,
      "step": 32730
    },
    {
      "entropy": 6.136154890060425,
      "epoch": 3.7778418926716677,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036402868800561696,
      "loss": 5.7003,
      "mean_token_accuracy": 0.19722906947135926,
      "num_tokens": 82974510.0,
      "step": 32735
    },
    {
      "entropy": 6.022222185134888,
      "epoch": 3.778418926716676,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003639907898590143,
      "loss": 5.6466,
      "mean_token_accuracy": 0.20063384771347045,
      "num_tokens": 82988086.0,
      "step": 32740
    },
    {
      "entropy": 6.170497941970825,
      "epoch": 3.778995960761685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036395288871925066,
      "loss": 5.7086,
      "mean_token_accuracy": 0.19816850423812865,
      "num_tokens": 83000787.0,
      "step": 32745
    },
    {
      "entropy": 6.096740818023681,
      "epoch": 3.779572994806694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003639149845876008,
      "loss": 5.5825,
      "mean_token_accuracy": 0.21283321976661682,
      "num_tokens": 83013365.0,
      "step": 32750
    },
    {
      "entropy": 6.082970905303955,
      "epoch": 3.7801500288517023,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036387707746533956,
      "loss": 5.6463,
      "mean_token_accuracy": 0.20189248770475388,
      "num_tokens": 83025184.0,
      "step": 32755
    },
    {
      "entropy": 6.1701476097106935,
      "epoch": 3.7807270628967107,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036383916735374205,
      "loss": 5.6839,
      "mean_token_accuracy": 0.1987305536866188,
      "num_tokens": 83037490.0,
      "step": 32760
    },
    {
      "entropy": 6.159075736999512,
      "epoch": 3.7813040969417195,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036380125425408325,
      "loss": 5.61,
      "mean_token_accuracy": 0.19836454689502717,
      "num_tokens": 83049831.0,
      "step": 32765
    },
    {
      "entropy": 6.107798957824707,
      "epoch": 3.7818811309867284,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003637633381676383,
      "loss": 5.6009,
      "mean_token_accuracy": 0.21274805814027786,
      "num_tokens": 83063299.0,
      "step": 32770
    },
    {
      "entropy": 6.126123046875,
      "epoch": 3.782458165031737,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003637254190956826,
      "loss": 5.6562,
      "mean_token_accuracy": 0.20219721794128417,
      "num_tokens": 83076085.0,
      "step": 32775
    },
    {
      "entropy": 6.108048439025879,
      "epoch": 3.7830351990767457,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003636874970394915,
      "loss": 5.5945,
      "mean_token_accuracy": 0.20824381858110427,
      "num_tokens": 83089466.0,
      "step": 32780
    },
    {
      "entropy": 6.139750051498413,
      "epoch": 3.783612233121754,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036364957200034055,
      "loss": 5.6455,
      "mean_token_accuracy": 0.19942262321710585,
      "num_tokens": 83102208.0,
      "step": 32785
    },
    {
      "entropy": 6.032274913787842,
      "epoch": 3.784189267166763,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003636116439795052,
      "loss": 5.4957,
      "mean_token_accuracy": 0.21652564853429795,
      "num_tokens": 83115804.0,
      "step": 32790
    },
    {
      "entropy": 6.087881088256836,
      "epoch": 3.7847663012117714,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00036357371297826124,
      "loss": 5.6633,
      "mean_token_accuracy": 0.19774015694856645,
      "num_tokens": 83129599.0,
      "step": 32795
    },
    {
      "entropy": 6.076316213607788,
      "epoch": 3.7853433352567802,
      "grad_norm": 1.0,
      "learning_rate": 0.0003635357789978845,
      "loss": 5.6161,
      "mean_token_accuracy": 0.19634360522031785,
      "num_tokens": 83142019.0,
      "step": 32800
    },
    {
      "entropy": 6.12406964302063,
      "epoch": 3.7859203693017887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036349784203965074,
      "loss": 5.6794,
      "mean_token_accuracy": 0.19823502749204636,
      "num_tokens": 83155099.0,
      "step": 32805
    },
    {
      "entropy": 6.10956301689148,
      "epoch": 3.7864974033467975,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036345990210483604,
      "loss": 5.6104,
      "mean_token_accuracy": 0.21280012428760528,
      "num_tokens": 83168294.0,
      "step": 32810
    },
    {
      "entropy": 6.132806968688965,
      "epoch": 3.787074437391806,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003634219591947165,
      "loss": 5.5843,
      "mean_token_accuracy": 0.20933948010206221,
      "num_tokens": 83181453.0,
      "step": 32815
    },
    {
      "entropy": 6.208439016342163,
      "epoch": 3.787651471436815,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036338401331056827,
      "loss": 5.7389,
      "mean_token_accuracy": 0.19418902397155763,
      "num_tokens": 83193942.0,
      "step": 32820
    },
    {
      "entropy": 6.11561598777771,
      "epoch": 3.7882285054818237,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003633460644536676,
      "loss": 5.6703,
      "mean_token_accuracy": 0.20308802276849747,
      "num_tokens": 83206507.0,
      "step": 32825
    },
    {
      "entropy": 6.078601121902466,
      "epoch": 3.788805539526832,
      "grad_norm": 1.015625,
      "learning_rate": 0.000363308112625291,
      "loss": 5.548,
      "mean_token_accuracy": 0.2040304198861122,
      "num_tokens": 83219383.0,
      "step": 32830
    },
    {
      "entropy": 6.072515392303467,
      "epoch": 3.7893825735718405,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00036327015782671487,
      "loss": 5.6083,
      "mean_token_accuracy": 0.2036228060722351,
      "num_tokens": 83231178.0,
      "step": 32835
    },
    {
      "entropy": 6.126521015167237,
      "epoch": 3.7899596076168494,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036323220005921584,
      "loss": 5.6053,
      "mean_token_accuracy": 0.2092042863368988,
      "num_tokens": 83242804.0,
      "step": 32840
    },
    {
      "entropy": 6.132390451431275,
      "epoch": 3.790536641661858,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003631942393240706,
      "loss": 5.6412,
      "mean_token_accuracy": 0.2018760234117508,
      "num_tokens": 83255483.0,
      "step": 32845
    },
    {
      "entropy": 6.074113941192627,
      "epoch": 3.7911136757068666,
      "grad_norm": 1.0,
      "learning_rate": 0.0003631562756225559,
      "loss": 5.6031,
      "mean_token_accuracy": 0.20220310091972352,
      "num_tokens": 83268442.0,
      "step": 32850
    },
    {
      "entropy": 6.151972961425781,
      "epoch": 3.7916907097518755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036311830895594866,
      "loss": 5.7149,
      "mean_token_accuracy": 0.20267511457204818,
      "num_tokens": 83281089.0,
      "step": 32855
    },
    {
      "entropy": 6.024322032928467,
      "epoch": 3.792267743796884,
      "grad_norm": 0.875,
      "learning_rate": 0.0003630803393255259,
      "loss": 5.4699,
      "mean_token_accuracy": 0.2122853457927704,
      "num_tokens": 83293134.0,
      "step": 32860
    },
    {
      "entropy": 6.017717504501343,
      "epoch": 3.792844777841893,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036304236673256467,
      "loss": 5.5166,
      "mean_token_accuracy": 0.2070764809846878,
      "num_tokens": 83305717.0,
      "step": 32865
    },
    {
      "entropy": 6.128232383728028,
      "epoch": 3.793421811886901,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003630043911783422,
      "loss": 5.557,
      "mean_token_accuracy": 0.21078261882066726,
      "num_tokens": 83317411.0,
      "step": 32870
    },
    {
      "entropy": 6.079075860977173,
      "epoch": 3.79399884593191,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036296641266413566,
      "loss": 5.6663,
      "mean_token_accuracy": 0.205397766828537,
      "num_tokens": 83330245.0,
      "step": 32875
    },
    {
      "entropy": 6.045082092285156,
      "epoch": 3.7945758799769185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003629284311912226,
      "loss": 5.5792,
      "mean_token_accuracy": 0.20603713691234588,
      "num_tokens": 83342617.0,
      "step": 32880
    },
    {
      "entropy": 6.065017795562744,
      "epoch": 3.7951529140219273,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036289044676088045,
      "loss": 5.5379,
      "mean_token_accuracy": 0.21593087613582612,
      "num_tokens": 83354238.0,
      "step": 32885
    },
    {
      "entropy": 6.043313121795654,
      "epoch": 3.7957299480669358,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003628524593743867,
      "loss": 5.5122,
      "mean_token_accuracy": 0.20442054867744447,
      "num_tokens": 83366838.0,
      "step": 32890
    },
    {
      "entropy": 6.043250608444214,
      "epoch": 3.7963069821119446,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036281446903301917,
      "loss": 5.5633,
      "mean_token_accuracy": 0.20783358216285705,
      "num_tokens": 83378293.0,
      "step": 32895
    },
    {
      "entropy": 6.143496990203857,
      "epoch": 3.7968840161569535,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036277647573805554,
      "loss": 5.6184,
      "mean_token_accuracy": 0.20373995006084442,
      "num_tokens": 83391345.0,
      "step": 32900
    },
    {
      "entropy": 6.148879528045654,
      "epoch": 3.797461050201962,
      "grad_norm": 0.9375,
      "learning_rate": 0.00036273847949077384,
      "loss": 5.6735,
      "mean_token_accuracy": 0.1974882960319519,
      "num_tokens": 83403764.0,
      "step": 32905
    },
    {
      "entropy": 6.078379726409912,
      "epoch": 3.7980380842469703,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003627004802924519,
      "loss": 5.5828,
      "mean_token_accuracy": 0.20430790930986403,
      "num_tokens": 83416496.0,
      "step": 32910
    },
    {
      "entropy": 6.101625728607178,
      "epoch": 3.798615118291979,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036266247814436794,
      "loss": 5.5799,
      "mean_token_accuracy": 0.2013986587524414,
      "num_tokens": 83429915.0,
      "step": 32915
    },
    {
      "entropy": 6.1453032970428465,
      "epoch": 3.799192152336988,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036262447304779996,
      "loss": 5.6366,
      "mean_token_accuracy": 0.1979958161711693,
      "num_tokens": 83442195.0,
      "step": 32920
    },
    {
      "entropy": 6.194523000717163,
      "epoch": 3.7997691863819965,
      "grad_norm": 1.0,
      "learning_rate": 0.00036258646500402636,
      "loss": 5.7321,
      "mean_token_accuracy": 0.19568776041269303,
      "num_tokens": 83454105.0,
      "step": 32925
    },
    {
      "entropy": 6.082583379745484,
      "epoch": 3.8003462204270053,
      "grad_norm": 1.109375,
      "learning_rate": 0.00036254845401432556,
      "loss": 5.5765,
      "mean_token_accuracy": 0.21255692839622498,
      "num_tokens": 83466487.0,
      "step": 32930
    },
    {
      "entropy": 6.049877691268921,
      "epoch": 3.8009232544720137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036251044007997607,
      "loss": 5.5233,
      "mean_token_accuracy": 0.21362363696098327,
      "num_tokens": 83478688.0,
      "step": 32935
    },
    {
      "entropy": 6.125126361846924,
      "epoch": 3.8015002885170226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003624724232022563,
      "loss": 5.68,
      "mean_token_accuracy": 0.20447041392326354,
      "num_tokens": 83492474.0,
      "step": 32940
    },
    {
      "entropy": 6.136812591552735,
      "epoch": 3.802077322562031,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036243440338244503,
      "loss": 5.6274,
      "mean_token_accuracy": 0.20964494496583938,
      "num_tokens": 83505409.0,
      "step": 32945
    },
    {
      "entropy": 6.062854051589966,
      "epoch": 3.80265435660704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003623963806218211,
      "loss": 5.5439,
      "mean_token_accuracy": 0.208920019865036,
      "num_tokens": 83518502.0,
      "step": 32950
    },
    {
      "entropy": 6.221317195892334,
      "epoch": 3.8032313906520487,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036235835492166326,
      "loss": 5.812,
      "mean_token_accuracy": 0.18976594656705856,
      "num_tokens": 83531688.0,
      "step": 32955
    },
    {
      "entropy": 6.185259485244751,
      "epoch": 3.803808424697057,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003623203262832505,
      "loss": 5.6891,
      "mean_token_accuracy": 0.19513944387435914,
      "num_tokens": 83544845.0,
      "step": 32960
    },
    {
      "entropy": 6.0592139720916744,
      "epoch": 3.8043854587420656,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003622822947078621,
      "loss": 5.586,
      "mean_token_accuracy": 0.20544871240854262,
      "num_tokens": 83558178.0,
      "step": 32965
    },
    {
      "entropy": 6.047280788421631,
      "epoch": 3.8049624927870744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003622442601967769,
      "loss": 5.6343,
      "mean_token_accuracy": 0.1995137721300125,
      "num_tokens": 83570405.0,
      "step": 32970
    },
    {
      "entropy": 6.122627830505371,
      "epoch": 3.8055395268320833,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003622062227512745,
      "loss": 5.6297,
      "mean_token_accuracy": 0.20263560712337494,
      "num_tokens": 83584225.0,
      "step": 32975
    },
    {
      "entropy": 6.146964168548584,
      "epoch": 3.8061165608770917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036216818237263393,
      "loss": 5.6139,
      "mean_token_accuracy": 0.2086881786584854,
      "num_tokens": 83596187.0,
      "step": 32980
    },
    {
      "entropy": 6.076435708999634,
      "epoch": 3.8066935949221,
      "grad_norm": 1.03125,
      "learning_rate": 0.000362130139062135,
      "loss": 5.6076,
      "mean_token_accuracy": 0.20205966681241988,
      "num_tokens": 83608798.0,
      "step": 32985
    },
    {
      "entropy": 6.103664541244507,
      "epoch": 3.807270628967109,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003620920928210571,
      "loss": 5.5517,
      "mean_token_accuracy": 0.20823033452033995,
      "num_tokens": 83621826.0,
      "step": 32990
    },
    {
      "entropy": 6.055540227890015,
      "epoch": 3.807847663012118,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003620540436506799,
      "loss": 5.6002,
      "mean_token_accuracy": 0.2017235904932022,
      "num_tokens": 83633632.0,
      "step": 32995
    },
    {
      "entropy": 6.098191022872925,
      "epoch": 3.8084246970571263,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036201599155228313,
      "loss": 5.6355,
      "mean_token_accuracy": 0.2051863595843315,
      "num_tokens": 83647659.0,
      "step": 33000
    },
    {
      "epoch": 3.8084246970571263,
      "eval_entropy": 6.002264706899336,
      "eval_loss": 5.746444225311279,
      "eval_mean_token_accuracy": 0.2049994945615098,
      "eval_num_tokens": 83647659.0,
      "eval_runtime": 22.9958,
      "eval_samples_per_second": 1223.53,
      "eval_steps_per_second": 152.941,
      "step": 33000
    },
    {
      "entropy": 6.115227508544922,
      "epoch": 3.809001731102135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003619779365271468,
      "loss": 5.5523,
      "mean_token_accuracy": 0.2084027573466301,
      "num_tokens": 83660950.0,
      "step": 33005
    },
    {
      "entropy": 6.137101650238037,
      "epoch": 3.8095787651471436,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036193987857655074,
      "loss": 5.62,
      "mean_token_accuracy": 0.20889806002378464,
      "num_tokens": 83673671.0,
      "step": 33010
    },
    {
      "entropy": 6.092999982833862,
      "epoch": 3.8101557991921524,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036190181770177503,
      "loss": 5.5949,
      "mean_token_accuracy": 0.20383886694908143,
      "num_tokens": 83685851.0,
      "step": 33015
    },
    {
      "entropy": 6.116231966018677,
      "epoch": 3.810732833237161,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003618637539040998,
      "loss": 5.5794,
      "mean_token_accuracy": 0.20814661234617232,
      "num_tokens": 83699138.0,
      "step": 33020
    },
    {
      "entropy": 6.08889799118042,
      "epoch": 3.8113098672821697,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003618256871848054,
      "loss": 5.6346,
      "mean_token_accuracy": 0.2052951991558075,
      "num_tokens": 83711392.0,
      "step": 33025
    },
    {
      "entropy": 6.04820442199707,
      "epoch": 3.8118869013271786,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003617876175451722,
      "loss": 5.604,
      "mean_token_accuracy": 0.21000177413225174,
      "num_tokens": 83726170.0,
      "step": 33030
    },
    {
      "entropy": 6.121216201782227,
      "epoch": 3.812463935372187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036174954498648055,
      "loss": 5.6153,
      "mean_token_accuracy": 0.21042023450136185,
      "num_tokens": 83738724.0,
      "step": 33035
    },
    {
      "entropy": 6.003577136993409,
      "epoch": 3.8130409694171954,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000361711469510011,
      "loss": 5.56,
      "mean_token_accuracy": 0.21309767961502074,
      "num_tokens": 83752177.0,
      "step": 33040
    },
    {
      "entropy": 6.0856188297271725,
      "epoch": 3.8136180034622043,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003616733911170443,
      "loss": 5.5925,
      "mean_token_accuracy": 0.20746787786483764,
      "num_tokens": 83764645.0,
      "step": 33045
    },
    {
      "entropy": 6.147206020355225,
      "epoch": 3.814195037507213,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036163530980886114,
      "loss": 5.5931,
      "mean_token_accuracy": 0.1981767639517784,
      "num_tokens": 83775843.0,
      "step": 33050
    },
    {
      "entropy": 6.113157415390015,
      "epoch": 3.8147720715522215,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003615972255867424,
      "loss": 5.5825,
      "mean_token_accuracy": 0.20405851304531097,
      "num_tokens": 83788308.0,
      "step": 33055
    },
    {
      "entropy": 6.139213609695434,
      "epoch": 3.8153491055972304,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036155913845196903,
      "loss": 5.6362,
      "mean_token_accuracy": 0.20707210898399353,
      "num_tokens": 83800947.0,
      "step": 33060
    },
    {
      "entropy": 6.087186336517334,
      "epoch": 3.815926139642239,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003615210484058219,
      "loss": 5.5891,
      "mean_token_accuracy": 0.21279811561107637,
      "num_tokens": 83814332.0,
      "step": 33065
    },
    {
      "entropy": 6.127863264083862,
      "epoch": 3.8165031736872477,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003614829554495823,
      "loss": 5.6754,
      "mean_token_accuracy": 0.19855415672063828,
      "num_tokens": 83825942.0,
      "step": 33070
    },
    {
      "entropy": 6.17516713142395,
      "epoch": 3.817080207732256,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036144485958453153,
      "loss": 5.6587,
      "mean_token_accuracy": 0.20230256766080856,
      "num_tokens": 83837841.0,
      "step": 33075
    },
    {
      "entropy": 6.080211496353149,
      "epoch": 3.817657241777265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003614067608119508,
      "loss": 5.5952,
      "mean_token_accuracy": 0.2020416870713234,
      "num_tokens": 83849675.0,
      "step": 33080
    },
    {
      "entropy": 6.109024286270142,
      "epoch": 3.8182342758222734,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003613686591331216,
      "loss": 5.6302,
      "mean_token_accuracy": 0.20066949725151062,
      "num_tokens": 83861834.0,
      "step": 33085
    },
    {
      "entropy": 6.0655725479125975,
      "epoch": 3.8188113098672822,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003613305545493255,
      "loss": 5.5421,
      "mean_token_accuracy": 0.20998494774103166,
      "num_tokens": 83874360.0,
      "step": 33090
    },
    {
      "entropy": 6.124150657653809,
      "epoch": 3.8193883439122907,
      "grad_norm": 0.984375,
      "learning_rate": 0.000361292447061844,
      "loss": 5.6752,
      "mean_token_accuracy": 0.20430647134780883,
      "num_tokens": 83889262.0,
      "step": 33095
    },
    {
      "entropy": 6.205819034576416,
      "epoch": 3.8199653779572995,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036125433667195896,
      "loss": 5.686,
      "mean_token_accuracy": 0.19616526067256929,
      "num_tokens": 83901967.0,
      "step": 33100
    },
    {
      "entropy": 6.085394954681396,
      "epoch": 3.8205424120023084,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003612162233809521,
      "loss": 5.5914,
      "mean_token_accuracy": 0.2094152271747589,
      "num_tokens": 83914909.0,
      "step": 33105
    },
    {
      "entropy": 6.0280091762542725,
      "epoch": 3.821119446047317,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003611781071901054,
      "loss": 5.6194,
      "mean_token_accuracy": 0.2058122530579567,
      "num_tokens": 83927414.0,
      "step": 33110
    },
    {
      "entropy": 6.151128101348877,
      "epoch": 3.821696480092325,
      "grad_norm": 0.8203125,
      "learning_rate": 0.00036113998810070094,
      "loss": 5.659,
      "mean_token_accuracy": 0.20045981258153917,
      "num_tokens": 83940341.0,
      "step": 33115
    },
    {
      "entropy": 6.171204614639282,
      "epoch": 3.822273514137334,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036110186611402074,
      "loss": 5.626,
      "mean_token_accuracy": 0.19879001677036284,
      "num_tokens": 83952509.0,
      "step": 33120
    },
    {
      "entropy": 6.104555416107178,
      "epoch": 3.822850548182343,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000361063741231347,
      "loss": 5.645,
      "mean_token_accuracy": 0.19978587478399276,
      "num_tokens": 83964455.0,
      "step": 33125
    },
    {
      "entropy": 6.100145387649536,
      "epoch": 3.8234275822273514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003610256134539621,
      "loss": 5.5868,
      "mean_token_accuracy": 0.2091189667582512,
      "num_tokens": 83977003.0,
      "step": 33130
    },
    {
      "entropy": 6.170768928527832,
      "epoch": 3.82400461627236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003609874827831484,
      "loss": 5.6778,
      "mean_token_accuracy": 0.19716740101575853,
      "num_tokens": 83988740.0,
      "step": 33135
    },
    {
      "entropy": 6.089428853988648,
      "epoch": 3.8245816503173686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036094934922018847,
      "loss": 5.6189,
      "mean_token_accuracy": 0.20273886620998383,
      "num_tokens": 84001814.0,
      "step": 33140
    },
    {
      "entropy": 6.039244985580444,
      "epoch": 3.8251586843623775,
      "grad_norm": 1.0,
      "learning_rate": 0.00036091121276636486,
      "loss": 5.5637,
      "mean_token_accuracy": 0.21307864636182786,
      "num_tokens": 84013889.0,
      "step": 33145
    },
    {
      "entropy": 6.0446113586425785,
      "epoch": 3.825735718407386,
      "grad_norm": 1.8125,
      "learning_rate": 0.0003608730734229602,
      "loss": 5.5768,
      "mean_token_accuracy": 0.2095784768462181,
      "num_tokens": 84025453.0,
      "step": 33150
    },
    {
      "entropy": 6.085889053344727,
      "epoch": 3.8263127524523948,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003608349311912575,
      "loss": 5.5547,
      "mean_token_accuracy": 0.19999430775642396,
      "num_tokens": 84038246.0,
      "step": 33155
    },
    {
      "entropy": 6.198304986953735,
      "epoch": 3.826889786497403,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036079678607253945,
      "loss": 5.7339,
      "mean_token_accuracy": 0.19320725351572038,
      "num_tokens": 84049559.0,
      "step": 33160
    },
    {
      "entropy": 6.095018768310547,
      "epoch": 3.827466820542412,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003607586380680892,
      "loss": 5.6464,
      "mean_token_accuracy": 0.21408105045557022,
      "num_tokens": 84060969.0,
      "step": 33165
    },
    {
      "entropy": 6.165418672561645,
      "epoch": 3.8280438545874205,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003607204871791896,
      "loss": 5.6829,
      "mean_token_accuracy": 0.19782189279794693,
      "num_tokens": 84074623.0,
      "step": 33170
    },
    {
      "entropy": 5.951971817016601,
      "epoch": 3.8286208886324293,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036068233340712415,
      "loss": 5.4868,
      "mean_token_accuracy": 0.21891556829214096,
      "num_tokens": 84087187.0,
      "step": 33175
    },
    {
      "entropy": 6.096566867828369,
      "epoch": 3.829197922677438,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003606441767531759,
      "loss": 5.6831,
      "mean_token_accuracy": 0.19806020557880402,
      "num_tokens": 84099392.0,
      "step": 33180
    },
    {
      "entropy": 6.056458854675293,
      "epoch": 3.8297749567224466,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036060601721862835,
      "loss": 5.5279,
      "mean_token_accuracy": 0.21461428552865983,
      "num_tokens": 84114191.0,
      "step": 33185
    },
    {
      "entropy": 6.036667156219482,
      "epoch": 3.830351990767455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003605678548047649,
      "loss": 5.5752,
      "mean_token_accuracy": 0.2035483315587044,
      "num_tokens": 84126013.0,
      "step": 33190
    },
    {
      "entropy": 6.0725034236907955,
      "epoch": 3.830929024812464,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00036052968951286917,
      "loss": 5.6617,
      "mean_token_accuracy": 0.2068752095103264,
      "num_tokens": 84138702.0,
      "step": 33195
    },
    {
      "entropy": 6.125632381439209,
      "epoch": 3.8315060588574728,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003604915213442248,
      "loss": 5.61,
      "mean_token_accuracy": 0.2049739181995392,
      "num_tokens": 84151854.0,
      "step": 33200
    },
    {
      "entropy": 6.055310153961182,
      "epoch": 3.832083092902481,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00036045335030011563,
      "loss": 5.6128,
      "mean_token_accuracy": 0.20610808283090593,
      "num_tokens": 84165183.0,
      "step": 33205
    },
    {
      "entropy": 6.07265453338623,
      "epoch": 3.83266012694749,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003604151763818254,
      "loss": 5.565,
      "mean_token_accuracy": 0.19883307218551635,
      "num_tokens": 84177082.0,
      "step": 33210
    },
    {
      "entropy": 6.142414283752442,
      "epoch": 3.8332371609924984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003603769995906382,
      "loss": 5.6089,
      "mean_token_accuracy": 0.20537890940904618,
      "num_tokens": 84189289.0,
      "step": 33215
    },
    {
      "entropy": 6.032017850875855,
      "epoch": 3.8338141950375073,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000360338819927838,
      "loss": 5.5256,
      "mean_token_accuracy": 0.21574822068214417,
      "num_tokens": 84201877.0,
      "step": 33220
    },
    {
      "entropy": 6.16310749053955,
      "epoch": 3.8343912290825157,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000360300637394709,
      "loss": 5.6747,
      "mean_token_accuracy": 0.19754361659288405,
      "num_tokens": 84212963.0,
      "step": 33225
    },
    {
      "entropy": 6.070192956924439,
      "epoch": 3.8349682631275246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003602624519925354,
      "loss": 5.5556,
      "mean_token_accuracy": 0.21068055331707,
      "num_tokens": 84227577.0,
      "step": 33230
    },
    {
      "entropy": 6.1115833759307865,
      "epoch": 3.8355452971725335,
      "grad_norm": 0.90625,
      "learning_rate": 0.00036022426372260156,
      "loss": 5.7118,
      "mean_token_accuracy": 0.19522076100111008,
      "num_tokens": 84240285.0,
      "step": 33235
    },
    {
      "entropy": 6.104603958129883,
      "epoch": 3.836122331217542,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036018607258619196,
      "loss": 5.6036,
      "mean_token_accuracy": 0.20627150684595108,
      "num_tokens": 84253308.0,
      "step": 33240
    },
    {
      "entropy": 6.095502185821533,
      "epoch": 3.8366993652625503,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003601478785845912,
      "loss": 5.595,
      "mean_token_accuracy": 0.20843304842710494,
      "num_tokens": 84265213.0,
      "step": 33245
    },
    {
      "entropy": 6.1630088806152346,
      "epoch": 3.837276399307559,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003601096817190837,
      "loss": 5.6728,
      "mean_token_accuracy": 0.20044924467802047,
      "num_tokens": 84277865.0,
      "step": 33250
    },
    {
      "entropy": 6.159406042098999,
      "epoch": 3.837853433352568,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003600714819909544,
      "loss": 5.7262,
      "mean_token_accuracy": 0.19874559193849564,
      "num_tokens": 84289771.0,
      "step": 33255
    },
    {
      "entropy": 6.191400241851807,
      "epoch": 3.8384304673975764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003600332794014881,
      "loss": 5.6583,
      "mean_token_accuracy": 0.20253454446792601,
      "num_tokens": 84300716.0,
      "step": 33260
    },
    {
      "entropy": 6.099560737609863,
      "epoch": 3.839007501442585,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003599950739519697,
      "loss": 5.6785,
      "mean_token_accuracy": 0.20405016839504242,
      "num_tokens": 84312274.0,
      "step": 33265
    },
    {
      "entropy": 6.152928352355957,
      "epoch": 3.8395845354875937,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003599568656436841,
      "loss": 5.6313,
      "mean_token_accuracy": 0.20496930330991744,
      "num_tokens": 84324550.0,
      "step": 33270
    },
    {
      "entropy": 6.09661979675293,
      "epoch": 3.8401615695326026,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003599186544779166,
      "loss": 5.6012,
      "mean_token_accuracy": 0.20848660171031952,
      "num_tokens": 84336935.0,
      "step": 33275
    },
    {
      "entropy": 6.110539293289184,
      "epoch": 3.840738603577611,
      "grad_norm": 0.875,
      "learning_rate": 0.0003598804404559523,
      "loss": 5.65,
      "mean_token_accuracy": 0.20124222189188004,
      "num_tokens": 84349757.0,
      "step": 33280
    },
    {
      "entropy": 6.139134454727173,
      "epoch": 3.84131563762262,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035984222357907667,
      "loss": 5.589,
      "mean_token_accuracy": 0.2093803256750107,
      "num_tokens": 84361149.0,
      "step": 33285
    },
    {
      "entropy": 5.919655656814575,
      "epoch": 3.8418926716676283,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035980400384857486,
      "loss": 5.4104,
      "mean_token_accuracy": 0.22745767384767532,
      "num_tokens": 84373714.0,
      "step": 33290
    },
    {
      "entropy": 6.180468368530273,
      "epoch": 3.842469705712637,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003597657812657325,
      "loss": 5.6793,
      "mean_token_accuracy": 0.19682663232088088,
      "num_tokens": 84386193.0,
      "step": 33295
    },
    {
      "entropy": 6.185771417617798,
      "epoch": 3.8430467397576455,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035972755583183526,
      "loss": 5.6793,
      "mean_token_accuracy": 0.19723526984453202,
      "num_tokens": 84399122.0,
      "step": 33300
    },
    {
      "entropy": 6.039923858642578,
      "epoch": 3.8436237738026544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035968932754816877,
      "loss": 5.6108,
      "mean_token_accuracy": 0.20566056966781615,
      "num_tokens": 84413058.0,
      "step": 33305
    },
    {
      "entropy": 6.129089593887329,
      "epoch": 3.8442008078476633,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003596510964160188,
      "loss": 5.5797,
      "mean_token_accuracy": 0.20427815169095992,
      "num_tokens": 84425003.0,
      "step": 33310
    },
    {
      "entropy": 6.022801113128662,
      "epoch": 3.8447778418926717,
      "grad_norm": 1.421875,
      "learning_rate": 0.00035961286243667116,
      "loss": 5.4697,
      "mean_token_accuracy": 0.2220272034406662,
      "num_tokens": 84437873.0,
      "step": 33315
    },
    {
      "entropy": 6.0829442024230955,
      "epoch": 3.84535487593768,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000359574625611412,
      "loss": 5.6626,
      "mean_token_accuracy": 0.20509454607963562,
      "num_tokens": 84450991.0,
      "step": 33320
    },
    {
      "entropy": 6.022344446182251,
      "epoch": 3.845931909982689,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003595363859415273,
      "loss": 5.5577,
      "mean_token_accuracy": 0.20722992420196534,
      "num_tokens": 84464464.0,
      "step": 33325
    },
    {
      "entropy": 6.083918523788452,
      "epoch": 3.846508944027698,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035949814342830334,
      "loss": 5.5997,
      "mean_token_accuracy": 0.20815321505069734,
      "num_tokens": 84477757.0,
      "step": 33330
    },
    {
      "entropy": 6.119990015029908,
      "epoch": 3.8470859780727062,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003594598980730262,
      "loss": 5.4876,
      "mean_token_accuracy": 0.20616916865110396,
      "num_tokens": 84490470.0,
      "step": 33335
    },
    {
      "entropy": 6.165686893463135,
      "epoch": 3.847663012117715,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003594216498769824,
      "loss": 5.602,
      "mean_token_accuracy": 0.20258891731500625,
      "num_tokens": 84503093.0,
      "step": 33340
    },
    {
      "entropy": 6.046669244766235,
      "epoch": 3.8482400461627235,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035938339884145824,
      "loss": 5.6079,
      "mean_token_accuracy": 0.20831372886896132,
      "num_tokens": 84515135.0,
      "step": 33345
    },
    {
      "entropy": 6.1239931106567385,
      "epoch": 3.8488170802077324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003593451449677405,
      "loss": 5.6386,
      "mean_token_accuracy": 0.2001243159174919,
      "num_tokens": 84526988.0,
      "step": 33350
    },
    {
      "entropy": 6.136247539520264,
      "epoch": 3.849394114252741,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003593068882571156,
      "loss": 5.5566,
      "mean_token_accuracy": 0.2071867674589157,
      "num_tokens": 84539350.0,
      "step": 33355
    },
    {
      "entropy": 6.036378526687622,
      "epoch": 3.8499711482977497,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035926862871087046,
      "loss": 5.5548,
      "mean_token_accuracy": 0.20641928315162658,
      "num_tokens": 84551562.0,
      "step": 33360
    },
    {
      "entropy": 6.111506557464599,
      "epoch": 3.850548182342758,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003592303663302919,
      "loss": 5.6251,
      "mean_token_accuracy": 0.20090053528547286,
      "num_tokens": 84563487.0,
      "step": 33365
    },
    {
      "entropy": 6.125816679000854,
      "epoch": 3.851125216387767,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003591921011166667,
      "loss": 5.5992,
      "mean_token_accuracy": 0.20233289003372193,
      "num_tokens": 84575865.0,
      "step": 33370
    },
    {
      "entropy": 6.08932785987854,
      "epoch": 3.8517022504327754,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000359153833071282,
      "loss": 5.6471,
      "mean_token_accuracy": 0.20459658801555633,
      "num_tokens": 84589183.0,
      "step": 33375
    },
    {
      "entropy": 6.065012645721436,
      "epoch": 3.8522792844777842,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035911556219542495,
      "loss": 5.5694,
      "mean_token_accuracy": 0.20598982125520707,
      "num_tokens": 84603798.0,
      "step": 33380
    },
    {
      "entropy": 6.103428173065185,
      "epoch": 3.852856318522793,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003590772884903828,
      "loss": 5.5781,
      "mean_token_accuracy": 0.2167849987745285,
      "num_tokens": 84616056.0,
      "step": 33385
    },
    {
      "entropy": 6.1010619640350345,
      "epoch": 3.8534333525678015,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035903901195744276,
      "loss": 5.5789,
      "mean_token_accuracy": 0.2054425969719887,
      "num_tokens": 84629854.0,
      "step": 33390
    },
    {
      "entropy": 6.087514495849609,
      "epoch": 3.85401038661281,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003590007325978923,
      "loss": 5.663,
      "mean_token_accuracy": 0.1989055022597313,
      "num_tokens": 84642015.0,
      "step": 33395
    },
    {
      "entropy": 6.140673685073852,
      "epoch": 3.854587420657819,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000358962450413019,
      "loss": 5.6457,
      "mean_token_accuracy": 0.2022153079509735,
      "num_tokens": 84654732.0,
      "step": 33400
    },
    {
      "entropy": 6.1412034034729,
      "epoch": 3.8551644547028276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003589241654041103,
      "loss": 5.6813,
      "mean_token_accuracy": 0.2025569811463356,
      "num_tokens": 84667315.0,
      "step": 33405
    },
    {
      "entropy": 6.141792297363281,
      "epoch": 3.855741488747836,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000358885877572454,
      "loss": 5.6459,
      "mean_token_accuracy": 0.19607073813676834,
      "num_tokens": 84680291.0,
      "step": 33410
    },
    {
      "entropy": 6.064030408859253,
      "epoch": 3.856318522792845,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035884758691933787,
      "loss": 5.5756,
      "mean_token_accuracy": 0.20241304934024812,
      "num_tokens": 84692716.0,
      "step": 33415
    },
    {
      "entropy": 6.074345111846924,
      "epoch": 3.8568955568378533,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003588092934460498,
      "loss": 5.5397,
      "mean_token_accuracy": 0.2167934238910675,
      "num_tokens": 84704822.0,
      "step": 33420
    },
    {
      "entropy": 6.087766408920288,
      "epoch": 3.857472590882862,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035877099715387776,
      "loss": 5.6489,
      "mean_token_accuracy": 0.19916362017393113,
      "num_tokens": 84718047.0,
      "step": 33425
    },
    {
      "entropy": 6.169582557678223,
      "epoch": 3.8580496249278706,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035873269804410993,
      "loss": 5.7337,
      "mean_token_accuracy": 0.19794039726257323,
      "num_tokens": 84731209.0,
      "step": 33430
    },
    {
      "entropy": 6.105790710449218,
      "epoch": 3.8586266589728795,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003586943961180343,
      "loss": 5.6645,
      "mean_token_accuracy": 0.19722446501255037,
      "num_tokens": 84745014.0,
      "step": 33435
    },
    {
      "entropy": 6.09491753578186,
      "epoch": 3.859203693017888,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035865609137693933,
      "loss": 5.5895,
      "mean_token_accuracy": 0.21143753230571746,
      "num_tokens": 84758751.0,
      "step": 33440
    },
    {
      "entropy": 6.14051570892334,
      "epoch": 3.8597807270628968,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003586177838221132,
      "loss": 5.6027,
      "mean_token_accuracy": 0.1978860840201378,
      "num_tokens": 84772108.0,
      "step": 33445
    },
    {
      "entropy": 6.110732841491699,
      "epoch": 3.860357761107905,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035857947345484454,
      "loss": 5.6628,
      "mean_token_accuracy": 0.20033389180898667,
      "num_tokens": 84785120.0,
      "step": 33450
    },
    {
      "entropy": 6.180313491821289,
      "epoch": 3.860934795152914,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003585411602764217,
      "loss": 5.688,
      "mean_token_accuracy": 0.19840385317802428,
      "num_tokens": 84797941.0,
      "step": 33455
    },
    {
      "entropy": 6.056058740615844,
      "epoch": 3.861511829197923,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035850284428813346,
      "loss": 5.5353,
      "mean_token_accuracy": 0.21057915687561035,
      "num_tokens": 84809065.0,
      "step": 33460
    },
    {
      "entropy": 6.097518968582153,
      "epoch": 3.8620888632429313,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003584645254912685,
      "loss": 5.6428,
      "mean_token_accuracy": 0.20579753965139388,
      "num_tokens": 84821865.0,
      "step": 33465
    },
    {
      "entropy": 6.101900672912597,
      "epoch": 3.8626658972879397,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035842620388711575,
      "loss": 5.5593,
      "mean_token_accuracy": 0.20282736718654631,
      "num_tokens": 84835451.0,
      "step": 33470
    },
    {
      "entropy": 6.084393548965454,
      "epoch": 3.8632429313329486,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003583878794769641,
      "loss": 5.6032,
      "mean_token_accuracy": 0.20311120748519898,
      "num_tokens": 84847732.0,
      "step": 33475
    },
    {
      "entropy": 6.057758474349976,
      "epoch": 3.8638199653779575,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003583495522621024,
      "loss": 5.5535,
      "mean_token_accuracy": 0.2112577587366104,
      "num_tokens": 84860431.0,
      "step": 33480
    },
    {
      "entropy": 6.035004901885986,
      "epoch": 3.864396999422966,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035831122224382,
      "loss": 5.573,
      "mean_token_accuracy": 0.2079429879784584,
      "num_tokens": 84872810.0,
      "step": 33485
    },
    {
      "entropy": 6.062376546859741,
      "epoch": 3.8649740334679747,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035827288942340605,
      "loss": 5.6099,
      "mean_token_accuracy": 0.2063120722770691,
      "num_tokens": 84885394.0,
      "step": 33490
    },
    {
      "entropy": 6.10299859046936,
      "epoch": 3.865551067512983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035823455380214984,
      "loss": 5.6076,
      "mean_token_accuracy": 0.20408399701118468,
      "num_tokens": 84898416.0,
      "step": 33495
    },
    {
      "entropy": 6.079794597625733,
      "epoch": 3.866128101557992,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003581962153813406,
      "loss": 5.5482,
      "mean_token_accuracy": 0.2129485249519348,
      "num_tokens": 84910550.0,
      "step": 33500
    },
    {
      "entropy": 6.043974924087524,
      "epoch": 3.8667051356030004,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003581578741622682,
      "loss": 5.5745,
      "mean_token_accuracy": 0.20866378545761108,
      "num_tokens": 84923177.0,
      "step": 33505
    },
    {
      "entropy": 6.090998697280884,
      "epoch": 3.8672821696480093,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003581195301462219,
      "loss": 5.6049,
      "mean_token_accuracy": 0.2031246930360794,
      "num_tokens": 84935032.0,
      "step": 33510
    },
    {
      "entropy": 5.993401050567627,
      "epoch": 3.8678592036930177,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003580811833344915,
      "loss": 5.4978,
      "mean_token_accuracy": 0.21389926671981813,
      "num_tokens": 84947660.0,
      "step": 33515
    },
    {
      "entropy": 6.118982887268066,
      "epoch": 3.8684362377380266,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003580428337283667,
      "loss": 5.6738,
      "mean_token_accuracy": 0.2047731786966324,
      "num_tokens": 84960916.0,
      "step": 33520
    },
    {
      "entropy": 6.154660415649414,
      "epoch": 3.869013271783035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035800448132913755,
      "loss": 5.6245,
      "mean_token_accuracy": 0.2007632613182068,
      "num_tokens": 84974022.0,
      "step": 33525
    },
    {
      "entropy": 6.087403440475464,
      "epoch": 3.869590305828044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003579661261380938,
      "loss": 5.5966,
      "mean_token_accuracy": 0.2088603690266609,
      "num_tokens": 84987654.0,
      "step": 33530
    },
    {
      "entropy": 6.098627519607544,
      "epoch": 3.8701673398730527,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003579277681565257,
      "loss": 5.6741,
      "mean_token_accuracy": 0.19669127762317656,
      "num_tokens": 85000781.0,
      "step": 33535
    },
    {
      "entropy": 6.154697370529175,
      "epoch": 3.870744373918061,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035788940738572326,
      "loss": 5.6863,
      "mean_token_accuracy": 0.19641603529453278,
      "num_tokens": 85013145.0,
      "step": 33540
    },
    {
      "entropy": 6.104619455337525,
      "epoch": 3.8713214079630696,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035785104382697673,
      "loss": 5.6071,
      "mean_token_accuracy": 0.20106724053621292,
      "num_tokens": 85025795.0,
      "step": 33545
    },
    {
      "entropy": 6.126614999771118,
      "epoch": 3.8718984420080784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003578126774815766,
      "loss": 5.6084,
      "mean_token_accuracy": 0.20979945361614227,
      "num_tokens": 85038123.0,
      "step": 33550
    },
    {
      "entropy": 6.104243898391724,
      "epoch": 3.8724754760530873,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003577743083508132,
      "loss": 5.6244,
      "mean_token_accuracy": 0.2064577892422676,
      "num_tokens": 85050323.0,
      "step": 33555
    },
    {
      "entropy": 6.0725977420806885,
      "epoch": 3.8730525100980957,
      "grad_norm": 0.890625,
      "learning_rate": 0.000357735936435977,
      "loss": 5.605,
      "mean_token_accuracy": 0.20157973766326903,
      "num_tokens": 85061897.0,
      "step": 33560
    },
    {
      "entropy": 6.089364719390869,
      "epoch": 3.8736295441431046,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035769756173835866,
      "loss": 5.5729,
      "mean_token_accuracy": 0.20399502217769622,
      "num_tokens": 85074536.0,
      "step": 33565
    },
    {
      "entropy": 6.132237291336059,
      "epoch": 3.874206578188113,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035765918425924894,
      "loss": 5.6443,
      "mean_token_accuracy": 0.2067454054951668,
      "num_tokens": 85087493.0,
      "step": 33570
    },
    {
      "entropy": 5.971921682357788,
      "epoch": 3.874783612233122,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00035762080399993863,
      "loss": 5.446,
      "mean_token_accuracy": 0.2301473543047905,
      "num_tokens": 85101037.0,
      "step": 33575
    },
    {
      "entropy": 5.985307025909424,
      "epoch": 3.8753606462781303,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035758242096171856,
      "loss": 5.4892,
      "mean_token_accuracy": 0.2119239941239357,
      "num_tokens": 85114268.0,
      "step": 33580
    },
    {
      "entropy": 6.125138998031616,
      "epoch": 3.875937680323139,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003575440351458798,
      "loss": 5.7185,
      "mean_token_accuracy": 0.19590566903352738,
      "num_tokens": 85127059.0,
      "step": 33585
    },
    {
      "entropy": 6.152217149734497,
      "epoch": 3.876514714368148,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003575056465537135,
      "loss": 5.619,
      "mean_token_accuracy": 0.20546552985906602,
      "num_tokens": 85139353.0,
      "step": 33590
    },
    {
      "entropy": 5.932551670074463,
      "epoch": 3.8770917484131564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003574672551865107,
      "loss": 5.4573,
      "mean_token_accuracy": 0.21776998043060303,
      "num_tokens": 85153312.0,
      "step": 33595
    },
    {
      "entropy": 6.13020601272583,
      "epoch": 3.877668782458165,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003574288610455627,
      "loss": 5.6524,
      "mean_token_accuracy": 0.19965772479772567,
      "num_tokens": 85165440.0,
      "step": 33600
    },
    {
      "entropy": 6.128564500808716,
      "epoch": 3.8782458165031737,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035739046413216095,
      "loss": 5.6839,
      "mean_token_accuracy": 0.19749677032232285,
      "num_tokens": 85178673.0,
      "step": 33605
    },
    {
      "entropy": 6.088786935806274,
      "epoch": 3.8788228505481825,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003573520644475968,
      "loss": 5.7057,
      "mean_token_accuracy": 0.20382700860500336,
      "num_tokens": 85192587.0,
      "step": 33610
    },
    {
      "entropy": 6.027975034713745,
      "epoch": 3.879399884593191,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035731366199316195,
      "loss": 5.5131,
      "mean_token_accuracy": 0.20946575999259948,
      "num_tokens": 85204247.0,
      "step": 33615
    },
    {
      "entropy": 6.07987265586853,
      "epoch": 3.8799769186381994,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000357275256770148,
      "loss": 5.5702,
      "mean_token_accuracy": 0.20465978235006332,
      "num_tokens": 85217096.0,
      "step": 33620
    },
    {
      "entropy": 6.093013525009155,
      "epoch": 3.8805539526832082,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003572368487798465,
      "loss": 5.6047,
      "mean_token_accuracy": 0.20413179099559783,
      "num_tokens": 85229988.0,
      "step": 33625
    },
    {
      "entropy": 6.136398601531982,
      "epoch": 3.881130986728217,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003571984380235496,
      "loss": 5.6681,
      "mean_token_accuracy": 0.19391278624534608,
      "num_tokens": 85242386.0,
      "step": 33630
    },
    {
      "entropy": 6.104293966293335,
      "epoch": 3.8817080207732255,
      "grad_norm": 1.0,
      "learning_rate": 0.000357160024502549,
      "loss": 5.6218,
      "mean_token_accuracy": 0.1969421312212944,
      "num_tokens": 85253723.0,
      "step": 33635
    },
    {
      "entropy": 6.153084325790405,
      "epoch": 3.8822850548182344,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0003571216082181368,
      "loss": 5.6366,
      "mean_token_accuracy": 0.20404088497161865,
      "num_tokens": 85267296.0,
      "step": 33640
    },
    {
      "entropy": 6.043382883071899,
      "epoch": 3.882862088863243,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035708318917160504,
      "loss": 5.5437,
      "mean_token_accuracy": 0.2082097128033638,
      "num_tokens": 85278620.0,
      "step": 33645
    },
    {
      "entropy": 6.086602592468262,
      "epoch": 3.8834391229082517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003570447673642461,
      "loss": 5.6513,
      "mean_token_accuracy": 0.20131202787160873,
      "num_tokens": 85290522.0,
      "step": 33650
    },
    {
      "entropy": 6.182232999801636,
      "epoch": 3.88401615695326,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035700634279735207,
      "loss": 5.6537,
      "mean_token_accuracy": 0.20291696041822432,
      "num_tokens": 85304629.0,
      "step": 33655
    },
    {
      "entropy": 6.1675389289855955,
      "epoch": 3.884593190998269,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003569679154722155,
      "loss": 5.6923,
      "mean_token_accuracy": 0.19452993869781493,
      "num_tokens": 85317844.0,
      "step": 33660
    },
    {
      "entropy": 6.055772495269776,
      "epoch": 3.885170225043278,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003569294853901288,
      "loss": 5.5664,
      "mean_token_accuracy": 0.20586709231138228,
      "num_tokens": 85330819.0,
      "step": 33665
    },
    {
      "entropy": 6.05358772277832,
      "epoch": 3.885747259088286,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003568910525523844,
      "loss": 5.5775,
      "mean_token_accuracy": 0.20565457195043563,
      "num_tokens": 85342833.0,
      "step": 33670
    },
    {
      "entropy": 6.082759428024292,
      "epoch": 3.8863242931332946,
      "grad_norm": 1.0,
      "learning_rate": 0.00035685261696027524,
      "loss": 5.6196,
      "mean_token_accuracy": 0.2065345123410225,
      "num_tokens": 85355250.0,
      "step": 33675
    },
    {
      "entropy": 6.096784448623657,
      "epoch": 3.8869013271783035,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000356814178615094,
      "loss": 5.6002,
      "mean_token_accuracy": 0.20772335529327393,
      "num_tokens": 85367112.0,
      "step": 33680
    },
    {
      "entropy": 6.052044630050659,
      "epoch": 3.8874783612233124,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003567757375181335,
      "loss": 5.5937,
      "mean_token_accuracy": 0.2097228616476059,
      "num_tokens": 85379935.0,
      "step": 33685
    },
    {
      "entropy": 6.060841178894043,
      "epoch": 3.888055395268321,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003567372936706867,
      "loss": 5.6187,
      "mean_token_accuracy": 0.1991729959845543,
      "num_tokens": 85393378.0,
      "step": 33690
    },
    {
      "entropy": 6.147786855697632,
      "epoch": 3.8886324293133296,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035669884707404656,
      "loss": 5.594,
      "mean_token_accuracy": 0.20705071836709976,
      "num_tokens": 85407181.0,
      "step": 33695
    },
    {
      "entropy": 6.092485475540161,
      "epoch": 3.889209463358338,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003566603977295063,
      "loss": 5.5539,
      "mean_token_accuracy": 0.2103789210319519,
      "num_tokens": 85419795.0,
      "step": 33700
    },
    {
      "entropy": 6.018665981292725,
      "epoch": 3.889786497403347,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003566219456383592,
      "loss": 5.6047,
      "mean_token_accuracy": 0.20543065369129182,
      "num_tokens": 85432973.0,
      "step": 33705
    },
    {
      "entropy": 6.144924163818359,
      "epoch": 3.8903635314483553,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035658349080189847,
      "loss": 5.6155,
      "mean_token_accuracy": 0.20156259536743165,
      "num_tokens": 85445256.0,
      "step": 33710
    },
    {
      "entropy": 6.105436849594116,
      "epoch": 3.890940565493364,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003565450332214175,
      "loss": 5.665,
      "mean_token_accuracy": 0.2001728817820549,
      "num_tokens": 85458400.0,
      "step": 33715
    },
    {
      "entropy": 6.163173580169678,
      "epoch": 3.8915175995383726,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035650657289820986,
      "loss": 5.664,
      "mean_token_accuracy": 0.1991086483001709,
      "num_tokens": 85469711.0,
      "step": 33720
    },
    {
      "entropy": 6.164878940582275,
      "epoch": 3.8920946335833815,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003564681098335692,
      "loss": 5.6295,
      "mean_token_accuracy": 0.20740173161029815,
      "num_tokens": 85481672.0,
      "step": 33725
    },
    {
      "entropy": 6.196621465682983,
      "epoch": 3.89267166762839,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035642964402878897,
      "loss": 5.7435,
      "mean_token_accuracy": 0.19491058588027954,
      "num_tokens": 85494511.0,
      "step": 33730
    },
    {
      "entropy": 6.032270240783691,
      "epoch": 3.8932487016733988,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035639117548516325,
      "loss": 5.5114,
      "mean_token_accuracy": 0.21603468507528306,
      "num_tokens": 85507085.0,
      "step": 33735
    },
    {
      "entropy": 6.0698925971984865,
      "epoch": 3.8938257357184076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035635270420398575,
      "loss": 5.6263,
      "mean_token_accuracy": 0.2079391896724701,
      "num_tokens": 85520688.0,
      "step": 33740
    },
    {
      "entropy": 6.131998300552368,
      "epoch": 3.894402769763416,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003563142301865505,
      "loss": 5.6057,
      "mean_token_accuracy": 0.20017666220664979,
      "num_tokens": 85533427.0,
      "step": 33745
    },
    {
      "entropy": 6.161380863189697,
      "epoch": 3.8949798038084245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035627575343415136,
      "loss": 5.6341,
      "mean_token_accuracy": 0.20435797423124313,
      "num_tokens": 85545362.0,
      "step": 33750
    },
    {
      "entropy": 6.079386758804321,
      "epoch": 3.8955568378534333,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003562372739480827,
      "loss": 5.5923,
      "mean_token_accuracy": 0.20815217941999437,
      "num_tokens": 85558189.0,
      "step": 33755
    },
    {
      "entropy": 5.971133279800415,
      "epoch": 3.896133871898442,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003561987917296387,
      "loss": 5.5222,
      "mean_token_accuracy": 0.2101283222436905,
      "num_tokens": 85571513.0,
      "step": 33760
    },
    {
      "entropy": 6.060505390167236,
      "epoch": 3.8967109059434506,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003561603067801137,
      "loss": 5.5749,
      "mean_token_accuracy": 0.2111502930521965,
      "num_tokens": 85583977.0,
      "step": 33765
    },
    {
      "entropy": 6.132311201095581,
      "epoch": 3.8972879399884595,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035612181910080204,
      "loss": 5.6349,
      "mean_token_accuracy": 0.20717811584472656,
      "num_tokens": 85595543.0,
      "step": 33770
    },
    {
      "entropy": 6.079429674148559,
      "epoch": 3.897864974033468,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035608332869299837,
      "loss": 5.5534,
      "mean_token_accuracy": 0.2110853835940361,
      "num_tokens": 85608621.0,
      "step": 33775
    },
    {
      "entropy": 6.0897375583648685,
      "epoch": 3.8984420080784767,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003560448355579971,
      "loss": 5.5337,
      "mean_token_accuracy": 0.21005466878414153,
      "num_tokens": 85621607.0,
      "step": 33780
    },
    {
      "entropy": 5.973836946487427,
      "epoch": 3.899019042123485,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003560063396970932,
      "loss": 5.5175,
      "mean_token_accuracy": 0.21489454060792923,
      "num_tokens": 85633957.0,
      "step": 33785
    },
    {
      "entropy": 6.126867055892944,
      "epoch": 3.899596076168494,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003559678411115811,
      "loss": 5.738,
      "mean_token_accuracy": 0.19460658580064774,
      "num_tokens": 85646930.0,
      "step": 33790
    },
    {
      "entropy": 6.117756128311157,
      "epoch": 3.9001731102135024,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00035592933980275603,
      "loss": 5.6057,
      "mean_token_accuracy": 0.2055487647652626,
      "num_tokens": 85659305.0,
      "step": 33795
    },
    {
      "entropy": 6.130749702453613,
      "epoch": 3.9007501442585113,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035589083577191276,
      "loss": 5.6281,
      "mean_token_accuracy": 0.21119747310876846,
      "num_tokens": 85671669.0,
      "step": 33800
    },
    {
      "entropy": 6.11531252861023,
      "epoch": 3.9013271783035197,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003558523290203465,
      "loss": 5.6396,
      "mean_token_accuracy": 0.20355551093816757,
      "num_tokens": 85683725.0,
      "step": 33805
    },
    {
      "entropy": 6.1662764072418215,
      "epoch": 3.9019042123485286,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003558138195493522,
      "loss": 5.6128,
      "mean_token_accuracy": 0.20136215686798095,
      "num_tokens": 85695193.0,
      "step": 33810
    },
    {
      "entropy": 6.0813981056213375,
      "epoch": 3.9024812463935374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035577530736022523,
      "loss": 5.6346,
      "mean_token_accuracy": 0.20283491313457488,
      "num_tokens": 85708972.0,
      "step": 33815
    },
    {
      "entropy": 6.105281019210816,
      "epoch": 3.903058280438546,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035573679245426095,
      "loss": 5.5747,
      "mean_token_accuracy": 0.20464861392974854,
      "num_tokens": 85721333.0,
      "step": 33820
    },
    {
      "entropy": 6.002027130126953,
      "epoch": 3.9036353144835543,
      "grad_norm": 0.890625,
      "learning_rate": 0.00035569827483275475,
      "loss": 5.5463,
      "mean_token_accuracy": 0.2111913576722145,
      "num_tokens": 85735248.0,
      "step": 33825
    },
    {
      "entropy": 6.051641178131104,
      "epoch": 3.904212348528563,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035565975449700223,
      "loss": 5.5299,
      "mean_token_accuracy": 0.2117503046989441,
      "num_tokens": 85746575.0,
      "step": 33830
    },
    {
      "entropy": 6.066372632980347,
      "epoch": 3.904789382573572,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003556212314482988,
      "loss": 5.5742,
      "mean_token_accuracy": 0.20425924211740493,
      "num_tokens": 85760086.0,
      "step": 33835
    },
    {
      "entropy": 6.127329301834107,
      "epoch": 3.9053664166185804,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035558270568794044,
      "loss": 5.6624,
      "mean_token_accuracy": 0.2022235855460167,
      "num_tokens": 85771952.0,
      "step": 33840
    },
    {
      "entropy": 6.079333734512329,
      "epoch": 3.9059434506635893,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00035554417721722273,
      "loss": 5.5872,
      "mean_token_accuracy": 0.2000998467206955,
      "num_tokens": 85784736.0,
      "step": 33845
    },
    {
      "entropy": 6.049238538742065,
      "epoch": 3.9065204847085977,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035550564603744163,
      "loss": 5.5478,
      "mean_token_accuracy": 0.21463964134454727,
      "num_tokens": 85797152.0,
      "step": 33850
    },
    {
      "entropy": 6.203660726547241,
      "epoch": 3.9070975187536066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003554671121498931,
      "loss": 5.7027,
      "mean_token_accuracy": 0.1982257753610611,
      "num_tokens": 85807935.0,
      "step": 33855
    },
    {
      "entropy": 6.094870138168335,
      "epoch": 3.907674552798615,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003554285755558732,
      "loss": 5.6192,
      "mean_token_accuracy": 0.20157722681760787,
      "num_tokens": 85819876.0,
      "step": 33860
    },
    {
      "entropy": 6.115404272079468,
      "epoch": 3.908251586843624,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003553900362566782,
      "loss": 5.6322,
      "mean_token_accuracy": 0.2049118772149086,
      "num_tokens": 85832563.0,
      "step": 33865
    },
    {
      "entropy": 6.077067708969116,
      "epoch": 3.9088286208886327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003553514942536042,
      "loss": 5.5502,
      "mean_token_accuracy": 0.21306779980659485,
      "num_tokens": 85845260.0,
      "step": 33870
    },
    {
      "entropy": 6.10785231590271,
      "epoch": 3.909405654933641,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035531294954794764,
      "loss": 5.7037,
      "mean_token_accuracy": 0.1998100146651268,
      "num_tokens": 85857536.0,
      "step": 33875
    },
    {
      "entropy": 6.0291038990020756,
      "epoch": 3.9099826889786495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003552744021410049,
      "loss": 5.4841,
      "mean_token_accuracy": 0.21486638933420182,
      "num_tokens": 85870090.0,
      "step": 33880
    },
    {
      "entropy": 6.054075908660889,
      "epoch": 3.9105597230236584,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003552358520340725,
      "loss": 5.5379,
      "mean_token_accuracy": 0.21507245302200317,
      "num_tokens": 85883363.0,
      "step": 33885
    },
    {
      "entropy": 6.090999794006348,
      "epoch": 3.9111367570686673,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035519729922844705,
      "loss": 5.6359,
      "mean_token_accuracy": 0.20017725676298143,
      "num_tokens": 85896080.0,
      "step": 33890
    },
    {
      "entropy": 6.088406133651733,
      "epoch": 3.9117137911136757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035515874372542527,
      "loss": 5.6332,
      "mean_token_accuracy": 0.2104274958372116,
      "num_tokens": 85908961.0,
      "step": 33895
    },
    {
      "entropy": 6.100196266174317,
      "epoch": 3.912290825158684,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003551201855263041,
      "loss": 5.4948,
      "mean_token_accuracy": 0.21800185590982438,
      "num_tokens": 85920496.0,
      "step": 33900
    },
    {
      "entropy": 6.1130064010620115,
      "epoch": 3.912867859203693,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003550816246323802,
      "loss": 5.6702,
      "mean_token_accuracy": 0.20012075304985047,
      "num_tokens": 85933149.0,
      "step": 33905
    },
    {
      "entropy": 6.041610527038574,
      "epoch": 3.913444893248702,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003550430610449506,
      "loss": 5.5822,
      "mean_token_accuracy": 0.20414189696311952,
      "num_tokens": 85945644.0,
      "step": 33910
    },
    {
      "entropy": 6.100792407989502,
      "epoch": 3.9140219272937102,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035500449476531245,
      "loss": 5.5798,
      "mean_token_accuracy": 0.20422907024621964,
      "num_tokens": 85957796.0,
      "step": 33915
    },
    {
      "entropy": 6.099258279800415,
      "epoch": 3.914598961338719,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003549659257947628,
      "loss": 5.5717,
      "mean_token_accuracy": 0.20179478377103804,
      "num_tokens": 85970186.0,
      "step": 33920
    },
    {
      "entropy": 6.170908880233765,
      "epoch": 3.9151759953837275,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000354927354134599,
      "loss": 5.6642,
      "mean_token_accuracy": 0.2043335035443306,
      "num_tokens": 85984056.0,
      "step": 33925
    },
    {
      "entropy": 6.095730209350586,
      "epoch": 3.9157530294287364,
      "grad_norm": 1.140625,
      "learning_rate": 0.00035488877978611836,
      "loss": 5.5829,
      "mean_token_accuracy": 0.21233381628990172,
      "num_tokens": 85996802.0,
      "step": 33930
    },
    {
      "entropy": 6.120232105255127,
      "epoch": 3.916330063473745,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035485020275061825,
      "loss": 5.6285,
      "mean_token_accuracy": 0.20354021787643434,
      "num_tokens": 86009063.0,
      "step": 33935
    },
    {
      "entropy": 6.054386138916016,
      "epoch": 3.9169070975187537,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003548116230293962,
      "loss": 5.557,
      "mean_token_accuracy": 0.2120618298649788,
      "num_tokens": 86021793.0,
      "step": 33940
    },
    {
      "entropy": 6.1094207763671875,
      "epoch": 3.9174841315637625,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035477304062374993,
      "loss": 5.6342,
      "mean_token_accuracy": 0.2037036806344986,
      "num_tokens": 86033707.0,
      "step": 33945
    },
    {
      "entropy": 6.092756652832032,
      "epoch": 3.918061165608771,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003547344555349771,
      "loss": 5.5615,
      "mean_token_accuracy": 0.2100572556257248,
      "num_tokens": 86045841.0,
      "step": 33950
    },
    {
      "entropy": 5.985999059677124,
      "epoch": 3.9186381996537794,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0003546958677643754,
      "loss": 5.5233,
      "mean_token_accuracy": 0.21066182255744934,
      "num_tokens": 86059875.0,
      "step": 33955
    },
    {
      "entropy": 6.0427368640899655,
      "epoch": 3.919215233698788,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035465727731324277,
      "loss": 5.5958,
      "mean_token_accuracy": 0.2058490887284279,
      "num_tokens": 86073536.0,
      "step": 33960
    },
    {
      "entropy": 6.059695434570313,
      "epoch": 3.919792267743797,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003546186841828772,
      "loss": 5.5058,
      "mean_token_accuracy": 0.22149794548749924,
      "num_tokens": 86085874.0,
      "step": 33965
    },
    {
      "entropy": 6.097062349319458,
      "epoch": 3.9203693017888055,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003545800883745767,
      "loss": 5.6243,
      "mean_token_accuracy": 0.19834984540939332,
      "num_tokens": 86096805.0,
      "step": 33970
    },
    {
      "entropy": 6.156413269042969,
      "epoch": 3.9209463358338144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035454148988963953,
      "loss": 5.6447,
      "mean_token_accuracy": 0.20302644073963166,
      "num_tokens": 86108138.0,
      "step": 33975
    },
    {
      "entropy": 6.123766660690308,
      "epoch": 3.9215233698788228,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003545028887293638,
      "loss": 5.6137,
      "mean_token_accuracy": 0.2032058998942375,
      "num_tokens": 86120329.0,
      "step": 33980
    },
    {
      "entropy": 6.09746994972229,
      "epoch": 3.9221004039238316,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003544642848950479,
      "loss": 5.6125,
      "mean_token_accuracy": 0.20709274858236312,
      "num_tokens": 86133014.0,
      "step": 33985
    },
    {
      "entropy": 6.072454738616943,
      "epoch": 3.92267743796884,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00035442567838799027,
      "loss": 5.5731,
      "mean_token_accuracy": 0.20567707419395448,
      "num_tokens": 86146458.0,
      "step": 33990
    },
    {
      "entropy": 6.055431938171386,
      "epoch": 3.923254472013849,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035438706920948944,
      "loss": 5.5139,
      "mean_token_accuracy": 0.20984887480735778,
      "num_tokens": 86158700.0,
      "step": 33995
    },
    {
      "entropy": 6.124488258361817,
      "epoch": 3.9238315060588573,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035434845736084385,
      "loss": 5.7165,
      "mean_token_accuracy": 0.1972483292222023,
      "num_tokens": 86172038.0,
      "step": 34000
    },
    {
      "entropy": 6.09891791343689,
      "epoch": 3.924408540103866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035430984284335243,
      "loss": 5.5725,
      "mean_token_accuracy": 0.21213222593069075,
      "num_tokens": 86184300.0,
      "step": 34005
    },
    {
      "entropy": 6.028426837921143,
      "epoch": 3.9249855741488746,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00035427122565831384,
      "loss": 5.5198,
      "mean_token_accuracy": 0.21397937536239625,
      "num_tokens": 86198042.0,
      "step": 34010
    },
    {
      "entropy": 6.09245080947876,
      "epoch": 3.9255626081938835,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035423260580702695,
      "loss": 5.6045,
      "mean_token_accuracy": 0.21403784304857254,
      "num_tokens": 86209999.0,
      "step": 34015
    },
    {
      "entropy": 6.111929655075073,
      "epoch": 3.9261396422388923,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003541939832907907,
      "loss": 5.6226,
      "mean_token_accuracy": 0.20569241493940355,
      "num_tokens": 86224503.0,
      "step": 34020
    },
    {
      "entropy": 6.061177492141724,
      "epoch": 3.9267166762839008,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035415535811090404,
      "loss": 5.5962,
      "mean_token_accuracy": 0.20923683792352676,
      "num_tokens": 86238586.0,
      "step": 34025
    },
    {
      "entropy": 6.020795679092407,
      "epoch": 3.927293710328909,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035411673026866636,
      "loss": 5.5637,
      "mean_token_accuracy": 0.20788252353668213,
      "num_tokens": 86252244.0,
      "step": 34030
    },
    {
      "entropy": 6.1387275695800785,
      "epoch": 3.927870744373918,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035407809976537675,
      "loss": 5.653,
      "mean_token_accuracy": 0.2002102926373482,
      "num_tokens": 86263427.0,
      "step": 34035
    },
    {
      "entropy": 6.190771627426147,
      "epoch": 3.928447778418927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035403946660233457,
      "loss": 5.6629,
      "mean_token_accuracy": 0.20265832990407945,
      "num_tokens": 86275972.0,
      "step": 34040
    },
    {
      "entropy": 6.137822818756104,
      "epoch": 3.9290248124639353,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035400083078083913,
      "loss": 5.6717,
      "mean_token_accuracy": 0.19944227039813994,
      "num_tokens": 86289330.0,
      "step": 34045
    },
    {
      "entropy": 6.089179849624633,
      "epoch": 3.929601846508944,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035396219230218994,
      "loss": 5.6048,
      "mean_token_accuracy": 0.20957937389612197,
      "num_tokens": 86301584.0,
      "step": 34050
    },
    {
      "entropy": 6.047177696228028,
      "epoch": 3.9301788805539526,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003539235511676867,
      "loss": 5.5488,
      "mean_token_accuracy": 0.21274682730436326,
      "num_tokens": 86313796.0,
      "step": 34055
    },
    {
      "entropy": 6.0519298076629635,
      "epoch": 3.9307559145989615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035388490737862905,
      "loss": 5.4976,
      "mean_token_accuracy": 0.21602849811315536,
      "num_tokens": 86326379.0,
      "step": 34060
    },
    {
      "entropy": 6.077120113372803,
      "epoch": 3.93133294864397,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003538462609363166,
      "loss": 5.6307,
      "mean_token_accuracy": 0.19990917295217514,
      "num_tokens": 86339119.0,
      "step": 34065
    },
    {
      "entropy": 6.075619554519653,
      "epoch": 3.9319099826889787,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035380761184204945,
      "loss": 5.5616,
      "mean_token_accuracy": 0.2067670777440071,
      "num_tokens": 86350840.0,
      "step": 34070
    },
    {
      "entropy": 6.07916841506958,
      "epoch": 3.932487016733987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003537689600971274,
      "loss": 5.5766,
      "mean_token_accuracy": 0.2027121141552925,
      "num_tokens": 86364520.0,
      "step": 34075
    },
    {
      "entropy": 6.149055099487304,
      "epoch": 3.933064050778996,
      "grad_norm": 0.984375,
      "learning_rate": 0.00035373030570285053,
      "loss": 5.6367,
      "mean_token_accuracy": 0.1999443918466568,
      "num_tokens": 86375888.0,
      "step": 34080
    },
    {
      "entropy": 6.11506757736206,
      "epoch": 3.9336410848240044,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035369164866051885,
      "loss": 5.6036,
      "mean_token_accuracy": 0.20841236412525177,
      "num_tokens": 86388251.0,
      "step": 34085
    },
    {
      "entropy": 6.092437219619751,
      "epoch": 3.9342181188690133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003536529889714327,
      "loss": 5.5961,
      "mean_token_accuracy": 0.21188106387853622,
      "num_tokens": 86401491.0,
      "step": 34090
    },
    {
      "entropy": 6.124885320663452,
      "epoch": 3.934795152914022,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035361432663689237,
      "loss": 5.621,
      "mean_token_accuracy": 0.20653427243232728,
      "num_tokens": 86414931.0,
      "step": 34095
    },
    {
      "entropy": 6.0764488697052,
      "epoch": 3.9353721869590306,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035357566165819814,
      "loss": 5.6112,
      "mean_token_accuracy": 0.19901864975690842,
      "num_tokens": 86426720.0,
      "step": 34100
    },
    {
      "entropy": 6.0422587394714355,
      "epoch": 3.935949221004039,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003535369940366506,
      "loss": 5.518,
      "mean_token_accuracy": 0.2147801086306572,
      "num_tokens": 86439559.0,
      "step": 34105
    },
    {
      "entropy": 6.097068643569946,
      "epoch": 3.936526255049048,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003534983237735503,
      "loss": 5.5462,
      "mean_token_accuracy": 0.20530952513217926,
      "num_tokens": 86450810.0,
      "step": 34110
    },
    {
      "entropy": 6.063716411590576,
      "epoch": 3.9371032890940567,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003534596508701979,
      "loss": 5.6593,
      "mean_token_accuracy": 0.2004273533821106,
      "num_tokens": 86465034.0,
      "step": 34115
    },
    {
      "entropy": 6.044253778457642,
      "epoch": 3.937680323139065,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035342097532789404,
      "loss": 5.441,
      "mean_token_accuracy": 0.2190190151333809,
      "num_tokens": 86477678.0,
      "step": 34120
    },
    {
      "entropy": 5.9745402336120605,
      "epoch": 3.938257357184074,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00035338229714793957,
      "loss": 5.4768,
      "mean_token_accuracy": 0.2211019903421402,
      "num_tokens": 86490209.0,
      "step": 34125
    },
    {
      "entropy": 5.966192817687988,
      "epoch": 3.9388343912290824,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035334361633163557,
      "loss": 5.5014,
      "mean_token_accuracy": 0.2125539854168892,
      "num_tokens": 86504554.0,
      "step": 34130
    },
    {
      "entropy": 6.09911527633667,
      "epoch": 3.9394114252740913,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000353304932880283,
      "loss": 5.6291,
      "mean_token_accuracy": 0.207482248544693,
      "num_tokens": 86516571.0,
      "step": 34135
    },
    {
      "entropy": 6.092695236206055,
      "epoch": 3.9399884593190997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035326624679518284,
      "loss": 5.6512,
      "mean_token_accuracy": 0.1950303226709366,
      "num_tokens": 86529191.0,
      "step": 34140
    },
    {
      "entropy": 6.0369203090667725,
      "epoch": 3.9405654933641086,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003532275580776363,
      "loss": 5.5179,
      "mean_token_accuracy": 0.20641598999500274,
      "num_tokens": 86542109.0,
      "step": 34145
    },
    {
      "entropy": 5.9909604549407955,
      "epoch": 3.9411425274091174,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035318886672894474,
      "loss": 5.4947,
      "mean_token_accuracy": 0.21550468355417252,
      "num_tokens": 86554267.0,
      "step": 34150
    },
    {
      "entropy": 6.089364719390869,
      "epoch": 3.941719561454126,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003531501727504095,
      "loss": 5.5284,
      "mean_token_accuracy": 0.20988584607839583,
      "num_tokens": 86567841.0,
      "step": 34155
    },
    {
      "entropy": 6.098076200485229,
      "epoch": 3.9422965954991342,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035311147614333204,
      "loss": 5.6007,
      "mean_token_accuracy": 0.21065972447395326,
      "num_tokens": 86579729.0,
      "step": 34160
    },
    {
      "entropy": 6.107254505157471,
      "epoch": 3.942873629544143,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003530727769090138,
      "loss": 5.6665,
      "mean_token_accuracy": 0.2033235639333725,
      "num_tokens": 86592375.0,
      "step": 34165
    },
    {
      "entropy": 6.09987096786499,
      "epoch": 3.943450663589152,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00035303407504875655,
      "loss": 5.5452,
      "mean_token_accuracy": 0.21543767750263215,
      "num_tokens": 86604934.0,
      "step": 34170
    },
    {
      "entropy": 6.056863641738891,
      "epoch": 3.9440276976341604,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000352995370563862,
      "loss": 5.5802,
      "mean_token_accuracy": 0.20921788960695267,
      "num_tokens": 86618705.0,
      "step": 34175
    },
    {
      "entropy": 6.134924507141113,
      "epoch": 3.944604731679169,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003529566634556319,
      "loss": 5.5335,
      "mean_token_accuracy": 0.21077499389648438,
      "num_tokens": 86631116.0,
      "step": 34180
    },
    {
      "entropy": 6.123343896865845,
      "epoch": 3.9451817657241777,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000352917953725368,
      "loss": 5.6637,
      "mean_token_accuracy": 0.20194773823022844,
      "num_tokens": 86643649.0,
      "step": 34185
    },
    {
      "entropy": 6.106112337112426,
      "epoch": 3.9457587997691865,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003528792413743725,
      "loss": 5.5815,
      "mean_token_accuracy": 0.21331805288791655,
      "num_tokens": 86656269.0,
      "step": 34190
    },
    {
      "entropy": 6.1351221084594725,
      "epoch": 3.946335833814195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003528405264039475,
      "loss": 5.6044,
      "mean_token_accuracy": 0.20997445583343505,
      "num_tokens": 86669245.0,
      "step": 34195
    },
    {
      "entropy": 6.090846061706543,
      "epoch": 3.946912867859204,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000352801808815395,
      "loss": 5.6061,
      "mean_token_accuracy": 0.20263096392154695,
      "num_tokens": 86682005.0,
      "step": 34200
    },
    {
      "entropy": 5.996529388427734,
      "epoch": 3.9474899019042122,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035276308861001724,
      "loss": 5.5134,
      "mean_token_accuracy": 0.2083194762468338,
      "num_tokens": 86695468.0,
      "step": 34205
    },
    {
      "entropy": 6.141702604293823,
      "epoch": 3.948066935949221,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035272436578911667,
      "loss": 5.6828,
      "mean_token_accuracy": 0.19777455180883408,
      "num_tokens": 86708021.0,
      "step": 34210
    },
    {
      "entropy": 6.106406021118164,
      "epoch": 3.9486439699942295,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003526856403539957,
      "loss": 5.6038,
      "mean_token_accuracy": 0.21075500696897506,
      "num_tokens": 86720427.0,
      "step": 34215
    },
    {
      "entropy": 6.138281202316284,
      "epoch": 3.9492210040392384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035264691230595677,
      "loss": 5.6208,
      "mean_token_accuracy": 0.20158367455005646,
      "num_tokens": 86732562.0,
      "step": 34220
    },
    {
      "entropy": 6.0598243236541744,
      "epoch": 3.9497980380842472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003526081816463025,
      "loss": 5.5617,
      "mean_token_accuracy": 0.2076629713177681,
      "num_tokens": 86746303.0,
      "step": 34225
    },
    {
      "entropy": 6.0711382865905765,
      "epoch": 3.9503750721292556,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003525694483763356,
      "loss": 5.6141,
      "mean_token_accuracy": 0.20496369153261185,
      "num_tokens": 86759317.0,
      "step": 34230
    },
    {
      "entropy": 6.152827739715576,
      "epoch": 3.950952106174264,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035253071249735884,
      "loss": 5.6332,
      "mean_token_accuracy": 0.20343178659677505,
      "num_tokens": 86771103.0,
      "step": 34235
    },
    {
      "entropy": 6.086332988739014,
      "epoch": 3.951529140219273,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003524919740106751,
      "loss": 5.6096,
      "mean_token_accuracy": 0.20374724566936492,
      "num_tokens": 86783210.0,
      "step": 34240
    },
    {
      "entropy": 6.065157651901245,
      "epoch": 3.952106174264282,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003524532329175873,
      "loss": 5.5341,
      "mean_token_accuracy": 0.2141427293419838,
      "num_tokens": 86795293.0,
      "step": 34245
    },
    {
      "entropy": 6.102271509170532,
      "epoch": 3.95268320830929,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003524144892193984,
      "loss": 5.6131,
      "mean_token_accuracy": 0.2004251480102539,
      "num_tokens": 86807526.0,
      "step": 34250
    },
    {
      "entropy": 6.031961679458618,
      "epoch": 3.953260242354299,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003523757429174117,
      "loss": 5.523,
      "mean_token_accuracy": 0.21525313705205917,
      "num_tokens": 86821568.0,
      "step": 34255
    },
    {
      "entropy": 6.150691938400269,
      "epoch": 3.9538372763993075,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035233699401293037,
      "loss": 5.6832,
      "mean_token_accuracy": 0.20531666427850723,
      "num_tokens": 86834504.0,
      "step": 34260
    },
    {
      "entropy": 6.072914934158325,
      "epoch": 3.9544143104443163,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035229824250725757,
      "loss": 5.6305,
      "mean_token_accuracy": 0.20377937704324722,
      "num_tokens": 86847212.0,
      "step": 34265
    },
    {
      "entropy": 6.142857217788697,
      "epoch": 3.9549913444893248,
      "grad_norm": 1.0,
      "learning_rate": 0.00035225948840169675,
      "loss": 5.6421,
      "mean_token_accuracy": 0.20208427757024766,
      "num_tokens": 86859493.0,
      "step": 34270
    },
    {
      "entropy": 6.0962928295135494,
      "epoch": 3.9555683785343336,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035222073169755144,
      "loss": 5.4915,
      "mean_token_accuracy": 0.2181245669722557,
      "num_tokens": 86873409.0,
      "step": 34275
    },
    {
      "entropy": 6.029172086715699,
      "epoch": 3.956145412579342,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003521819723961253,
      "loss": 5.6175,
      "mean_token_accuracy": 0.2116535097360611,
      "num_tokens": 86886715.0,
      "step": 34280
    },
    {
      "entropy": 6.0778555393219,
      "epoch": 3.956722446624351,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003521432104987217,
      "loss": 5.5307,
      "mean_token_accuracy": 0.20899896770715715,
      "num_tokens": 86898415.0,
      "step": 34285
    },
    {
      "entropy": 6.025340032577515,
      "epoch": 3.9572994806693593,
      "grad_norm": 1.0,
      "learning_rate": 0.00035210444600664454,
      "loss": 5.5369,
      "mean_token_accuracy": 0.2082712933421135,
      "num_tokens": 86911399.0,
      "step": 34290
    },
    {
      "entropy": 5.994102668762207,
      "epoch": 3.957876514714368,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035206567892119773,
      "loss": 5.5092,
      "mean_token_accuracy": 0.2213461682200432,
      "num_tokens": 86923935.0,
      "step": 34295
    },
    {
      "entropy": 6.105885982513428,
      "epoch": 3.958453548759377,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003520269092436851,
      "loss": 5.6046,
      "mean_token_accuracy": 0.2052093654870987,
      "num_tokens": 86937116.0,
      "step": 34300
    },
    {
      "entropy": 6.126477241516113,
      "epoch": 3.9590305828043855,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003519881369754106,
      "loss": 5.6757,
      "mean_token_accuracy": 0.20096003711223603,
      "num_tokens": 86949701.0,
      "step": 34305
    },
    {
      "entropy": 6.085348510742188,
      "epoch": 3.959607616849394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035194936211767824,
      "loss": 5.5577,
      "mean_token_accuracy": 0.20440681129693986,
      "num_tokens": 86963431.0,
      "step": 34310
    },
    {
      "entropy": 6.157261848449707,
      "epoch": 3.9601846508944027,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003519105846717924,
      "loss": 5.6978,
      "mean_token_accuracy": 0.201397080719471,
      "num_tokens": 86975965.0,
      "step": 34315
    },
    {
      "entropy": 6.039133930206299,
      "epoch": 3.9607616849394116,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035187180463905733,
      "loss": 5.5313,
      "mean_token_accuracy": 0.21336714625358583,
      "num_tokens": 86988575.0,
      "step": 34320
    },
    {
      "entropy": 6.0220293521881105,
      "epoch": 3.96133871898442,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035183302202077725,
      "loss": 5.5428,
      "mean_token_accuracy": 0.21505910903215408,
      "num_tokens": 87000623.0,
      "step": 34325
    },
    {
      "entropy": 6.068096017837524,
      "epoch": 3.961915753029429,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035179423681825663,
      "loss": 5.559,
      "mean_token_accuracy": 0.2072800800204277,
      "num_tokens": 87013129.0,
      "step": 34330
    },
    {
      "entropy": 6.09013090133667,
      "epoch": 3.9624927870744373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003517554490327999,
      "loss": 5.6564,
      "mean_token_accuracy": 0.19848361909389495,
      "num_tokens": 87024906.0,
      "step": 34335
    },
    {
      "entropy": 6.125682687759399,
      "epoch": 3.963069821119446,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003517166586657119,
      "loss": 5.691,
      "mean_token_accuracy": 0.2008557692170143,
      "num_tokens": 87037974.0,
      "step": 34340
    },
    {
      "entropy": 6.1886663913726805,
      "epoch": 3.9636468551644546,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003516778657182971,
      "loss": 5.6453,
      "mean_token_accuracy": 0.20642592906951904,
      "num_tokens": 87051149.0,
      "step": 34345
    },
    {
      "entropy": 6.036720275878906,
      "epoch": 3.9642238892094634,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035163907019186046,
      "loss": 5.5614,
      "mean_token_accuracy": 0.20383604764938354,
      "num_tokens": 87062682.0,
      "step": 34350
    },
    {
      "entropy": 6.147052145004272,
      "epoch": 3.964800923254472,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003516002720877067,
      "loss": 5.6469,
      "mean_token_accuracy": 0.19538031965494157,
      "num_tokens": 87076150.0,
      "step": 34355
    },
    {
      "entropy": 5.9898974895477295,
      "epoch": 3.9653779572994807,
      "grad_norm": 0.7890625,
      "learning_rate": 0.0003515614714071408,
      "loss": 5.4767,
      "mean_token_accuracy": 0.21158430874347686,
      "num_tokens": 87089717.0,
      "step": 34360
    },
    {
      "entropy": 6.1503664493560795,
      "epoch": 3.965954991344489,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000351522668151468,
      "loss": 5.6275,
      "mean_token_accuracy": 0.2006218194961548,
      "num_tokens": 87101621.0,
      "step": 34365
    },
    {
      "entropy": 6.031438827514648,
      "epoch": 3.966532025389498,
      "grad_norm": 0.875,
      "learning_rate": 0.0003514838623219932,
      "loss": 5.5821,
      "mean_token_accuracy": 0.20545547604560851,
      "num_tokens": 87114347.0,
      "step": 34370
    },
    {
      "entropy": 6.134225654602051,
      "epoch": 3.967109059434507,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003514450539200216,
      "loss": 5.6732,
      "mean_token_accuracy": 0.2022982656955719,
      "num_tokens": 87126197.0,
      "step": 34375
    },
    {
      "entropy": 6.088818359375,
      "epoch": 3.9676860934795153,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003514062429468586,
      "loss": 5.6006,
      "mean_token_accuracy": 0.20101480633020402,
      "num_tokens": 87137307.0,
      "step": 34380
    },
    {
      "entropy": 6.098271131515503,
      "epoch": 3.9682631275245237,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003513674294038096,
      "loss": 5.5939,
      "mean_token_accuracy": 0.2085874229669571,
      "num_tokens": 87149892.0,
      "step": 34385
    },
    {
      "entropy": 6.167271709442138,
      "epoch": 3.9688401615695326,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003513286132921801,
      "loss": 5.6051,
      "mean_token_accuracy": 0.2024668499827385,
      "num_tokens": 87161112.0,
      "step": 34390
    },
    {
      "entropy": 6.1227614402771,
      "epoch": 3.9694171956145414,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035128979461327553,
      "loss": 5.6074,
      "mean_token_accuracy": 0.20293567031621934,
      "num_tokens": 87171515.0,
      "step": 34395
    },
    {
      "entropy": 6.102714252471924,
      "epoch": 3.96999422965955,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00035125097336840167,
      "loss": 5.6258,
      "mean_token_accuracy": 0.20371688455343245,
      "num_tokens": 87183792.0,
      "step": 34400
    },
    {
      "entropy": 6.1023924350738525,
      "epoch": 3.9705712637045587,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00035121214955886415,
      "loss": 5.5976,
      "mean_token_accuracy": 0.20437540113925934,
      "num_tokens": 87195737.0,
      "step": 34405
    },
    {
      "entropy": 6.084439277648926,
      "epoch": 3.971148297749567,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035117332318596885,
      "loss": 5.6293,
      "mean_token_accuracy": 0.20420658588409424,
      "num_tokens": 87208522.0,
      "step": 34410
    },
    {
      "entropy": 6.045976877212524,
      "epoch": 3.971725331794576,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003511344942510217,
      "loss": 5.5711,
      "mean_token_accuracy": 0.21201349794864655,
      "num_tokens": 87220893.0,
      "step": 34415
    },
    {
      "entropy": 6.138035106658935,
      "epoch": 3.9723023658395844,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035109566275532854,
      "loss": 5.6308,
      "mean_token_accuracy": 0.20123392194509507,
      "num_tokens": 87233245.0,
      "step": 34420
    },
    {
      "entropy": 6.062977504730225,
      "epoch": 3.9728793998845933,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035105682870019574,
      "loss": 5.6715,
      "mean_token_accuracy": 0.19215962737798692,
      "num_tokens": 87246029.0,
      "step": 34425
    },
    {
      "entropy": 6.130942106246948,
      "epoch": 3.9734564339296017,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003510179920869291,
      "loss": 5.5993,
      "mean_token_accuracy": 0.20385899245738984,
      "num_tokens": 87259740.0,
      "step": 34430
    },
    {
      "entropy": 6.108812761306763,
      "epoch": 3.9740334679746105,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035097915291683524,
      "loss": 5.6076,
      "mean_token_accuracy": 0.2115843877196312,
      "num_tokens": 87272771.0,
      "step": 34435
    },
    {
      "entropy": 6.006356811523437,
      "epoch": 3.974610502019619,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003509403111912202,
      "loss": 5.501,
      "mean_token_accuracy": 0.22150854468345643,
      "num_tokens": 87285684.0,
      "step": 34440
    },
    {
      "entropy": 6.116124773025513,
      "epoch": 3.975187536064628,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003509014669113905,
      "loss": 5.6191,
      "mean_token_accuracy": 0.20603700429201127,
      "num_tokens": 87298722.0,
      "step": 34445
    },
    {
      "entropy": 6.028958749771118,
      "epoch": 3.9757645701096367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003508626200786527,
      "loss": 5.5554,
      "mean_token_accuracy": 0.20523949712514877,
      "num_tokens": 87311668.0,
      "step": 34450
    },
    {
      "entropy": 6.143246078491211,
      "epoch": 3.976341604154645,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035082377069431343,
      "loss": 5.6738,
      "mean_token_accuracy": 0.2047056719660759,
      "num_tokens": 87324786.0,
      "step": 34455
    },
    {
      "entropy": 6.066857957839966,
      "epoch": 3.9769186381996535,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035078491875967916,
      "loss": 5.5362,
      "mean_token_accuracy": 0.21922508627176285,
      "num_tokens": 87337873.0,
      "step": 34460
    },
    {
      "entropy": 6.1323166847229,
      "epoch": 3.9774956722446624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003507460642760569,
      "loss": 5.7295,
      "mean_token_accuracy": 0.20169539302587508,
      "num_tokens": 87350464.0,
      "step": 34465
    },
    {
      "entropy": 6.044235801696777,
      "epoch": 3.9780727062896712,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003507072072447534,
      "loss": 5.5242,
      "mean_token_accuracy": 0.20335896462202072,
      "num_tokens": 87362448.0,
      "step": 34470
    },
    {
      "entropy": 6.102519607543945,
      "epoch": 3.9786497403346797,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035066834766707546,
      "loss": 5.5532,
      "mean_token_accuracy": 0.21501004099845886,
      "num_tokens": 87374456.0,
      "step": 34475
    },
    {
      "entropy": 6.079614162445068,
      "epoch": 3.9792267743796885,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035062948554433025,
      "loss": 5.5441,
      "mean_token_accuracy": 0.20899610072374344,
      "num_tokens": 87387467.0,
      "step": 34480
    },
    {
      "entropy": 6.128240442276001,
      "epoch": 3.979803808424697,
      "grad_norm": 0.921875,
      "learning_rate": 0.00035059062087782497,
      "loss": 5.6293,
      "mean_token_accuracy": 0.20750930458307265,
      "num_tokens": 87402236.0,
      "step": 34485
    },
    {
      "entropy": 6.06376953125,
      "epoch": 3.980380842469706,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003505517536688667,
      "loss": 5.5046,
      "mean_token_accuracy": 0.21335405111312866,
      "num_tokens": 87416514.0,
      "step": 34490
    },
    {
      "entropy": 6.0376605033874515,
      "epoch": 3.9809578765147142,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003505128839187626,
      "loss": 5.574,
      "mean_token_accuracy": 0.2085227608680725,
      "num_tokens": 87428474.0,
      "step": 34495
    },
    {
      "entropy": 6.040627193450928,
      "epoch": 3.981534910559723,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003504740116288203,
      "loss": 5.6181,
      "mean_token_accuracy": 0.203532937169075,
      "num_tokens": 87440945.0,
      "step": 34500
    },
    {
      "entropy": 6.063632154464722,
      "epoch": 3.982111944604732,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000350435136800347,
      "loss": 5.5314,
      "mean_token_accuracy": 0.21654371172189713,
      "num_tokens": 87454193.0,
      "step": 34505
    },
    {
      "entropy": 6.1012945652008055,
      "epoch": 3.9826889786497404,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0003503962594346504,
      "loss": 5.5648,
      "mean_token_accuracy": 0.20970981568098068,
      "num_tokens": 87467775.0,
      "step": 34510
    },
    {
      "entropy": 6.013328218460083,
      "epoch": 3.983266012694749,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003503573795330381,
      "loss": 5.5939,
      "mean_token_accuracy": 0.2044184222817421,
      "num_tokens": 87481210.0,
      "step": 34515
    },
    {
      "entropy": 6.112621450424195,
      "epoch": 3.9838430467397576,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003503184970968176,
      "loss": 5.6126,
      "mean_token_accuracy": 0.2013202890753746,
      "num_tokens": 87493464.0,
      "step": 34520
    },
    {
      "entropy": 6.059024953842163,
      "epoch": 3.9844200807847665,
      "grad_norm": 0.875,
      "learning_rate": 0.00035027961212729704,
      "loss": 5.5301,
      "mean_token_accuracy": 0.21107590794563294,
      "num_tokens": 87505980.0,
      "step": 34525
    },
    {
      "entropy": 6.004796171188355,
      "epoch": 3.984997114829775,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003502407246257841,
      "loss": 5.5083,
      "mean_token_accuracy": 0.20867787301540375,
      "num_tokens": 87518941.0,
      "step": 34530
    },
    {
      "entropy": 6.128394603729248,
      "epoch": 3.9855741488747833,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035020183459358675,
      "loss": 5.664,
      "mean_token_accuracy": 0.20174337327480316,
      "num_tokens": 87532342.0,
      "step": 34535
    },
    {
      "entropy": 6.130295324325561,
      "epoch": 3.986151182919792,
      "grad_norm": 0.96875,
      "learning_rate": 0.000350162942032013,
      "loss": 5.6104,
      "mean_token_accuracy": 0.21007633656263353,
      "num_tokens": 87544921.0,
      "step": 34540
    },
    {
      "entropy": 6.10325517654419,
      "epoch": 3.986728216964801,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035012404694237105,
      "loss": 5.6131,
      "mean_token_accuracy": 0.20724945962429048,
      "num_tokens": 87559131.0,
      "step": 34545
    },
    {
      "entropy": 6.081124210357666,
      "epoch": 3.9873052510098095,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035008514932596905,
      "loss": 5.5983,
      "mean_token_accuracy": 0.20876982808113098,
      "num_tokens": 87571776.0,
      "step": 34550
    },
    {
      "entropy": 6.136280298233032,
      "epoch": 3.9878822850548183,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035004624918411547,
      "loss": 5.6615,
      "mean_token_accuracy": 0.1971665322780609,
      "num_tokens": 87582901.0,
      "step": 34555
    },
    {
      "entropy": 6.064534664154053,
      "epoch": 3.9884593190998268,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003500073465181185,
      "loss": 5.6144,
      "mean_token_accuracy": 0.20179163813591003,
      "num_tokens": 87596225.0,
      "step": 34560
    },
    {
      "entropy": 6.172553634643554,
      "epoch": 3.9890363531448356,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003499684413292867,
      "loss": 5.6469,
      "mean_token_accuracy": 0.2037851795554161,
      "num_tokens": 87608977.0,
      "step": 34565
    },
    {
      "entropy": 6.137198114395142,
      "epoch": 3.989613387189844,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003499295336189286,
      "loss": 5.6596,
      "mean_token_accuracy": 0.20181216448545455,
      "num_tokens": 87622132.0,
      "step": 34570
    },
    {
      "entropy": 6.132321453094482,
      "epoch": 3.990190421234853,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034989062338835293,
      "loss": 5.6279,
      "mean_token_accuracy": 0.19931786209344865,
      "num_tokens": 87634939.0,
      "step": 34575
    },
    {
      "entropy": 5.941410684585572,
      "epoch": 3.9907674552798618,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034985171063886826,
      "loss": 5.4978,
      "mean_token_accuracy": 0.21616897284984588,
      "num_tokens": 87647323.0,
      "step": 34580
    },
    {
      "entropy": 5.928447341918945,
      "epoch": 3.99134448932487,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003498127953717835,
      "loss": 5.3827,
      "mean_token_accuracy": 0.22963026016950608,
      "num_tokens": 87660426.0,
      "step": 34585
    },
    {
      "entropy": 6.096912479400634,
      "epoch": 3.9919215233698786,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003497738775884075,
      "loss": 5.6128,
      "mean_token_accuracy": 0.2134929731488228,
      "num_tokens": 87673445.0,
      "step": 34590
    },
    {
      "entropy": 6.094749689102173,
      "epoch": 3.9924985574148875,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034973495729004935,
      "loss": 5.6564,
      "mean_token_accuracy": 0.20441095978021623,
      "num_tokens": 87686970.0,
      "step": 34595
    },
    {
      "entropy": 6.039253807067871,
      "epoch": 3.9930755914598963,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000349696034478018,
      "loss": 5.634,
      "mean_token_accuracy": 0.20482555478811265,
      "num_tokens": 87699062.0,
      "step": 34600
    },
    {
      "entropy": 6.185580396652222,
      "epoch": 3.9936526255049047,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034965710915362257,
      "loss": 5.6252,
      "mean_token_accuracy": 0.2027628555893898,
      "num_tokens": 87711329.0,
      "step": 34605
    },
    {
      "entropy": 6.097812128067017,
      "epoch": 3.9942296595499136,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003496181813181724,
      "loss": 5.5908,
      "mean_token_accuracy": 0.20764925926923752,
      "num_tokens": 87724650.0,
      "step": 34610
    },
    {
      "entropy": 6.012580919265747,
      "epoch": 3.994806693594922,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003495792509729768,
      "loss": 5.5586,
      "mean_token_accuracy": 0.21214479058980942,
      "num_tokens": 87738230.0,
      "step": 34615
    },
    {
      "entropy": 6.027355337142945,
      "epoch": 3.995383727639931,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003495403181193451,
      "loss": 5.4734,
      "mean_token_accuracy": 0.21015517860651017,
      "num_tokens": 87751716.0,
      "step": 34620
    },
    {
      "entropy": 6.214514446258545,
      "epoch": 3.9959607616849393,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003495013827585868,
      "loss": 5.6981,
      "mean_token_accuracy": 0.19951127469539642,
      "num_tokens": 87764626.0,
      "step": 34625
    },
    {
      "entropy": 6.107736492156983,
      "epoch": 3.996537795729948,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034946244489201156,
      "loss": 5.6938,
      "mean_token_accuracy": 0.1990364745259285,
      "num_tokens": 87776868.0,
      "step": 34630
    },
    {
      "entropy": 6.159733629226684,
      "epoch": 3.9971148297749566,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034942350452092896,
      "loss": 5.6297,
      "mean_token_accuracy": 0.203691266477108,
      "num_tokens": 87788870.0,
      "step": 34635
    },
    {
      "entropy": 6.105396318435669,
      "epoch": 3.9976918638199654,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003493845616466488,
      "loss": 5.5772,
      "mean_token_accuracy": 0.20774596631526948,
      "num_tokens": 87801306.0,
      "step": 34640
    },
    {
      "entropy": 6.0374228954315186,
      "epoch": 3.998268897864974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034934561627048076,
      "loss": 5.5535,
      "mean_token_accuracy": 0.20272466987371446,
      "num_tokens": 87814320.0,
      "step": 34645
    },
    {
      "entropy": 6.141267395019531,
      "epoch": 3.9988459319099827,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003493066683937349,
      "loss": 5.6372,
      "mean_token_accuracy": 0.20017722696065904,
      "num_tokens": 87826939.0,
      "step": 34650
    },
    {
      "entropy": 6.091874980926514,
      "epoch": 3.9994229659549916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003492677180177212,
      "loss": 5.5658,
      "mean_token_accuracy": 0.20687366873025895,
      "num_tokens": 87839769.0,
      "step": 34655
    },
    {
      "entropy": 6.094721698760987,
      "epoch": 4.0,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003492287651437496,
      "loss": 5.5966,
      "mean_token_accuracy": 0.20685495734214782,
      "num_tokens": 87852468.0,
      "step": 34660
    },
    {
      "entropy": 6.07154107093811,
      "epoch": 4.000577034045008,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034918980977313035,
      "loss": 5.5197,
      "mean_token_accuracy": 0.20995810776948928,
      "num_tokens": 87864594.0,
      "step": 34665
    },
    {
      "entropy": 6.073631811141968,
      "epoch": 4.001154068090018,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003491508519071738,
      "loss": 5.554,
      "mean_token_accuracy": 0.2091823026537895,
      "num_tokens": 87878129.0,
      "step": 34670
    },
    {
      "entropy": 6.096306991577149,
      "epoch": 4.001731102135026,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034911189154719017,
      "loss": 5.5819,
      "mean_token_accuracy": 0.207656329870224,
      "num_tokens": 87891110.0,
      "step": 34675
    },
    {
      "entropy": 6.062782049179077,
      "epoch": 4.002308136180035,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003490729286944899,
      "loss": 5.494,
      "mean_token_accuracy": 0.20850566476583482,
      "num_tokens": 87903015.0,
      "step": 34680
    },
    {
      "entropy": 6.093405628204346,
      "epoch": 4.002885170225043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003490339633503833,
      "loss": 5.5535,
      "mean_token_accuracy": 0.20809932947158813,
      "num_tokens": 87916832.0,
      "step": 34685
    },
    {
      "entropy": 6.122495079040528,
      "epoch": 4.003462204270052,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034899499551618125,
      "loss": 5.4839,
      "mean_token_accuracy": 0.21101561933755875,
      "num_tokens": 87929187.0,
      "step": 34690
    },
    {
      "entropy": 6.044330263137818,
      "epoch": 4.004039238315061,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034895602519319425,
      "loss": 5.4949,
      "mean_token_accuracy": 0.2122476115822792,
      "num_tokens": 87941966.0,
      "step": 34695
    },
    {
      "entropy": 6.04209156036377,
      "epoch": 4.004616272360069,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003489170523827331,
      "loss": 5.5199,
      "mean_token_accuracy": 0.20395033061504364,
      "num_tokens": 87953683.0,
      "step": 34700
    },
    {
      "entropy": 6.103100967407227,
      "epoch": 4.0051933064050775,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003488780770861086,
      "loss": 5.5468,
      "mean_token_accuracy": 0.20174019634723664,
      "num_tokens": 87965394.0,
      "step": 34705
    },
    {
      "entropy": 6.037567138671875,
      "epoch": 4.005770340450087,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034883909930463167,
      "loss": 5.4428,
      "mean_token_accuracy": 0.2161940738558769,
      "num_tokens": 87978478.0,
      "step": 34710
    },
    {
      "entropy": 5.922917222976684,
      "epoch": 4.006347374495095,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00034880011903961337,
      "loss": 5.3963,
      "mean_token_accuracy": 0.22060333490371703,
      "num_tokens": 87992029.0,
      "step": 34715
    },
    {
      "entropy": 6.0579345703125,
      "epoch": 4.006924408540104,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003487611362923646,
      "loss": 5.4871,
      "mean_token_accuracy": 0.21774370223283768,
      "num_tokens": 88004589.0,
      "step": 34720
    },
    {
      "entropy": 6.050927782058716,
      "epoch": 4.007501442585112,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003487221510641967,
      "loss": 5.5142,
      "mean_token_accuracy": 0.2204449087381363,
      "num_tokens": 88018711.0,
      "step": 34725
    },
    {
      "entropy": 6.143491458892822,
      "epoch": 4.008078476630121,
      "grad_norm": 1.0,
      "learning_rate": 0.0003486831633564209,
      "loss": 5.563,
      "mean_token_accuracy": 0.21282918155193328,
      "num_tokens": 88030944.0,
      "step": 34730
    },
    {
      "entropy": 6.090694427490234,
      "epoch": 4.00865551067513,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003486441731703485,
      "loss": 5.5893,
      "mean_token_accuracy": 0.2057602122426033,
      "num_tokens": 88044283.0,
      "step": 34735
    },
    {
      "entropy": 6.053262996673584,
      "epoch": 4.009232544720138,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003486051805072909,
      "loss": 5.5303,
      "mean_token_accuracy": 0.21125535070896148,
      "num_tokens": 88056658.0,
      "step": 34740
    },
    {
      "entropy": 6.114913654327393,
      "epoch": 4.0098095787651475,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003485661853685596,
      "loss": 5.6682,
      "mean_token_accuracy": 0.20412778109312057,
      "num_tokens": 88069880.0,
      "step": 34745
    },
    {
      "entropy": 6.189751148223877,
      "epoch": 4.010386612810156,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003485271877554662,
      "loss": 5.5933,
      "mean_token_accuracy": 0.20338854640722276,
      "num_tokens": 88082350.0,
      "step": 34750
    },
    {
      "entropy": 6.084666109085083,
      "epoch": 4.010963646855164,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034848818766932234,
      "loss": 5.5135,
      "mean_token_accuracy": 0.21101275831460953,
      "num_tokens": 88095330.0,
      "step": 34755
    },
    {
      "entropy": 6.0886983394622805,
      "epoch": 4.011540680900173,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003484491851114399,
      "loss": 5.5631,
      "mean_token_accuracy": 0.20131635814905166,
      "num_tokens": 88106689.0,
      "step": 34760
    },
    {
      "entropy": 6.088760423660278,
      "epoch": 4.012117714945182,
      "grad_norm": 1.0,
      "learning_rate": 0.00034841018008313054,
      "loss": 5.5676,
      "mean_token_accuracy": 0.2107717737555504,
      "num_tokens": 88118917.0,
      "step": 34765
    },
    {
      "entropy": 6.123544502258301,
      "epoch": 4.0126947489901905,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003483711725857062,
      "loss": 5.6126,
      "mean_token_accuracy": 0.20976841747760772,
      "num_tokens": 88131857.0,
      "step": 34770
    },
    {
      "entropy": 6.041328191757202,
      "epoch": 4.013271783035199,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034833216262047896,
      "loss": 5.4697,
      "mean_token_accuracy": 0.2147274672985077,
      "num_tokens": 88144197.0,
      "step": 34775
    },
    {
      "entropy": 6.042860412597657,
      "epoch": 4.013848817080207,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00034829315018876085,
      "loss": 5.5192,
      "mean_token_accuracy": 0.21206455826759338,
      "num_tokens": 88157091.0,
      "step": 34780
    },
    {
      "entropy": 6.034541320800781,
      "epoch": 4.014425851125217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000348254135291864,
      "loss": 5.4621,
      "mean_token_accuracy": 0.21467296183109283,
      "num_tokens": 88171332.0,
      "step": 34785
    },
    {
      "entropy": 6.137015628814697,
      "epoch": 4.015002885170225,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003482151179311008,
      "loss": 5.6597,
      "mean_token_accuracy": 0.19899375289678572,
      "num_tokens": 88183379.0,
      "step": 34790
    },
    {
      "entropy": 6.166791248321533,
      "epoch": 4.0155799192152335,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034817609810778335,
      "loss": 5.5885,
      "mean_token_accuracy": 0.21038224995136262,
      "num_tokens": 88195899.0,
      "step": 34795
    },
    {
      "entropy": 6.099807024002075,
      "epoch": 4.016156953260242,
      "grad_norm": 1.0,
      "learning_rate": 0.00034813707582322426,
      "loss": 5.4833,
      "mean_token_accuracy": 0.20996502935886383,
      "num_tokens": 88207395.0,
      "step": 34800
    },
    {
      "entropy": 6.047596263885498,
      "epoch": 4.016733987305251,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034809805107873594,
      "loss": 5.5208,
      "mean_token_accuracy": 0.2173555999994278,
      "num_tokens": 88220157.0,
      "step": 34805
    },
    {
      "entropy": 6.144260120391846,
      "epoch": 4.01731102135026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000348059023875631,
      "loss": 5.6472,
      "mean_token_accuracy": 0.20201577693223954,
      "num_tokens": 88233355.0,
      "step": 34810
    },
    {
      "entropy": 6.18580207824707,
      "epoch": 4.017888055395268,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000348019994215222,
      "loss": 5.5808,
      "mean_token_accuracy": 0.20198184996843338,
      "num_tokens": 88244906.0,
      "step": 34815
    },
    {
      "entropy": 6.142538738250733,
      "epoch": 4.018465089440277,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00034798096209882185,
      "loss": 5.5866,
      "mean_token_accuracy": 0.20428550243377686,
      "num_tokens": 88257777.0,
      "step": 34820
    },
    {
      "entropy": 5.976137924194336,
      "epoch": 4.019042123485286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003479419275277433,
      "loss": 5.4636,
      "mean_token_accuracy": 0.21240369230508804,
      "num_tokens": 88270532.0,
      "step": 34825
    },
    {
      "entropy": 6.037280797958374,
      "epoch": 4.019619157530294,
      "grad_norm": 2.28125,
      "learning_rate": 0.0003479028905032992,
      "loss": 5.4183,
      "mean_token_accuracy": 0.22472108155488968,
      "num_tokens": 88283084.0,
      "step": 34830
    },
    {
      "entropy": 6.03937029838562,
      "epoch": 4.020196191575303,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034786385102680264,
      "loss": 5.5017,
      "mean_token_accuracy": 0.21006458401679992,
      "num_tokens": 88296161.0,
      "step": 34835
    },
    {
      "entropy": 6.065213918685913,
      "epoch": 4.020773225620312,
      "grad_norm": 0.890625,
      "learning_rate": 0.00034782480909956653,
      "loss": 5.51,
      "mean_token_accuracy": 0.2059015691280365,
      "num_tokens": 88309008.0,
      "step": 34840
    },
    {
      "entropy": 6.134440469741821,
      "epoch": 4.02135025966532,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003477857647229043,
      "loss": 5.5697,
      "mean_token_accuracy": 0.20829159915447235,
      "num_tokens": 88323420.0,
      "step": 34845
    },
    {
      "entropy": 5.983515739440918,
      "epoch": 4.021927293710329,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003477467178981289,
      "loss": 5.3734,
      "mean_token_accuracy": 0.23136326670646667,
      "num_tokens": 88337154.0,
      "step": 34850
    },
    {
      "entropy": 6.010545635223389,
      "epoch": 4.022504327755337,
      "grad_norm": 1.171875,
      "learning_rate": 0.00034770766862655386,
      "loss": 5.4444,
      "mean_token_accuracy": 0.21841942369937897,
      "num_tokens": 88350397.0,
      "step": 34855
    },
    {
      "entropy": 6.070789909362793,
      "epoch": 4.0230813618003465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003476686169094925,
      "loss": 5.5244,
      "mean_token_accuracy": 0.21290499418973924,
      "num_tokens": 88363848.0,
      "step": 34860
    },
    {
      "entropy": 6.058571481704712,
      "epoch": 4.023658395845355,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00034762956274825827,
      "loss": 5.403,
      "mean_token_accuracy": 0.22446643561124802,
      "num_tokens": 88375811.0,
      "step": 34865
    },
    {
      "entropy": 6.059533023834229,
      "epoch": 4.024235429890363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003475905061441648,
      "loss": 5.5481,
      "mean_token_accuracy": 0.2168492391705513,
      "num_tokens": 88388797.0,
      "step": 34870
    },
    {
      "entropy": 6.045244407653809,
      "epoch": 4.024812463935373,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003475514470985257,
      "loss": 5.4841,
      "mean_token_accuracy": 0.21442991346120835,
      "num_tokens": 88401150.0,
      "step": 34875
    },
    {
      "entropy": 6.04743447303772,
      "epoch": 4.025389497980381,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034751238561265467,
      "loss": 5.4667,
      "mean_token_accuracy": 0.2109408125281334,
      "num_tokens": 88414359.0,
      "step": 34880
    },
    {
      "entropy": 6.023638677597046,
      "epoch": 4.0259665320253895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003474733216878656,
      "loss": 5.4559,
      "mean_token_accuracy": 0.21457384079694747,
      "num_tokens": 88427110.0,
      "step": 34885
    },
    {
      "entropy": 6.02948579788208,
      "epoch": 4.026543566070398,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003474342553254724,
      "loss": 5.4984,
      "mean_token_accuracy": 0.21211129575967788,
      "num_tokens": 88440221.0,
      "step": 34890
    },
    {
      "entropy": 6.075586748123169,
      "epoch": 4.027120600115407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000347395186526789,
      "loss": 5.5281,
      "mean_token_accuracy": 0.21080708503723145,
      "num_tokens": 88452177.0,
      "step": 34895
    },
    {
      "entropy": 6.135594749450684,
      "epoch": 4.027697634160416,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034735611529312937,
      "loss": 5.5853,
      "mean_token_accuracy": 0.20992309600114822,
      "num_tokens": 88464704.0,
      "step": 34900
    },
    {
      "entropy": 6.085616779327393,
      "epoch": 4.028274668205424,
      "grad_norm": 0.90625,
      "learning_rate": 0.00034731704162580787,
      "loss": 5.5849,
      "mean_token_accuracy": 0.2090852826833725,
      "num_tokens": 88478102.0,
      "step": 34905
    },
    {
      "entropy": 6.1202089309692385,
      "epoch": 4.028851702250432,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034727796552613854,
      "loss": 5.5246,
      "mean_token_accuracy": 0.20304690301418304,
      "num_tokens": 88490817.0,
      "step": 34910
    },
    {
      "entropy": 6.0862287998199465,
      "epoch": 4.029428736295442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034723888699543564,
      "loss": 5.5884,
      "mean_token_accuracy": 0.20916266441345216,
      "num_tokens": 88504118.0,
      "step": 34915
    },
    {
      "entropy": 6.033890008926392,
      "epoch": 4.03000577034045,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003471998060350138,
      "loss": 5.5026,
      "mean_token_accuracy": 0.20833633840084076,
      "num_tokens": 88517146.0,
      "step": 34920
    },
    {
      "entropy": 6.00624041557312,
      "epoch": 4.030582804385459,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003471607226461873,
      "loss": 5.4075,
      "mean_token_accuracy": 0.21588986814022065,
      "num_tokens": 88529338.0,
      "step": 34925
    },
    {
      "entropy": 6.022734308242798,
      "epoch": 4.031159838430467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003471216368302707,
      "loss": 5.4843,
      "mean_token_accuracy": 0.22091287821531297,
      "num_tokens": 88541795.0,
      "step": 34930
    },
    {
      "entropy": 6.011997127532959,
      "epoch": 4.031736872475476,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003470825485885787,
      "loss": 5.4811,
      "mean_token_accuracy": 0.22161715477705002,
      "num_tokens": 88554971.0,
      "step": 34935
    },
    {
      "entropy": 5.905759954452515,
      "epoch": 4.032313906520485,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034704345792242606,
      "loss": 5.4751,
      "mean_token_accuracy": 0.21288659572601318,
      "num_tokens": 88567037.0,
      "step": 34940
    },
    {
      "entropy": 6.1228782653808596,
      "epoch": 4.032890940565493,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003470043648331274,
      "loss": 5.5992,
      "mean_token_accuracy": 0.2074059873819351,
      "num_tokens": 88580773.0,
      "step": 34945
    },
    {
      "entropy": 6.1154704093933105,
      "epoch": 4.033467974610502,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003469652693219977,
      "loss": 5.6103,
      "mean_token_accuracy": 0.20879853367805482,
      "num_tokens": 88594829.0,
      "step": 34950
    },
    {
      "entropy": 6.057131290435791,
      "epoch": 4.034045008655511,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034692617139035195,
      "loss": 5.5198,
      "mean_token_accuracy": 0.2102598488330841,
      "num_tokens": 88607824.0,
      "step": 34955
    },
    {
      "entropy": 6.062717485427856,
      "epoch": 4.034622042700519,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034688707103950516,
      "loss": 5.5017,
      "mean_token_accuracy": 0.2109762579202652,
      "num_tokens": 88620743.0,
      "step": 34960
    },
    {
      "entropy": 6.085350418090821,
      "epoch": 4.035199076745528,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034684796827077243,
      "loss": 5.5337,
      "mean_token_accuracy": 0.21046335995197296,
      "num_tokens": 88632938.0,
      "step": 34965
    },
    {
      "entropy": 6.090941143035889,
      "epoch": 4.035776110790537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003468088630854689,
      "loss": 5.6041,
      "mean_token_accuracy": 0.2118572399020195,
      "num_tokens": 88644900.0,
      "step": 34970
    },
    {
      "entropy": 6.054655456542969,
      "epoch": 4.036353144835545,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034676975548491006,
      "loss": 5.4802,
      "mean_token_accuracy": 0.21850402504205704,
      "num_tokens": 88657828.0,
      "step": 34975
    },
    {
      "entropy": 6.093479061126709,
      "epoch": 4.036930178880554,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034673064547041107,
      "loss": 5.54,
      "mean_token_accuracy": 0.21199256032705308,
      "num_tokens": 88670346.0,
      "step": 34980
    },
    {
      "entropy": 6.032238292694092,
      "epoch": 4.037507212925562,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034669153304328745,
      "loss": 5.4768,
      "mean_token_accuracy": 0.21034955978393555,
      "num_tokens": 88682699.0,
      "step": 34985
    },
    {
      "entropy": 6.089202737808227,
      "epoch": 4.0380842469705716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003466524182048548,
      "loss": 5.5534,
      "mean_token_accuracy": 0.209945248067379,
      "num_tokens": 88694873.0,
      "step": 34990
    },
    {
      "entropy": 5.9667998313903805,
      "epoch": 4.03866128101558,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003466133009564286,
      "loss": 5.4528,
      "mean_token_accuracy": 0.22358045130968093,
      "num_tokens": 88707709.0,
      "step": 34995
    },
    {
      "entropy": 6.096726989746093,
      "epoch": 4.039238315060588,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034657418129932456,
      "loss": 5.6226,
      "mean_token_accuracy": 0.20594664067029952,
      "num_tokens": 88720263.0,
      "step": 35000
    },
    {
      "entropy": 6.084662342071534,
      "epoch": 4.039815349105597,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003465350592348586,
      "loss": 5.5159,
      "mean_token_accuracy": 0.20816581845283508,
      "num_tokens": 88733664.0,
      "step": 35005
    },
    {
      "entropy": 6.042846584320069,
      "epoch": 4.040392383150606,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034649593476434627,
      "loss": 5.4799,
      "mean_token_accuracy": 0.21679667979478837,
      "num_tokens": 88746441.0,
      "step": 35010
    },
    {
      "entropy": 6.101603841781616,
      "epoch": 4.0409694171956145,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034645680788910385,
      "loss": 5.5213,
      "mean_token_accuracy": 0.2122953027486801,
      "num_tokens": 88758194.0,
      "step": 35015
    },
    {
      "entropy": 6.043791246414185,
      "epoch": 4.041546451240623,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003464176786104471,
      "loss": 5.5445,
      "mean_token_accuracy": 0.21630258560180665,
      "num_tokens": 88771471.0,
      "step": 35020
    },
    {
      "entropy": 6.110648775100708,
      "epoch": 4.042123485285632,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034637854692969223,
      "loss": 5.5532,
      "mean_token_accuracy": 0.21201925724744797,
      "num_tokens": 88783592.0,
      "step": 35025
    },
    {
      "entropy": 6.021303176879883,
      "epoch": 4.042700519330641,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003463394128481554,
      "loss": 5.4586,
      "mean_token_accuracy": 0.21380839198827745,
      "num_tokens": 88796445.0,
      "step": 35030
    },
    {
      "entropy": 5.988835334777832,
      "epoch": 4.043277553375649,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003463002763671528,
      "loss": 5.4529,
      "mean_token_accuracy": 0.2160557672381401,
      "num_tokens": 88809118.0,
      "step": 35035
    },
    {
      "entropy": 6.0822868824005125,
      "epoch": 4.0438545874206575,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003462611374880009,
      "loss": 5.506,
      "mean_token_accuracy": 0.2135401651263237,
      "num_tokens": 88821163.0,
      "step": 35040
    },
    {
      "entropy": 6.133023691177368,
      "epoch": 4.044431621465667,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00034622199621201606,
      "loss": 5.6046,
      "mean_token_accuracy": 0.20615515410900115,
      "num_tokens": 88833323.0,
      "step": 35045
    },
    {
      "entropy": 6.1287635326385494,
      "epoch": 4.045008655510675,
      "grad_norm": 1.0,
      "learning_rate": 0.0003461828525405147,
      "loss": 5.5835,
      "mean_token_accuracy": 0.2061597466468811,
      "num_tokens": 88845378.0,
      "step": 35050
    },
    {
      "entropy": 6.080401229858398,
      "epoch": 4.045585689555684,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003461437064748135,
      "loss": 5.5322,
      "mean_token_accuracy": 0.20550114810466766,
      "num_tokens": 88858365.0,
      "step": 35055
    },
    {
      "entropy": 6.057364416122437,
      "epoch": 4.046162723600692,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000346104558016229,
      "loss": 5.5062,
      "mean_token_accuracy": 0.2132334053516388,
      "num_tokens": 88870879.0,
      "step": 35060
    },
    {
      "entropy": 6.100865411758423,
      "epoch": 4.046739757645701,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000346065407166078,
      "loss": 5.5484,
      "mean_token_accuracy": 0.2093985304236412,
      "num_tokens": 88883525.0,
      "step": 35065
    },
    {
      "entropy": 6.090194129943848,
      "epoch": 4.04731679169071,
      "grad_norm": 1.0,
      "learning_rate": 0.0003460262539256774,
      "loss": 5.5481,
      "mean_token_accuracy": 0.20774607062339784,
      "num_tokens": 88897293.0,
      "step": 35070
    },
    {
      "entropy": 6.069189834594726,
      "epoch": 4.047893825735718,
      "grad_norm": 0.921875,
      "learning_rate": 0.000345987098296344,
      "loss": 5.5602,
      "mean_token_accuracy": 0.20843006372451783,
      "num_tokens": 88910697.0,
      "step": 35075
    },
    {
      "entropy": 6.1488330364227295,
      "epoch": 4.048470859780727,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003459479402793948,
      "loss": 5.5731,
      "mean_token_accuracy": 0.20986740291118622,
      "num_tokens": 88924325.0,
      "step": 35080
    },
    {
      "entropy": 6.1072039127349855,
      "epoch": 4.049047893825736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034590877987614687,
      "loss": 5.5836,
      "mean_token_accuracy": 0.19938248693943023,
      "num_tokens": 88937218.0,
      "step": 35085
    },
    {
      "entropy": 5.942077255249023,
      "epoch": 4.049624927870744,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003458696170879174,
      "loss": 5.3773,
      "mean_token_accuracy": 0.23093982189893722,
      "num_tokens": 88949849.0,
      "step": 35090
    },
    {
      "entropy": 6.006182384490967,
      "epoch": 4.050201961915753,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003458304519160235,
      "loss": 5.4263,
      "mean_token_accuracy": 0.21751437485218048,
      "num_tokens": 88963466.0,
      "step": 35095
    },
    {
      "entropy": 6.096313238143921,
      "epoch": 4.050778995960762,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034579128436178257,
      "loss": 5.6251,
      "mean_token_accuracy": 0.2097773402929306,
      "num_tokens": 88977431.0,
      "step": 35100
    },
    {
      "entropy": 6.026976490020752,
      "epoch": 4.0513560300057705,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034575211442651195,
      "loss": 5.481,
      "mean_token_accuracy": 0.22147003710269927,
      "num_tokens": 88989822.0,
      "step": 35105
    },
    {
      "entropy": 6.1603217124938965,
      "epoch": 4.051933064050779,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003457129421115291,
      "loss": 5.591,
      "mean_token_accuracy": 0.20492201596498488,
      "num_tokens": 89002469.0,
      "step": 35110
    },
    {
      "entropy": 6.060435056686401,
      "epoch": 4.052510098095787,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003456737674181516,
      "loss": 5.5298,
      "mean_token_accuracy": 0.2116502732038498,
      "num_tokens": 89015658.0,
      "step": 35115
    },
    {
      "entropy": 6.064583253860474,
      "epoch": 4.053087132140797,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034563459034769693,
      "loss": 5.5803,
      "mean_token_accuracy": 0.20522839277982713,
      "num_tokens": 89028708.0,
      "step": 35120
    },
    {
      "entropy": 6.128100156784058,
      "epoch": 4.053664166185805,
      "grad_norm": 1.109375,
      "learning_rate": 0.000345595410901483,
      "loss": 5.5775,
      "mean_token_accuracy": 0.20922085344791413,
      "num_tokens": 89040935.0,
      "step": 35125
    },
    {
      "entropy": 6.068128156661987,
      "epoch": 4.0542412002308135,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034555622908082745,
      "loss": 5.5816,
      "mean_token_accuracy": 0.19945605099201202,
      "num_tokens": 89051848.0,
      "step": 35130
    },
    {
      "entropy": 6.065807962417603,
      "epoch": 4.054818234275822,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003455170448870482,
      "loss": 5.4788,
      "mean_token_accuracy": 0.2141215533018112,
      "num_tokens": 89064094.0,
      "step": 35135
    },
    {
      "entropy": 6.131045770645142,
      "epoch": 4.055395268320831,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003454778583214632,
      "loss": 5.6014,
      "mean_token_accuracy": 0.2104654550552368,
      "num_tokens": 89076460.0,
      "step": 35140
    },
    {
      "entropy": 6.164707040786743,
      "epoch": 4.05597230236584,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003454386693853904,
      "loss": 5.6821,
      "mean_token_accuracy": 0.20278911739587785,
      "num_tokens": 89089576.0,
      "step": 35145
    },
    {
      "entropy": 6.011860466003418,
      "epoch": 4.056549336410848,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034539947808014793,
      "loss": 5.4699,
      "mean_token_accuracy": 0.21467145383358002,
      "num_tokens": 89103289.0,
      "step": 35150
    },
    {
      "entropy": 6.121705341339111,
      "epoch": 4.057126370455857,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000345360284407054,
      "loss": 5.5781,
      "mean_token_accuracy": 0.2018578514456749,
      "num_tokens": 89115856.0,
      "step": 35155
    },
    {
      "entropy": 6.089470529556275,
      "epoch": 4.057703404500866,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034532108836742687,
      "loss": 5.5793,
      "mean_token_accuracy": 0.20609953850507737,
      "num_tokens": 89127623.0,
      "step": 35160
    },
    {
      "entropy": 6.123625326156616,
      "epoch": 4.058280438545874,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034528188996258487,
      "loss": 5.5591,
      "mean_token_accuracy": 0.2040885269641876,
      "num_tokens": 89142041.0,
      "step": 35165
    },
    {
      "entropy": 6.090643692016601,
      "epoch": 4.058857472590883,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034524268919384636,
      "loss": 5.5252,
      "mean_token_accuracy": 0.20539493411779403,
      "num_tokens": 89154025.0,
      "step": 35170
    },
    {
      "entropy": 6.016414642333984,
      "epoch": 4.059434506635892,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00034520348606252994,
      "loss": 5.4725,
      "mean_token_accuracy": 0.2165176197886467,
      "num_tokens": 89168168.0,
      "step": 35175
    },
    {
      "entropy": 6.029225063323975,
      "epoch": 4.0600115406809,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003451642805699541,
      "loss": 5.4753,
      "mean_token_accuracy": 0.21871232241392136,
      "num_tokens": 89181201.0,
      "step": 35180
    },
    {
      "entropy": 6.026586627960205,
      "epoch": 4.060588574725909,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034512507271743765,
      "loss": 5.4433,
      "mean_token_accuracy": 0.22248812019824982,
      "num_tokens": 89193691.0,
      "step": 35185
    },
    {
      "entropy": 6.082807731628418,
      "epoch": 4.061165608770917,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003450858625062991,
      "loss": 5.5998,
      "mean_token_accuracy": 0.19912096709012986,
      "num_tokens": 89205546.0,
      "step": 35190
    },
    {
      "entropy": 6.085789966583252,
      "epoch": 4.0617426428159265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003450466499378574,
      "loss": 5.5064,
      "mean_token_accuracy": 0.21895743608474733,
      "num_tokens": 89218102.0,
      "step": 35195
    },
    {
      "entropy": 6.104584360122681,
      "epoch": 4.062319676860935,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003450074350134314,
      "loss": 5.5486,
      "mean_token_accuracy": 0.2057262495160103,
      "num_tokens": 89229879.0,
      "step": 35200
    },
    {
      "entropy": 5.942384862899781,
      "epoch": 4.062896710905943,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034496821773434024,
      "loss": 5.4206,
      "mean_token_accuracy": 0.21874017566442489,
      "num_tokens": 89242644.0,
      "step": 35205
    },
    {
      "entropy": 6.103714323043823,
      "epoch": 4.063473744950952,
      "grad_norm": 1.0,
      "learning_rate": 0.0003449289981019026,
      "loss": 5.5917,
      "mean_token_accuracy": 0.20302655696868896,
      "num_tokens": 89254419.0,
      "step": 35210
    },
    {
      "entropy": 6.117969083786011,
      "epoch": 4.064050778995961,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000344889776117438,
      "loss": 5.5643,
      "mean_token_accuracy": 0.20394493341445924,
      "num_tokens": 89267274.0,
      "step": 35215
    },
    {
      "entropy": 6.110006809234619,
      "epoch": 4.064627813040969,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034485055178226545,
      "loss": 5.6056,
      "mean_token_accuracy": 0.20921236723661424,
      "num_tokens": 89279890.0,
      "step": 35220
    },
    {
      "entropy": 6.0223987102508545,
      "epoch": 4.065204847085978,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003448113250977043,
      "loss": 5.502,
      "mean_token_accuracy": 0.21321035027503968,
      "num_tokens": 89292583.0,
      "step": 35225
    },
    {
      "entropy": 6.046645641326904,
      "epoch": 4.065781881130987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034477209606507394,
      "loss": 5.5724,
      "mean_token_accuracy": 0.2063037022948265,
      "num_tokens": 89305218.0,
      "step": 35230
    },
    {
      "entropy": 6.119129419326782,
      "epoch": 4.066358915175996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034473286468569375,
      "loss": 5.5821,
      "mean_token_accuracy": 0.20413969606161117,
      "num_tokens": 89317900.0,
      "step": 35235
    },
    {
      "entropy": 6.119109964370727,
      "epoch": 4.066935949221004,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003446936309608833,
      "loss": 5.5804,
      "mean_token_accuracy": 0.20921165943145753,
      "num_tokens": 89332038.0,
      "step": 35240
    },
    {
      "entropy": 6.053035545349121,
      "epoch": 4.067512983266012,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034465439489196213,
      "loss": 5.4823,
      "mean_token_accuracy": 0.21430400013923645,
      "num_tokens": 89346701.0,
      "step": 35245
    },
    {
      "entropy": 5.9432168960571286,
      "epoch": 4.068090017311022,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034461515648025003,
      "loss": 5.3716,
      "mean_token_accuracy": 0.22382305413484574,
      "num_tokens": 89359766.0,
      "step": 35250
    },
    {
      "entropy": 6.0504148483276365,
      "epoch": 4.06866705135603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003445759157270668,
      "loss": 5.5363,
      "mean_token_accuracy": 0.21363147497177123,
      "num_tokens": 89372137.0,
      "step": 35255
    },
    {
      "entropy": 5.916083192825317,
      "epoch": 4.0692440854010385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034453667263373213,
      "loss": 5.3403,
      "mean_token_accuracy": 0.2247455358505249,
      "num_tokens": 89385949.0,
      "step": 35260
    },
    {
      "entropy": 6.037625646591186,
      "epoch": 4.069821119446047,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00034449742720156604,
      "loss": 5.5264,
      "mean_token_accuracy": 0.2109056830406189,
      "num_tokens": 89398464.0,
      "step": 35265
    },
    {
      "entropy": 6.076513957977295,
      "epoch": 4.070398153491056,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00034445817943188847,
      "loss": 5.5256,
      "mean_token_accuracy": 0.203271521627903,
      "num_tokens": 89410937.0,
      "step": 35270
    },
    {
      "entropy": 5.939506769180298,
      "epoch": 4.070975187536065,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003444189293260196,
      "loss": 5.4479,
      "mean_token_accuracy": 0.2124747022986412,
      "num_tokens": 89423887.0,
      "step": 35275
    },
    {
      "entropy": 6.090556478500366,
      "epoch": 4.071552221581073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034437967688527945,
      "loss": 5.5282,
      "mean_token_accuracy": 0.21709538400173187,
      "num_tokens": 89435297.0,
      "step": 35280
    },
    {
      "entropy": 5.99617109298706,
      "epoch": 4.0721292556260815,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034434042211098836,
      "loss": 5.4142,
      "mean_token_accuracy": 0.21990175247192384,
      "num_tokens": 89447514.0,
      "step": 35285
    },
    {
      "entropy": 6.121022415161133,
      "epoch": 4.072706289671091,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003443011650044666,
      "loss": 5.6505,
      "mean_token_accuracy": 0.198582661151886,
      "num_tokens": 89461078.0,
      "step": 35290
    },
    {
      "entropy": 6.137017631530762,
      "epoch": 4.073283323716099,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003442619055670346,
      "loss": 5.5499,
      "mean_token_accuracy": 0.2109602928161621,
      "num_tokens": 89473294.0,
      "step": 35295
    },
    {
      "entropy": 6.070271301269531,
      "epoch": 4.073860357761108,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00034422264380001285,
      "loss": 5.5216,
      "mean_token_accuracy": 0.21472931355237962,
      "num_tokens": 89485957.0,
      "step": 35300
    },
    {
      "entropy": 6.07611312866211,
      "epoch": 4.074437391806117,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003441833797047218,
      "loss": 5.5736,
      "mean_token_accuracy": 0.20550290048122405,
      "num_tokens": 89497702.0,
      "step": 35305
    },
    {
      "entropy": 6.057424068450928,
      "epoch": 4.075014425851125,
      "grad_norm": 0.875,
      "learning_rate": 0.00034414411328248214,
      "loss": 5.564,
      "mean_token_accuracy": 0.20878633111715317,
      "num_tokens": 89510520.0,
      "step": 35310
    },
    {
      "entropy": 6.131303405761718,
      "epoch": 4.075591459896134,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034410484453461463,
      "loss": 5.6081,
      "mean_token_accuracy": 0.20582737773656845,
      "num_tokens": 89523521.0,
      "step": 35315
    },
    {
      "entropy": 5.984634637832642,
      "epoch": 4.076168493941142,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003440655734624399,
      "loss": 5.4316,
      "mean_token_accuracy": 0.21949629634618759,
      "num_tokens": 89535543.0,
      "step": 35320
    },
    {
      "entropy": 5.942215347290039,
      "epoch": 4.0767455279861515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000344026300067279,
      "loss": 5.431,
      "mean_token_accuracy": 0.21890262514352798,
      "num_tokens": 89548521.0,
      "step": 35325
    },
    {
      "entropy": 6.14573450088501,
      "epoch": 4.07732256203116,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003439870243504528,
      "loss": 5.5861,
      "mean_token_accuracy": 0.2013203412294388,
      "num_tokens": 89561251.0,
      "step": 35330
    },
    {
      "entropy": 6.043168354034424,
      "epoch": 4.077899596076168,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00034394774631328227,
      "loss": 5.4607,
      "mean_token_accuracy": 0.2202374055981636,
      "num_tokens": 89573192.0,
      "step": 35335
    },
    {
      "entropy": 6.0670102596282955,
      "epoch": 4.078476630121177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003439084659570886,
      "loss": 5.5746,
      "mean_token_accuracy": 0.20714314728975297,
      "num_tokens": 89585960.0,
      "step": 35340
    },
    {
      "entropy": 6.055860805511474,
      "epoch": 4.079053664166186,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034386918328319283,
      "loss": 5.5136,
      "mean_token_accuracy": 0.21303393691778183,
      "num_tokens": 89599703.0,
      "step": 35345
    },
    {
      "entropy": 6.077237749099732,
      "epoch": 4.0796306982111945,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034382989829291627,
      "loss": 5.4843,
      "mean_token_accuracy": 0.21803004890680314,
      "num_tokens": 89611815.0,
      "step": 35350
    },
    {
      "entropy": 5.90229754447937,
      "epoch": 4.080207732256203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003437906109875803,
      "loss": 5.4214,
      "mean_token_accuracy": 0.214641210436821,
      "num_tokens": 89624556.0,
      "step": 35355
    },
    {
      "entropy": 6.0637694835662845,
      "epoch": 4.080784766301211,
      "grad_norm": 1.0,
      "learning_rate": 0.00034375132136850635,
      "loss": 5.5098,
      "mean_token_accuracy": 0.21228972673416138,
      "num_tokens": 89637685.0,
      "step": 35360
    },
    {
      "entropy": 6.109176588058472,
      "epoch": 4.081361800346221,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003437120294370158,
      "loss": 5.5671,
      "mean_token_accuracy": 0.21146745681762696,
      "num_tokens": 89650900.0,
      "step": 35365
    },
    {
      "entropy": 5.9502928256988525,
      "epoch": 4.081938834391229,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003436727351944303,
      "loss": 5.3942,
      "mean_token_accuracy": 0.2216118723154068,
      "num_tokens": 89662840.0,
      "step": 35370
    },
    {
      "entropy": 6.0189612865447994,
      "epoch": 4.0825158684362375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003436334386420714,
      "loss": 5.4818,
      "mean_token_accuracy": 0.21849386543035507,
      "num_tokens": 89675433.0,
      "step": 35375
    },
    {
      "entropy": 6.132223463058471,
      "epoch": 4.083092902481247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003435941397812609,
      "loss": 5.5892,
      "mean_token_accuracy": 0.20545320063829423,
      "num_tokens": 89687533.0,
      "step": 35380
    },
    {
      "entropy": 6.054631853103638,
      "epoch": 4.083669936526255,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034355483861332064,
      "loss": 5.4423,
      "mean_token_accuracy": 0.22038716673851014,
      "num_tokens": 89700757.0,
      "step": 35385
    },
    {
      "entropy": 6.007598733901977,
      "epoch": 4.084246970571264,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003435155351395724,
      "loss": 5.4891,
      "mean_token_accuracy": 0.2149067848920822,
      "num_tokens": 89713135.0,
      "step": 35390
    },
    {
      "entropy": 6.059642934799195,
      "epoch": 4.084824004616272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003434762293613382,
      "loss": 5.4926,
      "mean_token_accuracy": 0.2180089011788368,
      "num_tokens": 89723910.0,
      "step": 35395
    },
    {
      "entropy": 6.020151519775391,
      "epoch": 4.085401038661281,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00034343692127993995,
      "loss": 5.497,
      "mean_token_accuracy": 0.21171986311674118,
      "num_tokens": 89736603.0,
      "step": 35400
    },
    {
      "entropy": 6.010009622573852,
      "epoch": 4.08597807270629,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003433976108967,
      "loss": 5.4939,
      "mean_token_accuracy": 0.21410107016563415,
      "num_tokens": 89750203.0,
      "step": 35405
    },
    {
      "entropy": 6.155652093887329,
      "epoch": 4.086555106751298,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003433582982129403,
      "loss": 5.5418,
      "mean_token_accuracy": 0.20297714471817016,
      "num_tokens": 89762078.0,
      "step": 35410
    },
    {
      "entropy": 6.025065898895264,
      "epoch": 4.087132140796307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003433189832299831,
      "loss": 5.4692,
      "mean_token_accuracy": 0.21612975597381592,
      "num_tokens": 89775421.0,
      "step": 35415
    },
    {
      "entropy": 6.024143171310425,
      "epoch": 4.087709174841316,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034327966594915096,
      "loss": 5.566,
      "mean_token_accuracy": 0.20120743811130523,
      "num_tokens": 89788196.0,
      "step": 35420
    },
    {
      "entropy": 6.135048484802246,
      "epoch": 4.088286208886324,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034324034637176615,
      "loss": 5.5526,
      "mean_token_accuracy": 0.20458617210388183,
      "num_tokens": 89800167.0,
      "step": 35425
    },
    {
      "entropy": 6.111774444580078,
      "epoch": 4.088863242931333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003432010244991512,
      "loss": 5.5649,
      "mean_token_accuracy": 0.20744165629148484,
      "num_tokens": 89812102.0,
      "step": 35430
    },
    {
      "entropy": 6.113587760925293,
      "epoch": 4.089440276976342,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003431617003326286,
      "loss": 5.6284,
      "mean_token_accuracy": 0.20039559602737428,
      "num_tokens": 89824235.0,
      "step": 35435
    },
    {
      "entropy": 6.052518081665039,
      "epoch": 4.0900173110213505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003431223738735211,
      "loss": 5.5937,
      "mean_token_accuracy": 0.2117765724658966,
      "num_tokens": 89838535.0,
      "step": 35440
    },
    {
      "entropy": 6.075787591934204,
      "epoch": 4.090594345066359,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003430830451231514,
      "loss": 5.5976,
      "mean_token_accuracy": 0.20219520777463912,
      "num_tokens": 89851022.0,
      "step": 35445
    },
    {
      "entropy": 6.058403158187867,
      "epoch": 4.091171379111367,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034304371408284235,
      "loss": 5.4784,
      "mean_token_accuracy": 0.21215640902519226,
      "num_tokens": 89864466.0,
      "step": 35450
    },
    {
      "entropy": 6.028244543075561,
      "epoch": 4.091748413156377,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003430043807539167,
      "loss": 5.5544,
      "mean_token_accuracy": 0.21020198464393616,
      "num_tokens": 89878179.0,
      "step": 35455
    },
    {
      "entropy": 6.093299102783203,
      "epoch": 4.092325447201385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003429650451376975,
      "loss": 5.5574,
      "mean_token_accuracy": 0.21927787661552428,
      "num_tokens": 89890933.0,
      "step": 35460
    },
    {
      "entropy": 6.146709775924682,
      "epoch": 4.092902481246393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034292570723550773,
      "loss": 5.6325,
      "mean_token_accuracy": 0.20091886967420577,
      "num_tokens": 89903677.0,
      "step": 35465
    },
    {
      "entropy": 6.067857933044434,
      "epoch": 4.093479515291402,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00034288636704867063,
      "loss": 5.5238,
      "mean_token_accuracy": 0.21076389998197556,
      "num_tokens": 89917550.0,
      "step": 35470
    },
    {
      "entropy": 5.957700872421265,
      "epoch": 4.094056549336411,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034284702457850917,
      "loss": 5.3994,
      "mean_token_accuracy": 0.2170511394739151,
      "num_tokens": 89929730.0,
      "step": 35475
    },
    {
      "entropy": 6.062205076217651,
      "epoch": 4.09463358338142,
      "grad_norm": 1.0,
      "learning_rate": 0.0003428076798263469,
      "loss": 5.5141,
      "mean_token_accuracy": 0.21333808749914168,
      "num_tokens": 89941644.0,
      "step": 35480
    },
    {
      "entropy": 6.048794794082641,
      "epoch": 4.095210617426428,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003427683327935069,
      "loss": 5.4976,
      "mean_token_accuracy": 0.2092871129512787,
      "num_tokens": 89953966.0,
      "step": 35485
    },
    {
      "entropy": 6.052239847183228,
      "epoch": 4.095787651471436,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034272898348131276,
      "loss": 5.4875,
      "mean_token_accuracy": 0.20827424824237822,
      "num_tokens": 89967916.0,
      "step": 35490
    },
    {
      "entropy": 5.992624950408936,
      "epoch": 4.096364685516446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034268963189108777,
      "loss": 5.4722,
      "mean_token_accuracy": 0.21729318797588348,
      "num_tokens": 89981901.0,
      "step": 35495
    },
    {
      "entropy": 6.115233325958252,
      "epoch": 4.096941719561454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003426502780241558,
      "loss": 5.5972,
      "mean_token_accuracy": 0.20486102253198624,
      "num_tokens": 89994139.0,
      "step": 35500
    },
    {
      "entropy": 6.049506759643554,
      "epoch": 4.097518753606463,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034261092188184015,
      "loss": 5.5207,
      "mean_token_accuracy": 0.21402349025011064,
      "num_tokens": 90007259.0,
      "step": 35505
    },
    {
      "entropy": 6.167341518402099,
      "epoch": 4.098095787651472,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003425715634654649,
      "loss": 5.628,
      "mean_token_accuracy": 0.19943179786205292,
      "num_tokens": 90019771.0,
      "step": 35510
    },
    {
      "entropy": 6.110190343856812,
      "epoch": 4.09867282169648,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003425322027763536,
      "loss": 5.5744,
      "mean_token_accuracy": 0.21097375005483626,
      "num_tokens": 90032392.0,
      "step": 35515
    },
    {
      "entropy": 6.134899473190307,
      "epoch": 4.099249855741489,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003424928398158301,
      "loss": 5.5992,
      "mean_token_accuracy": 0.20266730338335037,
      "num_tokens": 90044467.0,
      "step": 35520
    },
    {
      "entropy": 6.073607349395752,
      "epoch": 4.099826889786497,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003424534745852186,
      "loss": 5.45,
      "mean_token_accuracy": 0.21179691851139068,
      "num_tokens": 90057554.0,
      "step": 35525
    },
    {
      "entropy": 5.977625894546509,
      "epoch": 4.100403923831506,
      "grad_norm": 1.171875,
      "learning_rate": 0.000342414107085843,
      "loss": 5.4596,
      "mean_token_accuracy": 0.2172801285982132,
      "num_tokens": 90070708.0,
      "step": 35530
    },
    {
      "entropy": 6.113221502304077,
      "epoch": 4.100980957876515,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034237473731902734,
      "loss": 5.612,
      "mean_token_accuracy": 0.20872834026813508,
      "num_tokens": 90083547.0,
      "step": 35535
    },
    {
      "entropy": 6.1904645442962645,
      "epoch": 4.101557991921523,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003423353652860959,
      "loss": 5.5555,
      "mean_token_accuracy": 0.20782464146614074,
      "num_tokens": 90095410.0,
      "step": 35540
    },
    {
      "entropy": 6.103205442428589,
      "epoch": 4.102135025966532,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003422959909883728,
      "loss": 5.5731,
      "mean_token_accuracy": 0.2067727565765381,
      "num_tokens": 90107704.0,
      "step": 35545
    },
    {
      "entropy": 5.874544477462768,
      "epoch": 4.102712060011541,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034225661442718246,
      "loss": 5.3214,
      "mean_token_accuracy": 0.22959463596343993,
      "num_tokens": 90120941.0,
      "step": 35550
    },
    {
      "entropy": 6.113426160812378,
      "epoch": 4.103289094056549,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034221723560384944,
      "loss": 5.5797,
      "mean_token_accuracy": 0.21372756063938142,
      "num_tokens": 90133432.0,
      "step": 35555
    },
    {
      "entropy": 6.076683235168457,
      "epoch": 4.103866128101558,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003421778545196981,
      "loss": 5.5013,
      "mean_token_accuracy": 0.21161381751298905,
      "num_tokens": 90146680.0,
      "step": 35560
    },
    {
      "entropy": 6.081414842605591,
      "epoch": 4.104443162146566,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003421384711760528,
      "loss": 5.5812,
      "mean_token_accuracy": 0.2071981281042099,
      "num_tokens": 90159257.0,
      "step": 35565
    },
    {
      "entropy": 6.047891426086426,
      "epoch": 4.1050201961915755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00034209908557423846,
      "loss": 5.5274,
      "mean_token_accuracy": 0.2077357366681099,
      "num_tokens": 90172369.0,
      "step": 35570
    },
    {
      "entropy": 6.129982948303223,
      "epoch": 4.105597230236584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003420596977155798,
      "loss": 5.5582,
      "mean_token_accuracy": 0.20510038137435913,
      "num_tokens": 90184840.0,
      "step": 35575
    },
    {
      "entropy": 6.033859348297119,
      "epoch": 4.106174264281592,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034202030760140144,
      "loss": 5.4897,
      "mean_token_accuracy": 0.22107934951782227,
      "num_tokens": 90197752.0,
      "step": 35580
    },
    {
      "entropy": 6.089110422134399,
      "epoch": 4.106751298326602,
      "grad_norm": 0.921875,
      "learning_rate": 0.00034198091523302834,
      "loss": 5.5947,
      "mean_token_accuracy": 0.20557233393192292,
      "num_tokens": 90211092.0,
      "step": 35585
    },
    {
      "entropy": 6.156796455383301,
      "epoch": 4.10732833237161,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003419415206117854,
      "loss": 5.6474,
      "mean_token_accuracy": 0.200355464220047,
      "num_tokens": 90225239.0,
      "step": 35590
    },
    {
      "entropy": 6.14080843925476,
      "epoch": 4.1079053664166185,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003419021237389977,
      "loss": 5.623,
      "mean_token_accuracy": 0.19885305762290956,
      "num_tokens": 90238484.0,
      "step": 35595
    },
    {
      "entropy": 6.035296249389648,
      "epoch": 4.108482400461627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034186272461599027,
      "loss": 5.5225,
      "mean_token_accuracy": 0.20796301513910292,
      "num_tokens": 90249975.0,
      "step": 35600
    },
    {
      "entropy": 6.127810716629028,
      "epoch": 4.109059434506636,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034182332324408834,
      "loss": 5.5727,
      "mean_token_accuracy": 0.21209159046411513,
      "num_tokens": 90262647.0,
      "step": 35605
    },
    {
      "entropy": 6.147694396972656,
      "epoch": 4.109636468551645,
      "grad_norm": 1.375,
      "learning_rate": 0.00034178391962461714,
      "loss": 5.6069,
      "mean_token_accuracy": 0.20603940039873123,
      "num_tokens": 90276088.0,
      "step": 35610
    },
    {
      "entropy": 5.99590573310852,
      "epoch": 4.110213502596653,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003417445137589019,
      "loss": 5.4803,
      "mean_token_accuracy": 0.2151528298854828,
      "num_tokens": 90288740.0,
      "step": 35615
    },
    {
      "entropy": 6.051789999008179,
      "epoch": 4.1107905366416615,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034170510564826816,
      "loss": 5.4712,
      "mean_token_accuracy": 0.2125127136707306,
      "num_tokens": 90300422.0,
      "step": 35620
    },
    {
      "entropy": 6.103476858139038,
      "epoch": 4.111367570686671,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003416656952940413,
      "loss": 5.5429,
      "mean_token_accuracy": 0.21694846749305724,
      "num_tokens": 90312595.0,
      "step": 35625
    },
    {
      "entropy": 6.040262365341187,
      "epoch": 4.111944604731679,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034162628269754694,
      "loss": 5.5166,
      "mean_token_accuracy": 0.2167874366044998,
      "num_tokens": 90325404.0,
      "step": 35630
    },
    {
      "entropy": 6.06687331199646,
      "epoch": 4.112521638776688,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003415868678601106,
      "loss": 5.517,
      "mean_token_accuracy": 0.21294075548648833,
      "num_tokens": 90337579.0,
      "step": 35635
    },
    {
      "entropy": 6.069899225234986,
      "epoch": 4.113098672821696,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034154745078305806,
      "loss": 5.5505,
      "mean_token_accuracy": 0.2093985930085182,
      "num_tokens": 90350535.0,
      "step": 35640
    },
    {
      "entropy": 6.129082679748535,
      "epoch": 4.113675706866705,
      "grad_norm": 1.0,
      "learning_rate": 0.000341508031467715,
      "loss": 5.5761,
      "mean_token_accuracy": 0.20228931903839112,
      "num_tokens": 90363635.0,
      "step": 35645
    },
    {
      "entropy": 6.052897357940674,
      "epoch": 4.114252740911714,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003414686099154074,
      "loss": 5.5255,
      "mean_token_accuracy": 0.2128541052341461,
      "num_tokens": 90376119.0,
      "step": 35650
    },
    {
      "entropy": 5.958323431015015,
      "epoch": 4.114829774956722,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00034142918612746124,
      "loss": 5.3818,
      "mean_token_accuracy": 0.21763400137424468,
      "num_tokens": 90388784.0,
      "step": 35655
    },
    {
      "entropy": 6.111651849746704,
      "epoch": 4.1154068090017315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003413897601052023,
      "loss": 5.6284,
      "mean_token_accuracy": 0.20257425010204316,
      "num_tokens": 90401273.0,
      "step": 35660
    },
    {
      "entropy": 6.10703125,
      "epoch": 4.11598384304674,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003413503318499566,
      "loss": 5.5258,
      "mean_token_accuracy": 0.20975951552391053,
      "num_tokens": 90414421.0,
      "step": 35665
    },
    {
      "entropy": 5.985723686218262,
      "epoch": 4.116560877091748,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034131090136305066,
      "loss": 5.3852,
      "mean_token_accuracy": 0.22481465935707093,
      "num_tokens": 90427391.0,
      "step": 35670
    },
    {
      "entropy": 6.100257396697998,
      "epoch": 4.117137911136757,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034127146864581045,
      "loss": 5.5691,
      "mean_token_accuracy": 0.21422185450792314,
      "num_tokens": 90439712.0,
      "step": 35675
    },
    {
      "entropy": 5.993046045303345,
      "epoch": 4.117714945181766,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034123203369956225,
      "loss": 5.5111,
      "mean_token_accuracy": 0.21299288868904115,
      "num_tokens": 90452514.0,
      "step": 35680
    },
    {
      "entropy": 6.132471942901612,
      "epoch": 4.1182919792267745,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00034119259652563253,
      "loss": 5.6189,
      "mean_token_accuracy": 0.2062388017773628,
      "num_tokens": 90465245.0,
      "step": 35685
    },
    {
      "entropy": 6.094356393814087,
      "epoch": 4.118869013271783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003411531571253477,
      "loss": 5.4881,
      "mean_token_accuracy": 0.22057535499334335,
      "num_tokens": 90477747.0,
      "step": 35690
    },
    {
      "entropy": 6.096591854095459,
      "epoch": 4.119446047316791,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034111371550003426,
      "loss": 5.525,
      "mean_token_accuracy": 0.20705520659685134,
      "num_tokens": 90489982.0,
      "step": 35695
    },
    {
      "entropy": 5.992465591430664,
      "epoch": 4.120023081361801,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003410742716510189,
      "loss": 5.4238,
      "mean_token_accuracy": 0.22328459620475768,
      "num_tokens": 90502336.0,
      "step": 35700
    },
    {
      "entropy": 6.086414861679077,
      "epoch": 4.120600115406809,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003410348255796282,
      "loss": 5.5187,
      "mean_token_accuracy": 0.21496592313051224,
      "num_tokens": 90514566.0,
      "step": 35705
    },
    {
      "entropy": 6.039079332351685,
      "epoch": 4.1211771494518175,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034099537728718893,
      "loss": 5.5345,
      "mean_token_accuracy": 0.20865318328142166,
      "num_tokens": 90528222.0,
      "step": 35710
    },
    {
      "entropy": 6.160186338424682,
      "epoch": 4.121754183496826,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034095592677502797,
      "loss": 5.5779,
      "mean_token_accuracy": 0.21000425815582274,
      "num_tokens": 90541189.0,
      "step": 35715
    },
    {
      "entropy": 6.082367181777954,
      "epoch": 4.122331217541835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034091647404447216,
      "loss": 5.5384,
      "mean_token_accuracy": 0.21051134467124938,
      "num_tokens": 90555211.0,
      "step": 35720
    },
    {
      "entropy": 6.073507165908813,
      "epoch": 4.122908251586844,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034087701909684846,
      "loss": 5.5497,
      "mean_token_accuracy": 0.2095015153288841,
      "num_tokens": 90568117.0,
      "step": 35725
    },
    {
      "entropy": 5.99535722732544,
      "epoch": 4.123485285631852,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000340837561933484,
      "loss": 5.449,
      "mean_token_accuracy": 0.21343269497156142,
      "num_tokens": 90582397.0,
      "step": 35730
    },
    {
      "entropy": 6.11025218963623,
      "epoch": 4.124062319676861,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034079810255570575,
      "loss": 5.5759,
      "mean_token_accuracy": 0.20200441926717758,
      "num_tokens": 90594878.0,
      "step": 35735
    },
    {
      "entropy": 6.010230922698975,
      "epoch": 4.12463935372187,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003407586409648411,
      "loss": 5.4875,
      "mean_token_accuracy": 0.20678807348012923,
      "num_tokens": 90607418.0,
      "step": 35740
    },
    {
      "entropy": 6.069612884521485,
      "epoch": 4.125216387766878,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00034071917716221715,
      "loss": 5.5257,
      "mean_token_accuracy": 0.21699086278676988,
      "num_tokens": 90619945.0,
      "step": 35745
    },
    {
      "entropy": 6.015505981445313,
      "epoch": 4.125793421811887,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003406797111491614,
      "loss": 5.4706,
      "mean_token_accuracy": 0.22581159323453903,
      "num_tokens": 90633983.0,
      "step": 35750
    },
    {
      "entropy": 6.134730672836303,
      "epoch": 4.126370455856896,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003406402429270011,
      "loss": 5.5355,
      "mean_token_accuracy": 0.20724848359823228,
      "num_tokens": 90646773.0,
      "step": 35755
    },
    {
      "entropy": 6.00944676399231,
      "epoch": 4.126947489901904,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034060077249706397,
      "loss": 5.5074,
      "mean_token_accuracy": 0.21261016577482222,
      "num_tokens": 90659771.0,
      "step": 35760
    },
    {
      "entropy": 5.9040679931640625,
      "epoch": 4.127524523946913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003405612998606774,
      "loss": 5.3254,
      "mean_token_accuracy": 0.2295476481318474,
      "num_tokens": 90673448.0,
      "step": 35765
    },
    {
      "entropy": 6.050318193435669,
      "epoch": 4.128101557991921,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003405218250191689,
      "loss": 5.53,
      "mean_token_accuracy": 0.20032575577497483,
      "num_tokens": 90686277.0,
      "step": 35770
    },
    {
      "entropy": 6.060693216323853,
      "epoch": 4.12867859203693,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00034048234797386654,
      "loss": 5.5522,
      "mean_token_accuracy": 0.21081956923007966,
      "num_tokens": 90698793.0,
      "step": 35775
    },
    {
      "entropy": 6.151148748397827,
      "epoch": 4.129255626081939,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003404428687260979,
      "loss": 5.6085,
      "mean_token_accuracy": 0.20680249333381653,
      "num_tokens": 90712325.0,
      "step": 35780
    },
    {
      "entropy": 6.111231803894043,
      "epoch": 4.129832660126947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034040338727719087,
      "loss": 5.5445,
      "mean_token_accuracy": 0.211186720430851,
      "num_tokens": 90724845.0,
      "step": 35785
    },
    {
      "entropy": 6.061575317382813,
      "epoch": 4.130409694171956,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034036390362847337,
      "loss": 5.5527,
      "mean_token_accuracy": 0.20894260406494142,
      "num_tokens": 90737136.0,
      "step": 35790
    },
    {
      "entropy": 6.093926000595093,
      "epoch": 4.130986728216965,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003403244177812735,
      "loss": 5.531,
      "mean_token_accuracy": 0.21399353742599486,
      "num_tokens": 90749703.0,
      "step": 35795
    },
    {
      "entropy": 6.020750999450684,
      "epoch": 4.131563762261973,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034028492973691923,
      "loss": 5.4501,
      "mean_token_accuracy": 0.21650319844484328,
      "num_tokens": 90763264.0,
      "step": 35800
    },
    {
      "entropy": 6.102464437484741,
      "epoch": 4.132140796306982,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003402454394967388,
      "loss": 5.629,
      "mean_token_accuracy": 0.20359870940446853,
      "num_tokens": 90775290.0,
      "step": 35805
    },
    {
      "entropy": 6.04753942489624,
      "epoch": 4.132717830351991,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003402059470620604,
      "loss": 5.4746,
      "mean_token_accuracy": 0.21019257009029388,
      "num_tokens": 90788765.0,
      "step": 35810
    },
    {
      "entropy": 6.030824041366577,
      "epoch": 4.133294864397,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034016645243421236,
      "loss": 5.4678,
      "mean_token_accuracy": 0.2205435112118721,
      "num_tokens": 90801141.0,
      "step": 35815
    },
    {
      "entropy": 6.14891881942749,
      "epoch": 4.133871898442008,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003401269556145231,
      "loss": 5.5944,
      "mean_token_accuracy": 0.20381052941083908,
      "num_tokens": 90813451.0,
      "step": 35820
    },
    {
      "entropy": 6.079682397842407,
      "epoch": 4.134448932487016,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034008745660432107,
      "loss": 5.5841,
      "mean_token_accuracy": 0.2068340927362442,
      "num_tokens": 90826283.0,
      "step": 35825
    },
    {
      "entropy": 6.017327165603637,
      "epoch": 4.135025966532026,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034004795540493464,
      "loss": 5.3992,
      "mean_token_accuracy": 0.2253447636961937,
      "num_tokens": 90838391.0,
      "step": 35830
    },
    {
      "entropy": 6.088587999343872,
      "epoch": 4.135603000577034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034000845201769263,
      "loss": 5.5711,
      "mean_token_accuracy": 0.21017614006996155,
      "num_tokens": 90850234.0,
      "step": 35835
    },
    {
      "entropy": 6.075516986846924,
      "epoch": 4.1361800346220425,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003399689464439236,
      "loss": 5.4644,
      "mean_token_accuracy": 0.21411296278238295,
      "num_tokens": 90862771.0,
      "step": 35840
    },
    {
      "entropy": 6.001401138305664,
      "epoch": 4.136757068667051,
      "grad_norm": 0.859375,
      "learning_rate": 0.00033992943868495635,
      "loss": 5.4976,
      "mean_token_accuracy": 0.21470560431480407,
      "num_tokens": 90877217.0,
      "step": 35845
    },
    {
      "entropy": 6.053483343124389,
      "epoch": 4.13733410271206,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003398899287421197,
      "loss": 5.5035,
      "mean_token_accuracy": 0.22157915383577348,
      "num_tokens": 90891283.0,
      "step": 35850
    },
    {
      "entropy": 6.114140796661377,
      "epoch": 4.137911136757069,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003398504166167425,
      "loss": 5.5392,
      "mean_token_accuracy": 0.2111232742667198,
      "num_tokens": 90902618.0,
      "step": 35855
    },
    {
      "entropy": 6.061532735824585,
      "epoch": 4.138488170802077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033981090231015375,
      "loss": 5.5588,
      "mean_token_accuracy": 0.20485687404870986,
      "num_tokens": 90914711.0,
      "step": 35860
    },
    {
      "entropy": 5.985019636154175,
      "epoch": 4.139065204847086,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003397713858236825,
      "loss": 5.5147,
      "mean_token_accuracy": 0.2078494518995285,
      "num_tokens": 90928579.0,
      "step": 35865
    },
    {
      "entropy": 6.08093671798706,
      "epoch": 4.139642238892095,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033973186715865783,
      "loss": 5.546,
      "mean_token_accuracy": 0.21357639580965043,
      "num_tokens": 90942196.0,
      "step": 35870
    },
    {
      "entropy": 6.034729242324829,
      "epoch": 4.140219272937103,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000339692346316409,
      "loss": 5.4648,
      "mean_token_accuracy": 0.2130606696009636,
      "num_tokens": 90955092.0,
      "step": 35875
    },
    {
      "entropy": 6.035179519653321,
      "epoch": 4.140796306982112,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033965282329826525,
      "loss": 5.4707,
      "mean_token_accuracy": 0.2174498438835144,
      "num_tokens": 90967887.0,
      "step": 35880
    },
    {
      "entropy": 6.006946420669555,
      "epoch": 4.141373341027121,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033961329810555587,
      "loss": 5.484,
      "mean_token_accuracy": 0.21400219202041626,
      "num_tokens": 90980077.0,
      "step": 35885
    },
    {
      "entropy": 6.172068929672241,
      "epoch": 4.141950375072129,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033957377073961034,
      "loss": 5.6863,
      "mean_token_accuracy": 0.19841102212667466,
      "num_tokens": 90991544.0,
      "step": 35890
    },
    {
      "entropy": 6.0838698387146,
      "epoch": 4.142527409117138,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033953424120175806,
      "loss": 5.5413,
      "mean_token_accuracy": 0.21244185715913771,
      "num_tokens": 91004739.0,
      "step": 35895
    },
    {
      "entropy": 6.018929576873779,
      "epoch": 4.143104443162146,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033949470949332864,
      "loss": 5.454,
      "mean_token_accuracy": 0.21465598344802855,
      "num_tokens": 91018090.0,
      "step": 35900
    },
    {
      "entropy": 5.949872636795044,
      "epoch": 4.1436814772071555,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033945517561565175,
      "loss": 5.4436,
      "mean_token_accuracy": 0.21365102678537368,
      "num_tokens": 91032294.0,
      "step": 35905
    },
    {
      "entropy": 6.015384149551392,
      "epoch": 4.144258511252164,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00033941563957005705,
      "loss": 5.465,
      "mean_token_accuracy": 0.2187521368265152,
      "num_tokens": 91045593.0,
      "step": 35910
    },
    {
      "entropy": 6.03531060218811,
      "epoch": 4.144835545297172,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003393761013578742,
      "loss": 5.4888,
      "mean_token_accuracy": 0.20976669043302537,
      "num_tokens": 91058163.0,
      "step": 35915
    },
    {
      "entropy": 6.124438714981079,
      "epoch": 4.145412579342181,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00033933656098043325,
      "loss": 5.5912,
      "mean_token_accuracy": 0.20663621723651887,
      "num_tokens": 91072155.0,
      "step": 35920
    },
    {
      "entropy": 6.081832361221314,
      "epoch": 4.14598961338719,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033929701843906404,
      "loss": 5.5765,
      "mean_token_accuracy": 0.21051104217767716,
      "num_tokens": 91085560.0,
      "step": 35925
    },
    {
      "entropy": 5.946006536483765,
      "epoch": 4.1465666474321985,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003392574737350966,
      "loss": 5.4435,
      "mean_token_accuracy": 0.21187988370656968,
      "num_tokens": 91097517.0,
      "step": 35930
    },
    {
      "entropy": 6.138819980621338,
      "epoch": 4.147143681477207,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00033921792686986076,
      "loss": 5.5822,
      "mean_token_accuracy": 0.20238389372825621,
      "num_tokens": 91109785.0,
      "step": 35935
    },
    {
      "entropy": 5.956558799743652,
      "epoch": 4.147720715522216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033917837784468695,
      "loss": 5.3888,
      "mean_token_accuracy": 0.2254834845662117,
      "num_tokens": 91122610.0,
      "step": 35940
    },
    {
      "entropy": 6.120345258712769,
      "epoch": 4.148297749567225,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003391388266609052,
      "loss": 5.5987,
      "mean_token_accuracy": 0.20726500153541566,
      "num_tokens": 91137596.0,
      "step": 35945
    },
    {
      "entropy": 6.048577070236206,
      "epoch": 4.148874783612233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000339099273319846,
      "loss": 5.517,
      "mean_token_accuracy": 0.21385516077280045,
      "num_tokens": 91149240.0,
      "step": 35950
    },
    {
      "entropy": 6.029694271087647,
      "epoch": 4.1494518176572415,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033905971782283964,
      "loss": 5.4923,
      "mean_token_accuracy": 0.21502895951271056,
      "num_tokens": 91161983.0,
      "step": 35955
    },
    {
      "entropy": 6.055621194839477,
      "epoch": 4.150028851702251,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033902016017121634,
      "loss": 5.5372,
      "mean_token_accuracy": 0.2109047681093216,
      "num_tokens": 91175496.0,
      "step": 35960
    },
    {
      "entropy": 6.112962913513184,
      "epoch": 4.150605885747259,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003389806003663068,
      "loss": 5.6688,
      "mean_token_accuracy": 0.20245399624109267,
      "num_tokens": 91189208.0,
      "step": 35965
    },
    {
      "entropy": 6.118751001358032,
      "epoch": 4.151182919792268,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003389410384094416,
      "loss": 5.5891,
      "mean_token_accuracy": 0.20861436426639557,
      "num_tokens": 91201027.0,
      "step": 35970
    },
    {
      "entropy": 6.089512348175049,
      "epoch": 4.151759953837276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003389014743019513,
      "loss": 5.585,
      "mean_token_accuracy": 0.20493076294660567,
      "num_tokens": 91213380.0,
      "step": 35975
    },
    {
      "entropy": 6.073336219787597,
      "epoch": 4.152336987882285,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003388619080451666,
      "loss": 5.5068,
      "mean_token_accuracy": 0.21620817482471466,
      "num_tokens": 91224860.0,
      "step": 35980
    },
    {
      "entropy": 6.023465347290039,
      "epoch": 4.152914021927294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033882233964041835,
      "loss": 5.4973,
      "mean_token_accuracy": 0.21763940453529357,
      "num_tokens": 91236844.0,
      "step": 35985
    },
    {
      "entropy": 6.075516176223755,
      "epoch": 4.153491055972302,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033878276908903756,
      "loss": 5.5435,
      "mean_token_accuracy": 0.20915375053882598,
      "num_tokens": 91249059.0,
      "step": 35990
    },
    {
      "entropy": 6.195187711715699,
      "epoch": 4.1540680900173115,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033874319639235487,
      "loss": 5.6374,
      "mean_token_accuracy": 0.20203358978033065,
      "num_tokens": 91260513.0,
      "step": 35995
    },
    {
      "entropy": 6.037358570098877,
      "epoch": 4.15464512406232,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003387036215517014,
      "loss": 5.4943,
      "mean_token_accuracy": 0.2176566556096077,
      "num_tokens": 91272488.0,
      "step": 36000
    },
    {
      "epoch": 4.15464512406232,
      "eval_entropy": 5.858769298853914,
      "eval_loss": 5.717448711395264,
      "eval_mean_token_accuracy": 0.20955150646684792,
      "eval_num_tokens": 91272488.0,
      "eval_runtime": 23.476,
      "eval_samples_per_second": 1198.499,
      "eval_steps_per_second": 149.812,
      "step": 36000
    },
    {
      "entropy": 5.978639698028564,
      "epoch": 4.155222158107328,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003386640445684084,
      "loss": 5.4509,
      "mean_token_accuracy": 0.21751538515090943,
      "num_tokens": 91284814.0,
      "step": 36005
    },
    {
      "entropy": 6.080470085144043,
      "epoch": 4.155799192152337,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003386244654438068,
      "loss": 5.5315,
      "mean_token_accuracy": 0.20885832309722902,
      "num_tokens": 91296635.0,
      "step": 36010
    },
    {
      "entropy": 6.075433397293091,
      "epoch": 4.156376226197346,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033858488417922794,
      "loss": 5.5553,
      "mean_token_accuracy": 0.2061830624938011,
      "num_tokens": 91309803.0,
      "step": 36015
    },
    {
      "entropy": 6.107651948928833,
      "epoch": 4.1569532602423545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033854530077600305,
      "loss": 5.6009,
      "mean_token_accuracy": 0.20393964797258377,
      "num_tokens": 91321796.0,
      "step": 36020
    },
    {
      "entropy": 6.129538440704346,
      "epoch": 4.157530294287363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033850571523546353,
      "loss": 5.5566,
      "mean_token_accuracy": 0.2077781468629837,
      "num_tokens": 91334290.0,
      "step": 36025
    },
    {
      "entropy": 6.017099761962891,
      "epoch": 4.158107328332371,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033846612755894084,
      "loss": 5.4585,
      "mean_token_accuracy": 0.2182576909661293,
      "num_tokens": 91348226.0,
      "step": 36030
    },
    {
      "entropy": 6.151995658874512,
      "epoch": 4.158684362377381,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033842653774776647,
      "loss": 5.6414,
      "mean_token_accuracy": 0.20287489742040635,
      "num_tokens": 91360195.0,
      "step": 36035
    },
    {
      "entropy": 6.101355075836182,
      "epoch": 4.159261396422389,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033838694580327205,
      "loss": 5.5066,
      "mean_token_accuracy": 0.21247724294662476,
      "num_tokens": 91373048.0,
      "step": 36040
    },
    {
      "entropy": 6.059773969650268,
      "epoch": 4.159838430467397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033834735172678915,
      "loss": 5.5594,
      "mean_token_accuracy": 0.2153088182210922,
      "num_tokens": 91385801.0,
      "step": 36045
    },
    {
      "entropy": 6.0652406215667725,
      "epoch": 4.160415464512406,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00033830775551964954,
      "loss": 5.5492,
      "mean_token_accuracy": 0.2091422498226166,
      "num_tokens": 91398803.0,
      "step": 36050
    },
    {
      "entropy": 6.088028192520142,
      "epoch": 4.160992498557415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003382681571831851,
      "loss": 5.5547,
      "mean_token_accuracy": 0.20253238677978516,
      "num_tokens": 91410083.0,
      "step": 36055
    },
    {
      "entropy": 6.1005878925323485,
      "epoch": 4.161569532602424,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003382285567187275,
      "loss": 5.5517,
      "mean_token_accuracy": 0.2081000491976738,
      "num_tokens": 91422306.0,
      "step": 36060
    },
    {
      "entropy": 5.976929378509522,
      "epoch": 4.162146566647432,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033818895412760885,
      "loss": 5.4847,
      "mean_token_accuracy": 0.22056014984846115,
      "num_tokens": 91434108.0,
      "step": 36065
    },
    {
      "entropy": 6.069995403289795,
      "epoch": 4.162723600692441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003381493494111612,
      "loss": 5.5096,
      "mean_token_accuracy": 0.20841262936592103,
      "num_tokens": 91446547.0,
      "step": 36070
    },
    {
      "entropy": 5.942376279830933,
      "epoch": 4.16330063473745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003381097425707164,
      "loss": 5.3638,
      "mean_token_accuracy": 0.22560459226369858,
      "num_tokens": 91458394.0,
      "step": 36075
    },
    {
      "entropy": 6.0690834522247314,
      "epoch": 4.163877668782458,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003380701336076068,
      "loss": 5.5104,
      "mean_token_accuracy": 0.21953116357326508,
      "num_tokens": 91471526.0,
      "step": 36080
    },
    {
      "entropy": 6.041635274887085,
      "epoch": 4.1644547028274665,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033803052252316456,
      "loss": 5.5275,
      "mean_token_accuracy": 0.21013310849666594,
      "num_tokens": 91484709.0,
      "step": 36085
    },
    {
      "entropy": 5.935430335998535,
      "epoch": 4.165031736872476,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033799090931872213,
      "loss": 5.3996,
      "mean_token_accuracy": 0.2252883121371269,
      "num_tokens": 91498405.0,
      "step": 36090
    },
    {
      "entropy": 6.011941194534302,
      "epoch": 4.165608770917484,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033795129399561167,
      "loss": 5.4644,
      "mean_token_accuracy": 0.22101012617349625,
      "num_tokens": 91511811.0,
      "step": 36095
    },
    {
      "entropy": 6.003114318847656,
      "epoch": 4.166185804962493,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00033791167655516577,
      "loss": 5.5126,
      "mean_token_accuracy": 0.2159534603357315,
      "num_tokens": 91525350.0,
      "step": 36100
    },
    {
      "entropy": 6.008171653747558,
      "epoch": 4.166762839007501,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033787205699871675,
      "loss": 5.4591,
      "mean_token_accuracy": 0.21674319356679916,
      "num_tokens": 91539036.0,
      "step": 36105
    },
    {
      "entropy": 6.1081925392150875,
      "epoch": 4.16733987305251,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003378324353275974,
      "loss": 5.6162,
      "mean_token_accuracy": 0.2078762486577034,
      "num_tokens": 91551188.0,
      "step": 36110
    },
    {
      "entropy": 6.074532175064087,
      "epoch": 4.167916907097519,
      "grad_norm": 1.0,
      "learning_rate": 0.00033779281154314037,
      "loss": 5.5102,
      "mean_token_accuracy": 0.21162545531988144,
      "num_tokens": 91563586.0,
      "step": 36115
    },
    {
      "entropy": 6.0588624477386475,
      "epoch": 4.168493941142527,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033775318564667825,
      "loss": 5.4547,
      "mean_token_accuracy": 0.22079792767763137,
      "num_tokens": 91575124.0,
      "step": 36120
    },
    {
      "entropy": 6.084282159805298,
      "epoch": 4.169070975187536,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003377135576395438,
      "loss": 5.551,
      "mean_token_accuracy": 0.2125624969601631,
      "num_tokens": 91587086.0,
      "step": 36125
    },
    {
      "entropy": 6.042050313949585,
      "epoch": 4.169648009232545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033767392752307015,
      "loss": 5.4265,
      "mean_token_accuracy": 0.22312956303358078,
      "num_tokens": 91598357.0,
      "step": 36130
    },
    {
      "entropy": 6.016818523406982,
      "epoch": 4.170225043277553,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00033763429529858996,
      "loss": 5.5073,
      "mean_token_accuracy": 0.21356083303689957,
      "num_tokens": 91612601.0,
      "step": 36135
    },
    {
      "entropy": 6.0931182384490965,
      "epoch": 4.170802077322562,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033759466096743647,
      "loss": 5.5235,
      "mean_token_accuracy": 0.2094171777367592,
      "num_tokens": 91625313.0,
      "step": 36140
    },
    {
      "entropy": 6.0659037113189695,
      "epoch": 4.171379111367571,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033755502453094256,
      "loss": 5.4995,
      "mean_token_accuracy": 0.21348541975021362,
      "num_tokens": 91638709.0,
      "step": 36145
    },
    {
      "entropy": 6.077402496337891,
      "epoch": 4.1719561454125795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003375153859904415,
      "loss": 5.5708,
      "mean_token_accuracy": 0.2069834738969803,
      "num_tokens": 91650860.0,
      "step": 36150
    },
    {
      "entropy": 6.077566528320313,
      "epoch": 4.172533179457588,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00033747574534726655,
      "loss": 5.5815,
      "mean_token_accuracy": 0.20579287707805632,
      "num_tokens": 91663162.0,
      "step": 36155
    },
    {
      "entropy": 6.144608879089356,
      "epoch": 4.173110213502596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003374361026027509,
      "loss": 5.5292,
      "mean_token_accuracy": 0.21433990746736525,
      "num_tokens": 91674874.0,
      "step": 36160
    },
    {
      "entropy": 6.0306562900543215,
      "epoch": 4.173687247547606,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003373964577582278,
      "loss": 5.5631,
      "mean_token_accuracy": 0.2087247058749199,
      "num_tokens": 91687592.0,
      "step": 36165
    },
    {
      "entropy": 6.122583627700806,
      "epoch": 4.174264281592614,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003373568108150311,
      "loss": 5.6388,
      "mean_token_accuracy": 0.20406082421541213,
      "num_tokens": 91699406.0,
      "step": 36170
    },
    {
      "entropy": 6.096256923675537,
      "epoch": 4.1748413156376225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003373171617744939,
      "loss": 5.5064,
      "mean_token_accuracy": 0.21386712044477463,
      "num_tokens": 91711664.0,
      "step": 36175
    },
    {
      "entropy": 5.992486524581909,
      "epoch": 4.175418349682631,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033727751063794987,
      "loss": 5.4413,
      "mean_token_accuracy": 0.2214311107993126,
      "num_tokens": 91725622.0,
      "step": 36180
    },
    {
      "entropy": 6.040131330490112,
      "epoch": 4.17599538372764,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003372378574067328,
      "loss": 5.4766,
      "mean_token_accuracy": 0.21044879108667375,
      "num_tokens": 91737123.0,
      "step": 36185
    },
    {
      "entropy": 5.953706216812134,
      "epoch": 4.176572417772649,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003371982020821762,
      "loss": 5.4445,
      "mean_token_accuracy": 0.21513064205646515,
      "num_tokens": 91749774.0,
      "step": 36190
    },
    {
      "entropy": 6.060849285125732,
      "epoch": 4.177149451817657,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003371585446656141,
      "loss": 5.5457,
      "mean_token_accuracy": 0.2157866835594177,
      "num_tokens": 91762045.0,
      "step": 36195
    },
    {
      "entropy": 6.10784683227539,
      "epoch": 4.1777264858626655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003371188851583802,
      "loss": 5.5245,
      "mean_token_accuracy": 0.2092201367020607,
      "num_tokens": 91774277.0,
      "step": 36200
    },
    {
      "entropy": 6.055388689041138,
      "epoch": 4.178303519907675,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003370792235618084,
      "loss": 5.5769,
      "mean_token_accuracy": 0.20751824229955673,
      "num_tokens": 91786247.0,
      "step": 36205
    },
    {
      "entropy": 6.071255397796631,
      "epoch": 4.178880553952683,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033703955987723276,
      "loss": 5.4734,
      "mean_token_accuracy": 0.21457559019327163,
      "num_tokens": 91798964.0,
      "step": 36210
    },
    {
      "entropy": 5.987832403182983,
      "epoch": 4.179457587997692,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003369998941059874,
      "loss": 5.4681,
      "mean_token_accuracy": 0.21751263737678528,
      "num_tokens": 91812271.0,
      "step": 36215
    },
    {
      "entropy": 5.911073017120361,
      "epoch": 4.180034622042701,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003369602262494064,
      "loss": 5.4321,
      "mean_token_accuracy": 0.219881047308445,
      "num_tokens": 91826272.0,
      "step": 36220
    },
    {
      "entropy": 5.92738528251648,
      "epoch": 4.180611656087709,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033692055630882387,
      "loss": 5.4461,
      "mean_token_accuracy": 0.21457289755344391,
      "num_tokens": 91839012.0,
      "step": 36225
    },
    {
      "entropy": 6.102661943435669,
      "epoch": 4.181188690132718,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00033688088428557425,
      "loss": 5.5461,
      "mean_token_accuracy": 0.2107859417796135,
      "num_tokens": 91852768.0,
      "step": 36230
    },
    {
      "entropy": 6.023000717163086,
      "epoch": 4.181765724177726,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003368412101809918,
      "loss": 5.5019,
      "mean_token_accuracy": 0.21322470605373384,
      "num_tokens": 91865845.0,
      "step": 36235
    },
    {
      "entropy": 5.92189769744873,
      "epoch": 4.1823427582227355,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000336801533996411,
      "loss": 5.382,
      "mean_token_accuracy": 0.22423402220010757,
      "num_tokens": 91880151.0,
      "step": 36240
    },
    {
      "entropy": 6.0993907928466795,
      "epoch": 4.182919792267744,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003367618557331663,
      "loss": 5.5551,
      "mean_token_accuracy": 0.21040874421596528,
      "num_tokens": 91892565.0,
      "step": 36245
    },
    {
      "entropy": 6.1127783298492435,
      "epoch": 4.183496826312752,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003367221753925922,
      "loss": 5.5529,
      "mean_token_accuracy": 0.20397200882434846,
      "num_tokens": 91904504.0,
      "step": 36250
    },
    {
      "entropy": 6.143336200714112,
      "epoch": 4.184073860357761,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033668249297602343,
      "loss": 5.535,
      "mean_token_accuracy": 0.20913943350315095,
      "num_tokens": 91916822.0,
      "step": 36255
    },
    {
      "entropy": 6.063420104980469,
      "epoch": 4.18465089440277,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003366428084847947,
      "loss": 5.5257,
      "mean_token_accuracy": 0.2174711376428604,
      "num_tokens": 91930596.0,
      "step": 36260
    },
    {
      "entropy": 6.096650552749634,
      "epoch": 4.1852279284477785,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003366031219202406,
      "loss": 5.6003,
      "mean_token_accuracy": 0.21027973890304566,
      "num_tokens": 91944007.0,
      "step": 36265
    },
    {
      "entropy": 6.060778093338013,
      "epoch": 4.185804962492787,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033656343328369615,
      "loss": 5.4611,
      "mean_token_accuracy": 0.21989932805299758,
      "num_tokens": 91956374.0,
      "step": 36270
    },
    {
      "entropy": 6.046530914306641,
      "epoch": 4.186381996537795,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033652374257649625,
      "loss": 5.5896,
      "mean_token_accuracy": 0.20774901807308196,
      "num_tokens": 91967768.0,
      "step": 36275
    },
    {
      "entropy": 6.014452075958252,
      "epoch": 4.186959030582805,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003364840497999759,
      "loss": 5.4547,
      "mean_token_accuracy": 0.21547582298517226,
      "num_tokens": 91981660.0,
      "step": 36280
    },
    {
      "entropy": 5.987741231918335,
      "epoch": 4.187536064627813,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033644435495546994,
      "loss": 5.4055,
      "mean_token_accuracy": 0.22181856632232666,
      "num_tokens": 91994959.0,
      "step": 36285
    },
    {
      "entropy": 6.0608429431915285,
      "epoch": 4.188113098672821,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033640465804431365,
      "loss": 5.5555,
      "mean_token_accuracy": 0.20919679701328278,
      "num_tokens": 92007805.0,
      "step": 36290
    },
    {
      "entropy": 6.0707409381866455,
      "epoch": 4.188690132717831,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003363649590678422,
      "loss": 5.5156,
      "mean_token_accuracy": 0.21529030799865723,
      "num_tokens": 92020144.0,
      "step": 36295
    },
    {
      "entropy": 6.1062415599822994,
      "epoch": 4.189267166762839,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033632525802739085,
      "loss": 5.5759,
      "mean_token_accuracy": 0.20948645025491713,
      "num_tokens": 92033659.0,
      "step": 36300
    },
    {
      "entropy": 6.095681762695312,
      "epoch": 4.189844200807848,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033628555492429485,
      "loss": 5.5708,
      "mean_token_accuracy": 0.20862073749303817,
      "num_tokens": 92046569.0,
      "step": 36305
    },
    {
      "entropy": 6.024930620193482,
      "epoch": 4.190421234852856,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033624584975988967,
      "loss": 5.5196,
      "mean_token_accuracy": 0.21469273865222932,
      "num_tokens": 92059324.0,
      "step": 36310
    },
    {
      "entropy": 6.044795227050781,
      "epoch": 4.190998268897865,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003362061425355108,
      "loss": 5.4972,
      "mean_token_accuracy": 0.21813523173332214,
      "num_tokens": 92071971.0,
      "step": 36315
    },
    {
      "entropy": 6.094405221939087,
      "epoch": 4.191575302942874,
      "grad_norm": 1.0,
      "learning_rate": 0.0003361664332524937,
      "loss": 5.5328,
      "mean_token_accuracy": 0.21547052264213562,
      "num_tokens": 92083393.0,
      "step": 36320
    },
    {
      "entropy": 6.064061260223388,
      "epoch": 4.192152336987882,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033612672191217404,
      "loss": 5.4772,
      "mean_token_accuracy": 0.20958197116851807,
      "num_tokens": 92095321.0,
      "step": 36325
    },
    {
      "entropy": 6.076068639755249,
      "epoch": 4.192729371032891,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033608700851588737,
      "loss": 5.5472,
      "mean_token_accuracy": 0.20471175611019135,
      "num_tokens": 92108112.0,
      "step": 36330
    },
    {
      "entropy": 6.095101928710937,
      "epoch": 4.1933064050779,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033604729306496955,
      "loss": 5.6052,
      "mean_token_accuracy": 0.20969596952199937,
      "num_tokens": 92120965.0,
      "step": 36335
    },
    {
      "entropy": 6.094717407226563,
      "epoch": 4.193883439122908,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003360075755607564,
      "loss": 5.5452,
      "mean_token_accuracy": 0.2095690831542015,
      "num_tokens": 92132180.0,
      "step": 36340
    },
    {
      "entropy": 6.109365177154541,
      "epoch": 4.194460473167917,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003359678560045836,
      "loss": 5.507,
      "mean_token_accuracy": 0.20993925780057907,
      "num_tokens": 92144211.0,
      "step": 36345
    },
    {
      "entropy": 6.10888557434082,
      "epoch": 4.195037507212925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033592813439778743,
      "loss": 5.4497,
      "mean_token_accuracy": 0.21592015624046326,
      "num_tokens": 92156645.0,
      "step": 36350
    },
    {
      "entropy": 6.038032817840576,
      "epoch": 4.195614541257934,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003358884107417036,
      "loss": 5.5118,
      "mean_token_accuracy": 0.2106542021036148,
      "num_tokens": 92170324.0,
      "step": 36355
    },
    {
      "entropy": 6.061372947692871,
      "epoch": 4.196191575302943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003358486850376685,
      "loss": 5.5596,
      "mean_token_accuracy": 0.19926372468471526,
      "num_tokens": 92181691.0,
      "step": 36360
    },
    {
      "entropy": 6.024242639541626,
      "epoch": 4.196768609347951,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003358089572870179,
      "loss": 5.4796,
      "mean_token_accuracy": 0.21465114504098892,
      "num_tokens": 92193447.0,
      "step": 36365
    },
    {
      "entropy": 6.091886568069458,
      "epoch": 4.197345643392961,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003357692274910883,
      "loss": 5.5216,
      "mean_token_accuracy": 0.21384626775979995,
      "num_tokens": 92206269.0,
      "step": 36370
    },
    {
      "entropy": 6.0934326171875,
      "epoch": 4.197922677437969,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003357294956512159,
      "loss": 5.5558,
      "mean_token_accuracy": 0.20839828103780747,
      "num_tokens": 92219210.0,
      "step": 36375
    },
    {
      "entropy": 6.113658809661866,
      "epoch": 4.198499711482977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033568976176873714,
      "loss": 5.5904,
      "mean_token_accuracy": 0.20379132628440857,
      "num_tokens": 92232169.0,
      "step": 36380
    },
    {
      "entropy": 6.113116216659546,
      "epoch": 4.199076745527986,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003356500258449883,
      "loss": 5.5544,
      "mean_token_accuracy": 0.20726522505283357,
      "num_tokens": 92245249.0,
      "step": 36385
    },
    {
      "entropy": 6.061218643188477,
      "epoch": 4.199653779572995,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003356102878813061,
      "loss": 5.56,
      "mean_token_accuracy": 0.2075910300016403,
      "num_tokens": 92256545.0,
      "step": 36390
    },
    {
      "entropy": 6.054666137695312,
      "epoch": 4.2002308136180035,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003355705478790268,
      "loss": 5.5732,
      "mean_token_accuracy": 0.20639955699443818,
      "num_tokens": 92268819.0,
      "step": 36395
    },
    {
      "entropy": 6.169286251068115,
      "epoch": 4.200807847663012,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003355308058394874,
      "loss": 5.6439,
      "mean_token_accuracy": 0.2057584270834923,
      "num_tokens": 92280746.0,
      "step": 36400
    },
    {
      "entropy": 6.145860815048218,
      "epoch": 4.20138488170802,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033549106176402433,
      "loss": 5.5736,
      "mean_token_accuracy": 0.20965632945299148,
      "num_tokens": 92293368.0,
      "step": 36405
    },
    {
      "entropy": 6.163086891174316,
      "epoch": 4.20196191575303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033545131565397443,
      "loss": 5.6941,
      "mean_token_accuracy": 0.199113629758358,
      "num_tokens": 92305395.0,
      "step": 36410
    },
    {
      "entropy": 6.039909076690674,
      "epoch": 4.202538949798038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003354115675106745,
      "loss": 5.4674,
      "mean_token_accuracy": 0.2235013023018837,
      "num_tokens": 92320196.0,
      "step": 36415
    },
    {
      "entropy": 6.049898862838745,
      "epoch": 4.2031159838430465,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003353718173354615,
      "loss": 5.5441,
      "mean_token_accuracy": 0.21430354863405227,
      "num_tokens": 92332530.0,
      "step": 36420
    },
    {
      "entropy": 6.104065847396851,
      "epoch": 4.203693017888055,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003353320651296725,
      "loss": 5.5714,
      "mean_token_accuracy": 0.21216470152139663,
      "num_tokens": 92343985.0,
      "step": 36425
    },
    {
      "entropy": 5.988610410690308,
      "epoch": 4.204270051933064,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033529231089464445,
      "loss": 5.4763,
      "mean_token_accuracy": 0.21894935220479966,
      "num_tokens": 92356222.0,
      "step": 36430
    },
    {
      "entropy": 5.986466312408448,
      "epoch": 4.204847085978073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033525255463171444,
      "loss": 5.4053,
      "mean_token_accuracy": 0.22800421565771103,
      "num_tokens": 92368734.0,
      "step": 36435
    },
    {
      "entropy": 6.037365818023682,
      "epoch": 4.205424120023081,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033521279634221975,
      "loss": 5.519,
      "mean_token_accuracy": 0.21274163573980331,
      "num_tokens": 92382360.0,
      "step": 36440
    },
    {
      "entropy": 6.023008537292481,
      "epoch": 4.20600115406809,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033517303602749747,
      "loss": 5.5306,
      "mean_token_accuracy": 0.21234100908041,
      "num_tokens": 92394924.0,
      "step": 36445
    },
    {
      "entropy": 6.060986566543579,
      "epoch": 4.206578188113099,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033513327368888513,
      "loss": 5.5807,
      "mean_token_accuracy": 0.2081702783703804,
      "num_tokens": 92407939.0,
      "step": 36450
    },
    {
      "entropy": 6.071784448623657,
      "epoch": 4.207155222158107,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003350935093277199,
      "loss": 5.5056,
      "mean_token_accuracy": 0.21452544629573822,
      "num_tokens": 92419528.0,
      "step": 36455
    },
    {
      "entropy": 6.034239482879639,
      "epoch": 4.207732256203116,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033505374294533934,
      "loss": 5.4754,
      "mean_token_accuracy": 0.2186982214450836,
      "num_tokens": 92431244.0,
      "step": 36460
    },
    {
      "entropy": 5.965225553512573,
      "epoch": 4.208309290248125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033501397454308094,
      "loss": 5.3508,
      "mean_token_accuracy": 0.2274266764521599,
      "num_tokens": 92444912.0,
      "step": 36465
    },
    {
      "entropy": 5.927420282363892,
      "epoch": 4.208886324293133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003349742041222824,
      "loss": 5.4381,
      "mean_token_accuracy": 0.22362808734178544,
      "num_tokens": 92458302.0,
      "step": 36470
    },
    {
      "entropy": 6.131619548797607,
      "epoch": 4.209463358338142,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003349344316842812,
      "loss": 5.6113,
      "mean_token_accuracy": 0.2088275134563446,
      "num_tokens": 92470605.0,
      "step": 36475
    },
    {
      "entropy": 6.083338642120362,
      "epoch": 4.21004039238315,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003348946572304153,
      "loss": 5.5149,
      "mean_token_accuracy": 0.2127673640847206,
      "num_tokens": 92483139.0,
      "step": 36480
    },
    {
      "entropy": 6.062240028381348,
      "epoch": 4.2106174264281595,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003348548807620223,
      "loss": 5.5443,
      "mean_token_accuracy": 0.21471376419067384,
      "num_tokens": 92495334.0,
      "step": 36485
    },
    {
      "entropy": 6.106533813476562,
      "epoch": 4.211194460473168,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033481510228044006,
      "loss": 5.6199,
      "mean_token_accuracy": 0.2119092807173729,
      "num_tokens": 92508362.0,
      "step": 36490
    },
    {
      "entropy": 6.065465354919434,
      "epoch": 4.211771494518176,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003347753217870066,
      "loss": 5.5149,
      "mean_token_accuracy": 0.21460752338171005,
      "num_tokens": 92520869.0,
      "step": 36495
    },
    {
      "entropy": 5.977196121215821,
      "epoch": 4.212348528563186,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003347355392830599,
      "loss": 5.4589,
      "mean_token_accuracy": 0.21670080870389938,
      "num_tokens": 92532615.0,
      "step": 36500
    },
    {
      "entropy": 6.0456725597381595,
      "epoch": 4.212925562608194,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000334695754769938,
      "loss": 5.4946,
      "mean_token_accuracy": 0.22031136602163315,
      "num_tokens": 92544813.0,
      "step": 36505
    },
    {
      "entropy": 6.1500200748443605,
      "epoch": 4.2135025966532025,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000334655968248979,
      "loss": 5.602,
      "mean_token_accuracy": 0.20118348002433778,
      "num_tokens": 92557256.0,
      "step": 36510
    },
    {
      "entropy": 6.110830974578858,
      "epoch": 4.214079630698211,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003346161797215211,
      "loss": 5.5547,
      "mean_token_accuracy": 0.21246470212936402,
      "num_tokens": 92569629.0,
      "step": 36515
    },
    {
      "entropy": 6.104619359970092,
      "epoch": 4.21465666474322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033457638918890273,
      "loss": 5.561,
      "mean_token_accuracy": 0.210816653072834,
      "num_tokens": 92582543.0,
      "step": 36520
    },
    {
      "entropy": 6.102493762969971,
      "epoch": 4.215233698788229,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033453659665246205,
      "loss": 5.5936,
      "mean_token_accuracy": 0.2060997098684311,
      "num_tokens": 92594708.0,
      "step": 36525
    },
    {
      "entropy": 6.144442081451416,
      "epoch": 4.215810732833237,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033449680211353753,
      "loss": 5.5695,
      "mean_token_accuracy": 0.20857217460870742,
      "num_tokens": 92607430.0,
      "step": 36530
    },
    {
      "entropy": 6.041609287261963,
      "epoch": 4.2163877668782455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033445700557346766,
      "loss": 5.4364,
      "mean_token_accuracy": 0.2212654948234558,
      "num_tokens": 92618919.0,
      "step": 36535
    },
    {
      "entropy": 6.048991060256958,
      "epoch": 4.216964800923255,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003344172070335908,
      "loss": 5.5685,
      "mean_token_accuracy": 0.2035675898194313,
      "num_tokens": 92630092.0,
      "step": 36540
    },
    {
      "entropy": 6.154643964767456,
      "epoch": 4.217541834968263,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033437740649524584,
      "loss": 5.6238,
      "mean_token_accuracy": 0.2028792291879654,
      "num_tokens": 92642108.0,
      "step": 36545
    },
    {
      "entropy": 6.219081830978394,
      "epoch": 4.218118869013272,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003343376039597712,
      "loss": 5.6485,
      "mean_token_accuracy": 0.2034117251634598,
      "num_tokens": 92653662.0,
      "step": 36550
    },
    {
      "entropy": 6.084265661239624,
      "epoch": 4.21869590305828,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003342977994285058,
      "loss": 5.4962,
      "mean_token_accuracy": 0.21776273250579833,
      "num_tokens": 92665899.0,
      "step": 36555
    },
    {
      "entropy": 6.046296310424805,
      "epoch": 4.219272937103289,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003342579929027883,
      "loss": 5.5022,
      "mean_token_accuracy": 0.21380335986614227,
      "num_tokens": 92677475.0,
      "step": 36560
    },
    {
      "entropy": 6.15334997177124,
      "epoch": 4.219849971148298,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003342181843839578,
      "loss": 5.5629,
      "mean_token_accuracy": 0.21173277050256728,
      "num_tokens": 92690629.0,
      "step": 36565
    },
    {
      "entropy": 6.116700792312622,
      "epoch": 4.220427005193306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033417837387335293,
      "loss": 5.5598,
      "mean_token_accuracy": 0.20802867859601976,
      "num_tokens": 92702274.0,
      "step": 36570
    },
    {
      "entropy": 6.059780836105347,
      "epoch": 4.2210040392383155,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003341385613723129,
      "loss": 5.5397,
      "mean_token_accuracy": 0.20720929503440857,
      "num_tokens": 92715779.0,
      "step": 36575
    },
    {
      "entropy": 6.0751426219940186,
      "epoch": 4.221581073283324,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00033409874688217667,
      "loss": 5.5522,
      "mean_token_accuracy": 0.211833356320858,
      "num_tokens": 92727316.0,
      "step": 36580
    },
    {
      "entropy": 6.0430460453033445,
      "epoch": 4.222158107328332,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003340589304042836,
      "loss": 5.5333,
      "mean_token_accuracy": 0.21074459999799727,
      "num_tokens": 92741004.0,
      "step": 36585
    },
    {
      "entropy": 6.153020763397217,
      "epoch": 4.222735141373341,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00033401911193997266,
      "loss": 5.6303,
      "mean_token_accuracy": 0.19995790719985962,
      "num_tokens": 92754748.0,
      "step": 36590
    },
    {
      "entropy": 6.20211763381958,
      "epoch": 4.22331217541835,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00033397929149058317,
      "loss": 5.6376,
      "mean_token_accuracy": 0.2024789646267891,
      "num_tokens": 92766697.0,
      "step": 36595
    },
    {
      "entropy": 6.103918027877808,
      "epoch": 4.223889209463358,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033393946905745456,
      "loss": 5.5413,
      "mean_token_accuracy": 0.21390598863363267,
      "num_tokens": 92779283.0,
      "step": 36600
    },
    {
      "entropy": 6.045347070693969,
      "epoch": 4.224466243508367,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003338996446419261,
      "loss": 5.5728,
      "mean_token_accuracy": 0.2157061830163002,
      "num_tokens": 92792454.0,
      "step": 36605
    },
    {
      "entropy": 6.013524293899536,
      "epoch": 4.225043277553375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033385981824533747,
      "loss": 5.5023,
      "mean_token_accuracy": 0.21211120933294297,
      "num_tokens": 92804446.0,
      "step": 36610
    },
    {
      "entropy": 6.01087760925293,
      "epoch": 4.225620311598385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000333819989869028,
      "loss": 5.4885,
      "mean_token_accuracy": 0.20864471644163132,
      "num_tokens": 92817719.0,
      "step": 36615
    },
    {
      "entropy": 6.026328897476196,
      "epoch": 4.226197345643393,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003337801595143374,
      "loss": 5.4778,
      "mean_token_accuracy": 0.21421949714422225,
      "num_tokens": 92830224.0,
      "step": 36620
    },
    {
      "entropy": 6.086988019943237,
      "epoch": 4.226774379688401,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033374032718260535,
      "loss": 5.6056,
      "mean_token_accuracy": 0.2036181226372719,
      "num_tokens": 92841349.0,
      "step": 36625
    },
    {
      "entropy": 6.104330730438233,
      "epoch": 4.227351413733411,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003337004928751716,
      "loss": 5.5096,
      "mean_token_accuracy": 0.21468435525894164,
      "num_tokens": 92853713.0,
      "step": 36630
    },
    {
      "entropy": 6.043995904922485,
      "epoch": 4.227928447778419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033366065659337586,
      "loss": 5.4671,
      "mean_token_accuracy": 0.21673654168844222,
      "num_tokens": 92865767.0,
      "step": 36635
    },
    {
      "entropy": 6.025083875656128,
      "epoch": 4.228505481823428,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003336208183385581,
      "loss": 5.4773,
      "mean_token_accuracy": 0.21581257581710817,
      "num_tokens": 92878487.0,
      "step": 36640
    },
    {
      "entropy": 6.000425386428833,
      "epoch": 4.229082515868436,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003335809781120583,
      "loss": 5.4712,
      "mean_token_accuracy": 0.2150888293981552,
      "num_tokens": 92890515.0,
      "step": 36645
    },
    {
      "entropy": 6.124005365371704,
      "epoch": 4.229659549913445,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003335411359152163,
      "loss": 5.6558,
      "mean_token_accuracy": 0.20216683000326158,
      "num_tokens": 92902313.0,
      "step": 36650
    },
    {
      "entropy": 5.906117630004883,
      "epoch": 4.230236583958454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033350129174937237,
      "loss": 5.3512,
      "mean_token_accuracy": 0.23351780623197554,
      "num_tokens": 92915419.0,
      "step": 36655
    },
    {
      "entropy": 5.958498239517212,
      "epoch": 4.230813618003462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003334614456158665,
      "loss": 5.447,
      "mean_token_accuracy": 0.2175373911857605,
      "num_tokens": 92927631.0,
      "step": 36660
    },
    {
      "entropy": 6.165533971786499,
      "epoch": 4.2313906520484705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033342159751603896,
      "loss": 5.6076,
      "mean_token_accuracy": 0.20217154771089554,
      "num_tokens": 92940312.0,
      "step": 36665
    },
    {
      "entropy": 6.154601144790649,
      "epoch": 4.23196768609348,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033338174745123003,
      "loss": 5.5302,
      "mean_token_accuracy": 0.21061812490224838,
      "num_tokens": 92953103.0,
      "step": 36670
    },
    {
      "entropy": 6.135361814498902,
      "epoch": 4.232544720138488,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033334189542277994,
      "loss": 5.6492,
      "mean_token_accuracy": 0.20071909427642823,
      "num_tokens": 92967033.0,
      "step": 36675
    },
    {
      "entropy": 6.062571287155151,
      "epoch": 4.233121754183497,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033330204143202925,
      "loss": 5.6108,
      "mean_token_accuracy": 0.2030896618962288,
      "num_tokens": 92978840.0,
      "step": 36680
    },
    {
      "entropy": 5.999409532546997,
      "epoch": 4.233698788228505,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003332621854803183,
      "loss": 5.407,
      "mean_token_accuracy": 0.22284828275442123,
      "num_tokens": 92991370.0,
      "step": 36685
    },
    {
      "entropy": 6.072699642181396,
      "epoch": 4.234275822273514,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003332223275689877,
      "loss": 5.4929,
      "mean_token_accuracy": 0.2187053382396698,
      "num_tokens": 93004744.0,
      "step": 36690
    },
    {
      "entropy": 6.136489057540894,
      "epoch": 4.234852856318523,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003331824676993781,
      "loss": 5.5942,
      "mean_token_accuracy": 0.20887219458818435,
      "num_tokens": 93018395.0,
      "step": 36695
    },
    {
      "entropy": 6.07927885055542,
      "epoch": 4.235429890363531,
      "grad_norm": 1.0,
      "learning_rate": 0.00033314260587283,
      "loss": 5.5265,
      "mean_token_accuracy": 0.21734849065542222,
      "num_tokens": 93030912.0,
      "step": 36700
    },
    {
      "entropy": 6.0258125305175785,
      "epoch": 4.2360069244085405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003331027420906843,
      "loss": 5.5189,
      "mean_token_accuracy": 0.2087923526763916,
      "num_tokens": 93043439.0,
      "step": 36705
    },
    {
      "entropy": 6.119767808914185,
      "epoch": 4.236583958453549,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033306287635428175,
      "loss": 5.5733,
      "mean_token_accuracy": 0.20359193831682204,
      "num_tokens": 93055803.0,
      "step": 36710
    },
    {
      "entropy": 6.054043531417847,
      "epoch": 4.237160992498557,
      "grad_norm": 1.0,
      "learning_rate": 0.0003330230086649631,
      "loss": 5.5695,
      "mean_token_accuracy": 0.2102679044008255,
      "num_tokens": 93068495.0,
      "step": 36715
    },
    {
      "entropy": 5.954505348205567,
      "epoch": 4.237738026543566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033298313902406946,
      "loss": 5.3865,
      "mean_token_accuracy": 0.22322705388069153,
      "num_tokens": 93082406.0,
      "step": 36720
    },
    {
      "entropy": 6.086103057861328,
      "epoch": 4.238315060588575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033294326743294166,
      "loss": 5.4831,
      "mean_token_accuracy": 0.21080585569143295,
      "num_tokens": 93094606.0,
      "step": 36725
    },
    {
      "entropy": 6.086388778686524,
      "epoch": 4.2388920946335835,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003329033938929209,
      "loss": 5.5468,
      "mean_token_accuracy": 0.2204357847571373,
      "num_tokens": 93107389.0,
      "step": 36730
    },
    {
      "entropy": 6.087401294708252,
      "epoch": 4.239469128678592,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00033286351840534835,
      "loss": 5.4878,
      "mean_token_accuracy": 0.2128287747502327,
      "num_tokens": 93119022.0,
      "step": 36735
    },
    {
      "entropy": 6.032084798812866,
      "epoch": 4.2400461627236,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00033282364097156494,
      "loss": 5.4722,
      "mean_token_accuracy": 0.21232181191444396,
      "num_tokens": 93132636.0,
      "step": 36740
    },
    {
      "entropy": 6.060864877700806,
      "epoch": 4.24062319676861,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003327837615929121,
      "loss": 5.4977,
      "mean_token_accuracy": 0.21614027321338652,
      "num_tokens": 93145454.0,
      "step": 36745
    },
    {
      "entropy": 6.120479774475098,
      "epoch": 4.241200230813618,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033274388027073123,
      "loss": 5.5736,
      "mean_token_accuracy": 0.20774539709091186,
      "num_tokens": 93158465.0,
      "step": 36750
    },
    {
      "entropy": 6.056994533538818,
      "epoch": 4.2417772648586265,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033270399700636354,
      "loss": 5.5093,
      "mean_token_accuracy": 0.2096223771572113,
      "num_tokens": 93171285.0,
      "step": 36755
    },
    {
      "entropy": 6.1011322975158695,
      "epoch": 4.242354298903635,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003326641118011506,
      "loss": 5.5567,
      "mean_token_accuracy": 0.20576124340295793,
      "num_tokens": 93182402.0,
      "step": 36760
    },
    {
      "entropy": 6.051559972763061,
      "epoch": 4.242931332948644,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000332624224656434,
      "loss": 5.5538,
      "mean_token_accuracy": 0.21190761774778366,
      "num_tokens": 93194719.0,
      "step": 36765
    },
    {
      "entropy": 6.116247129440308,
      "epoch": 4.243508366993653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003325843355735551,
      "loss": 5.5583,
      "mean_token_accuracy": 0.20830961614847182,
      "num_tokens": 93207246.0,
      "step": 36770
    },
    {
      "entropy": 6.14764633178711,
      "epoch": 4.244085401038661,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033254444455385574,
      "loss": 5.5905,
      "mean_token_accuracy": 0.2091515764594078,
      "num_tokens": 93219838.0,
      "step": 36775
    },
    {
      "entropy": 6.14138970375061,
      "epoch": 4.24466243508367,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033250455159867755,
      "loss": 5.647,
      "mean_token_accuracy": 0.2065313458442688,
      "num_tokens": 93231690.0,
      "step": 36780
    },
    {
      "entropy": 6.070845460891723,
      "epoch": 4.245239469128679,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003324646567093624,
      "loss": 5.5388,
      "mean_token_accuracy": 0.21307261288166046,
      "num_tokens": 93245489.0,
      "step": 36785
    },
    {
      "entropy": 6.175360441207886,
      "epoch": 4.245816503173687,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003324247598872519,
      "loss": 5.6603,
      "mean_token_accuracy": 0.20349411368370057,
      "num_tokens": 93258532.0,
      "step": 36790
    },
    {
      "entropy": 6.025357055664062,
      "epoch": 4.246393537218696,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033238486113368825,
      "loss": 5.4999,
      "mean_token_accuracy": 0.21715618371963502,
      "num_tokens": 93271176.0,
      "step": 36795
    },
    {
      "entropy": 6.035335350036621,
      "epoch": 4.246970571263705,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003323449604500133,
      "loss": 5.5176,
      "mean_token_accuracy": 0.2138695552945137,
      "num_tokens": 93282476.0,
      "step": 36800
    },
    {
      "entropy": 6.027915573120117,
      "epoch": 4.247547605308713,
      "grad_norm": 1.03125,
      "learning_rate": 0.000332305057837569,
      "loss": 5.4676,
      "mean_token_accuracy": 0.21152154803276063,
      "num_tokens": 93294597.0,
      "step": 36805
    },
    {
      "entropy": 6.145900249481201,
      "epoch": 4.248124639353722,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003322651532976976,
      "loss": 5.5399,
      "mean_token_accuracy": 0.20556803196668624,
      "num_tokens": 93305324.0,
      "step": 36810
    },
    {
      "entropy": 5.961233520507813,
      "epoch": 4.24870167339873,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033222524683174126,
      "loss": 5.4928,
      "mean_token_accuracy": 0.21724716573953629,
      "num_tokens": 93319856.0,
      "step": 36815
    },
    {
      "entropy": 6.082091474533081,
      "epoch": 4.2492787074437395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033218533844104215,
      "loss": 5.51,
      "mean_token_accuracy": 0.21420908868312835,
      "num_tokens": 93331965.0,
      "step": 36820
    },
    {
      "entropy": 5.999950170516968,
      "epoch": 4.249855741488748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003321454281269426,
      "loss": 5.4852,
      "mean_token_accuracy": 0.20520174205303193,
      "num_tokens": 93344293.0,
      "step": 36825
    },
    {
      "entropy": 5.913121223449707,
      "epoch": 4.250432775533756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003321055158907849,
      "loss": 5.3519,
      "mean_token_accuracy": 0.2235862523317337,
      "num_tokens": 93356959.0,
      "step": 36830
    },
    {
      "entropy": 6.064392614364624,
      "epoch": 4.251009809578765,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003320656017339116,
      "loss": 5.5388,
      "mean_token_accuracy": 0.2102576643228531,
      "num_tokens": 93370810.0,
      "step": 36835
    },
    {
      "entropy": 6.074534559249878,
      "epoch": 4.251586843623774,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003320256856576651,
      "loss": 5.5458,
      "mean_token_accuracy": 0.21450243890285492,
      "num_tokens": 93383100.0,
      "step": 36840
    },
    {
      "entropy": 6.088781833648682,
      "epoch": 4.2521638776687825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000331985767663388,
      "loss": 5.5655,
      "mean_token_accuracy": 0.20969249159097672,
      "num_tokens": 93396091.0,
      "step": 36845
    },
    {
      "entropy": 6.115301942825317,
      "epoch": 4.252740911713791,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033194584775242294,
      "loss": 5.5624,
      "mean_token_accuracy": 0.2136012688279152,
      "num_tokens": 93409658.0,
      "step": 36850
    },
    {
      "entropy": 6.0740166187286375,
      "epoch": 4.2533179457588,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033190592592611254,
      "loss": 5.4885,
      "mean_token_accuracy": 0.2175256922841072,
      "num_tokens": 93420865.0,
      "step": 36855
    },
    {
      "entropy": 6.0554931640625,
      "epoch": 4.253894979803809,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003318660021857996,
      "loss": 5.5076,
      "mean_token_accuracy": 0.21305003017187119,
      "num_tokens": 93432995.0,
      "step": 36860
    },
    {
      "entropy": 6.118357181549072,
      "epoch": 4.254472013848817,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003318260765328269,
      "loss": 5.6151,
      "mean_token_accuracy": 0.2058679550886154,
      "num_tokens": 93445463.0,
      "step": 36865
    },
    {
      "entropy": 6.068684053421021,
      "epoch": 4.255049047893825,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033178614896853736,
      "loss": 5.5317,
      "mean_token_accuracy": 0.20980582684278487,
      "num_tokens": 93458153.0,
      "step": 36870
    },
    {
      "entropy": 6.075751638412475,
      "epoch": 4.255626081938835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003317462194942739,
      "loss": 5.4812,
      "mean_token_accuracy": 0.21523391753435134,
      "num_tokens": 93470800.0,
      "step": 36875
    },
    {
      "entropy": 6.078428554534912,
      "epoch": 4.256203115983843,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033170628811137956,
      "loss": 5.555,
      "mean_token_accuracy": 0.206775526702404,
      "num_tokens": 93483960.0,
      "step": 36880
    },
    {
      "entropy": 5.962617826461792,
      "epoch": 4.256780150028852,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003316663548211973,
      "loss": 5.464,
      "mean_token_accuracy": 0.22148096412420273,
      "num_tokens": 93497583.0,
      "step": 36885
    },
    {
      "entropy": 6.092488098144531,
      "epoch": 4.25735718407386,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033162641962507044,
      "loss": 5.5205,
      "mean_token_accuracy": 0.21008556187152863,
      "num_tokens": 93510009.0,
      "step": 36890
    },
    {
      "entropy": 6.0640229225158695,
      "epoch": 4.257934218118869,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000331586482524342,
      "loss": 5.5107,
      "mean_token_accuracy": 0.2138819396495819,
      "num_tokens": 93522928.0,
      "step": 36895
    },
    {
      "entropy": 6.037232398986816,
      "epoch": 4.258511252163878,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033154654352035546,
      "loss": 5.4735,
      "mean_token_accuracy": 0.22069062441587448,
      "num_tokens": 93536092.0,
      "step": 36900
    },
    {
      "entropy": 6.097039747238159,
      "epoch": 4.259088286208886,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003315066026144539,
      "loss": 5.6105,
      "mean_token_accuracy": 0.20378256738185882,
      "num_tokens": 93550257.0,
      "step": 36905
    },
    {
      "entropy": 6.082575845718384,
      "epoch": 4.2596653202538945,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003314666598079809,
      "loss": 5.5744,
      "mean_token_accuracy": 0.20678354501724244,
      "num_tokens": 93562603.0,
      "step": 36910
    },
    {
      "entropy": 6.015063047409058,
      "epoch": 4.260242354298904,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003314267151022798,
      "loss": 5.4917,
      "mean_token_accuracy": 0.2138369455933571,
      "num_tokens": 93574888.0,
      "step": 36915
    },
    {
      "entropy": 6.097263336181641,
      "epoch": 4.260819388343912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003313867684986942,
      "loss": 5.5081,
      "mean_token_accuracy": 0.21456323117017745,
      "num_tokens": 93588471.0,
      "step": 36920
    },
    {
      "entropy": 5.952257871627808,
      "epoch": 4.261396422388921,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033134681999856763,
      "loss": 5.4884,
      "mean_token_accuracy": 0.21704992204904555,
      "num_tokens": 93601292.0,
      "step": 36925
    },
    {
      "entropy": 6.0868504524230955,
      "epoch": 4.26197345643393,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003313068696032438,
      "loss": 5.5075,
      "mean_token_accuracy": 0.21078077107667922,
      "num_tokens": 93613495.0,
      "step": 36930
    },
    {
      "entropy": 6.040879726409912,
      "epoch": 4.262550490478938,
      "grad_norm": 0.859375,
      "learning_rate": 0.00033126691731406627,
      "loss": 5.4756,
      "mean_token_accuracy": 0.20989009588956833,
      "num_tokens": 93626961.0,
      "step": 36935
    },
    {
      "entropy": 6.091657638549805,
      "epoch": 4.263127524523947,
      "grad_norm": 1.0,
      "learning_rate": 0.000331226963132379,
      "loss": 5.6326,
      "mean_token_accuracy": 0.21292945146560668,
      "num_tokens": 93640268.0,
      "step": 36940
    },
    {
      "entropy": 5.944855117797852,
      "epoch": 4.263704558568955,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033118700705952575,
      "loss": 5.4868,
      "mean_token_accuracy": 0.21734652817249298,
      "num_tokens": 93653224.0,
      "step": 36945
    },
    {
      "entropy": 6.086842107772827,
      "epoch": 4.2642815926139646,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033114704909685045,
      "loss": 5.5676,
      "mean_token_accuracy": 0.20482129603624344,
      "num_tokens": 93665986.0,
      "step": 36950
    },
    {
      "entropy": 6.149227237701416,
      "epoch": 4.264858626658973,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000331107089245697,
      "loss": 5.5039,
      "mean_token_accuracy": 0.2117534503340721,
      "num_tokens": 93677353.0,
      "step": 36955
    },
    {
      "entropy": 6.113183879852295,
      "epoch": 4.265435660703981,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033106712750740946,
      "loss": 5.6199,
      "mean_token_accuracy": 0.19657696187496185,
      "num_tokens": 93690115.0,
      "step": 36960
    },
    {
      "entropy": 6.13093409538269,
      "epoch": 4.26601269474899,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000331027163883332,
      "loss": 5.5835,
      "mean_token_accuracy": 0.21113261431455613,
      "num_tokens": 93703101.0,
      "step": 36965
    },
    {
      "entropy": 6.088113117218017,
      "epoch": 4.266589728793999,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033098719837480875,
      "loss": 5.5843,
      "mean_token_accuracy": 0.2116773709654808,
      "num_tokens": 93715846.0,
      "step": 36970
    },
    {
      "entropy": 5.979647922515869,
      "epoch": 4.2671667628390075,
      "grad_norm": 0.890625,
      "learning_rate": 0.00033094723098318375,
      "loss": 5.4267,
      "mean_token_accuracy": 0.2239473730325699,
      "num_tokens": 93728678.0,
      "step": 36975
    },
    {
      "entropy": 6.024152231216431,
      "epoch": 4.267743796884016,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033090726170980153,
      "loss": 5.526,
      "mean_token_accuracy": 0.21221805214881898,
      "num_tokens": 93740189.0,
      "step": 36980
    },
    {
      "entropy": 6.0541150093078615,
      "epoch": 4.268320830929024,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033086729055600627,
      "loss": 5.5646,
      "mean_token_accuracy": 0.21142414063215256,
      "num_tokens": 93753292.0,
      "step": 36985
    },
    {
      "entropy": 6.075201845169067,
      "epoch": 4.268897864974034,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003308273175231425,
      "loss": 5.4829,
      "mean_token_accuracy": 0.215493343770504,
      "num_tokens": 93765732.0,
      "step": 36990
    },
    {
      "entropy": 6.039456129074097,
      "epoch": 4.269474899019042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003307873426125546,
      "loss": 5.5046,
      "mean_token_accuracy": 0.21588677316904067,
      "num_tokens": 93777372.0,
      "step": 36995
    },
    {
      "entropy": 6.061505937576294,
      "epoch": 4.2700519330640505,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003307473658255871,
      "loss": 5.5285,
      "mean_token_accuracy": 0.20815494805574417,
      "num_tokens": 93790537.0,
      "step": 37000
    },
    {
      "entropy": 6.118732738494873,
      "epoch": 4.27062896710906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00033070738716358474,
      "loss": 5.5863,
      "mean_token_accuracy": 0.20679013282060624,
      "num_tokens": 93803374.0,
      "step": 37005
    },
    {
      "entropy": 6.105179262161255,
      "epoch": 4.271206001154068,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003306674066278921,
      "loss": 5.5664,
      "mean_token_accuracy": 0.2113454282283783,
      "num_tokens": 93815289.0,
      "step": 37010
    },
    {
      "entropy": 6.068741369247436,
      "epoch": 4.271783035199077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033062742421985375,
      "loss": 5.5883,
      "mean_token_accuracy": 0.20627841353416443,
      "num_tokens": 93828313.0,
      "step": 37015
    },
    {
      "entropy": 6.086719226837158,
      "epoch": 4.272360069244085,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033058743994081476,
      "loss": 5.54,
      "mean_token_accuracy": 0.21010360717773438,
      "num_tokens": 93842322.0,
      "step": 37020
    },
    {
      "entropy": 6.039522695541382,
      "epoch": 4.272937103289094,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00033054745379211976,
      "loss": 5.502,
      "mean_token_accuracy": 0.21405955404043198,
      "num_tokens": 93855661.0,
      "step": 37025
    },
    {
      "entropy": 6.038384532928466,
      "epoch": 4.273514137334103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033050746577511375,
      "loss": 5.4888,
      "mean_token_accuracy": 0.21556802839040756,
      "num_tokens": 93867319.0,
      "step": 37030
    },
    {
      "entropy": 6.121209049224854,
      "epoch": 4.274091171379111,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003304674758911417,
      "loss": 5.6246,
      "mean_token_accuracy": 0.2007503941655159,
      "num_tokens": 93879868.0,
      "step": 37035
    },
    {
      "entropy": 6.130085563659668,
      "epoch": 4.27466820542412,
      "grad_norm": 1.0,
      "learning_rate": 0.0003304274841415487,
      "loss": 5.5391,
      "mean_token_accuracy": 0.21511986702680588,
      "num_tokens": 93892312.0,
      "step": 37040
    },
    {
      "entropy": 6.008040237426758,
      "epoch": 4.275245239469129,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00033038749052767975,
      "loss": 5.497,
      "mean_token_accuracy": 0.21065040826797485,
      "num_tokens": 93905465.0,
      "step": 37045
    },
    {
      "entropy": 6.016496658325195,
      "epoch": 4.275822273514137,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033034749505088005,
      "loss": 5.5083,
      "mean_token_accuracy": 0.21184003353118896,
      "num_tokens": 93917777.0,
      "step": 37050
    },
    {
      "entropy": 6.104601287841797,
      "epoch": 4.276399307559146,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000330307497712495,
      "loss": 5.5522,
      "mean_token_accuracy": 0.20863109678030015,
      "num_tokens": 93931029.0,
      "step": 37055
    },
    {
      "entropy": 6.069776439666748,
      "epoch": 4.276976341604154,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003302674985138695,
      "loss": 5.529,
      "mean_token_accuracy": 0.21211292147636412,
      "num_tokens": 93944443.0,
      "step": 37060
    },
    {
      "entropy": 6.115158033370972,
      "epoch": 4.2775533756491635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033022749745634933,
      "loss": 5.6339,
      "mean_token_accuracy": 0.20906228125095366,
      "num_tokens": 93956793.0,
      "step": 37065
    },
    {
      "entropy": 6.147492408752441,
      "epoch": 4.278130409694172,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00033018749454127963,
      "loss": 5.5958,
      "mean_token_accuracy": 0.20936065018177033,
      "num_tokens": 93969635.0,
      "step": 37070
    },
    {
      "entropy": 6.094863986968994,
      "epoch": 4.27870744373918,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033014748977000595,
      "loss": 5.6128,
      "mean_token_accuracy": 0.210265588760376,
      "num_tokens": 93982360.0,
      "step": 37075
    },
    {
      "entropy": 6.153415584564209,
      "epoch": 4.27928447778419,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033010748314387387,
      "loss": 5.5754,
      "mean_token_accuracy": 0.21480638682842254,
      "num_tokens": 93995760.0,
      "step": 37080
    },
    {
      "entropy": 6.15511794090271,
      "epoch": 4.279861511829198,
      "grad_norm": 1.03125,
      "learning_rate": 0.000330067474664229,
      "loss": 5.5941,
      "mean_token_accuracy": 0.2037149429321289,
      "num_tokens": 94009560.0,
      "step": 37085
    },
    {
      "entropy": 6.0596918106079105,
      "epoch": 4.2804385458742065,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003300274643324169,
      "loss": 5.5223,
      "mean_token_accuracy": 0.21079382300376892,
      "num_tokens": 94022122.0,
      "step": 37090
    },
    {
      "entropy": 5.946433782577515,
      "epoch": 4.281015579919215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032998745214978344,
      "loss": 5.4367,
      "mean_token_accuracy": 0.2178565517067909,
      "num_tokens": 94034393.0,
      "step": 37095
    },
    {
      "entropy": 6.020993375778199,
      "epoch": 4.281592613964224,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003299474381176743,
      "loss": 5.4577,
      "mean_token_accuracy": 0.2173493206501007,
      "num_tokens": 94046464.0,
      "step": 37100
    },
    {
      "entropy": 6.124178123474121,
      "epoch": 4.282169648009233,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032990742223743544,
      "loss": 5.5954,
      "mean_token_accuracy": 0.2051453560590744,
      "num_tokens": 94058494.0,
      "step": 37105
    },
    {
      "entropy": 6.102830600738526,
      "epoch": 4.282746682054241,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003298674045104126,
      "loss": 5.6093,
      "mean_token_accuracy": 0.20323268622159957,
      "num_tokens": 94071755.0,
      "step": 37110
    },
    {
      "entropy": 6.143615055084228,
      "epoch": 4.2833237160992494,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032982738493795194,
      "loss": 5.5293,
      "mean_token_accuracy": 0.20884324014186859,
      "num_tokens": 94084888.0,
      "step": 37115
    },
    {
      "entropy": 6.062659025192261,
      "epoch": 4.283900750144259,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032978736352139935,
      "loss": 5.5911,
      "mean_token_accuracy": 0.2056392550468445,
      "num_tokens": 94098799.0,
      "step": 37120
    },
    {
      "entropy": 6.085668992996216,
      "epoch": 4.284477784189267,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003297473402621011,
      "loss": 5.5559,
      "mean_token_accuracy": 0.2036893054842949,
      "num_tokens": 94112690.0,
      "step": 37125
    },
    {
      "entropy": 6.110815191268921,
      "epoch": 4.285054818234276,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003297073151614032,
      "loss": 5.5993,
      "mean_token_accuracy": 0.20985495299100876,
      "num_tokens": 94125788.0,
      "step": 37130
    },
    {
      "entropy": 6.124050903320312,
      "epoch": 4.285631852279285,
      "grad_norm": 1.0,
      "learning_rate": 0.00032966728822065195,
      "loss": 5.5508,
      "mean_token_accuracy": 0.20489866137504578,
      "num_tokens": 94138310.0,
      "step": 37135
    },
    {
      "entropy": 6.0855231285095215,
      "epoch": 4.286208886324293,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003296272594411936,
      "loss": 5.5575,
      "mean_token_accuracy": 0.20213690102100373,
      "num_tokens": 94151475.0,
      "step": 37140
    },
    {
      "entropy": 6.08251256942749,
      "epoch": 4.286785920369302,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003295872288243745,
      "loss": 5.5885,
      "mean_token_accuracy": 0.20802915543317796,
      "num_tokens": 94164049.0,
      "step": 37145
    },
    {
      "entropy": 5.946686315536499,
      "epoch": 4.28736295441431,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032954719637154107,
      "loss": 5.4234,
      "mean_token_accuracy": 0.21805307120084763,
      "num_tokens": 94176006.0,
      "step": 37150
    },
    {
      "entropy": 6.0492674827575685,
      "epoch": 4.2879399884593195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003295071620840398,
      "loss": 5.5318,
      "mean_token_accuracy": 0.21225418746471406,
      "num_tokens": 94188855.0,
      "step": 37155
    },
    {
      "entropy": 5.979133796691895,
      "epoch": 4.288517022504328,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003294671259632172,
      "loss": 5.4774,
      "mean_token_accuracy": 0.21421956568956374,
      "num_tokens": 94202089.0,
      "step": 37160
    },
    {
      "entropy": 6.073222255706787,
      "epoch": 4.289094056549336,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032942708801041994,
      "loss": 5.5408,
      "mean_token_accuracy": 0.21121585071086885,
      "num_tokens": 94215378.0,
      "step": 37165
    },
    {
      "entropy": 6.028712320327759,
      "epoch": 4.289671090594345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032938704822699455,
      "loss": 5.4492,
      "mean_token_accuracy": 0.2216843456029892,
      "num_tokens": 94227652.0,
      "step": 37170
    },
    {
      "entropy": 6.114678382873535,
      "epoch": 4.290248124639354,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003293470066142878,
      "loss": 5.5362,
      "mean_token_accuracy": 0.20887450724840165,
      "num_tokens": 94239978.0,
      "step": 37175
    },
    {
      "entropy": 6.079232168197632,
      "epoch": 4.290825158684362,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003293069631736465,
      "loss": 5.6049,
      "mean_token_accuracy": 0.20437259674072267,
      "num_tokens": 94252363.0,
      "step": 37180
    },
    {
      "entropy": 6.061615753173828,
      "epoch": 4.291402192729371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003292669179064174,
      "loss": 5.528,
      "mean_token_accuracy": 0.20877125561237336,
      "num_tokens": 94264691.0,
      "step": 37185
    },
    {
      "entropy": 6.1360156536102295,
      "epoch": 4.29197922677438,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003292268708139476,
      "loss": 5.5541,
      "mean_token_accuracy": 0.20805031061172485,
      "num_tokens": 94276744.0,
      "step": 37190
    },
    {
      "entropy": 6.102462482452393,
      "epoch": 4.292556260819389,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003291868218975839,
      "loss": 5.563,
      "mean_token_accuracy": 0.20221893787384032,
      "num_tokens": 94289094.0,
      "step": 37195
    },
    {
      "entropy": 6.03277177810669,
      "epoch": 4.293133294864397,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003291467711586733,
      "loss": 5.4851,
      "mean_token_accuracy": 0.2170744374394417,
      "num_tokens": 94301607.0,
      "step": 37200
    },
    {
      "entropy": 6.071797323226929,
      "epoch": 4.293710328909405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032910671859856294,
      "loss": 5.5604,
      "mean_token_accuracy": 0.2095652401447296,
      "num_tokens": 94314749.0,
      "step": 37205
    },
    {
      "entropy": 6.036343193054199,
      "epoch": 4.294287362954415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003290666642186,
      "loss": 5.4826,
      "mean_token_accuracy": 0.2189536601305008,
      "num_tokens": 94327445.0,
      "step": 37210
    },
    {
      "entropy": 5.964126253128052,
      "epoch": 4.294864396999423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0003290266080201318,
      "loss": 5.395,
      "mean_token_accuracy": 0.21892424821853637,
      "num_tokens": 94340299.0,
      "step": 37215
    },
    {
      "entropy": 6.016927480697632,
      "epoch": 4.2954414310444315,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003289865500045053,
      "loss": 5.4236,
      "mean_token_accuracy": 0.21890969574451447,
      "num_tokens": 94353102.0,
      "step": 37220
    },
    {
      "entropy": 6.126763248443604,
      "epoch": 4.29601846508944,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032894649017306803,
      "loss": 5.5541,
      "mean_token_accuracy": 0.21068424880504608,
      "num_tokens": 94365325.0,
      "step": 37225
    },
    {
      "entropy": 6.10079345703125,
      "epoch": 4.296595499134449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032890642852716743,
      "loss": 5.4988,
      "mean_token_accuracy": 0.20940596610307693,
      "num_tokens": 94376978.0,
      "step": 37230
    },
    {
      "entropy": 5.94501485824585,
      "epoch": 4.297172533179458,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003288663650681508,
      "loss": 5.3932,
      "mean_token_accuracy": 0.22746245563030243,
      "num_tokens": 94390088.0,
      "step": 37235
    },
    {
      "entropy": 6.073201036453247,
      "epoch": 4.297749567224466,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032882629979736586,
      "loss": 5.5832,
      "mean_token_accuracy": 0.2104768618941307,
      "num_tokens": 94403042.0,
      "step": 37240
    },
    {
      "entropy": 6.044022703170777,
      "epoch": 4.2983266012694745,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032878623271616,
      "loss": 5.5016,
      "mean_token_accuracy": 0.2118683859705925,
      "num_tokens": 94415026.0,
      "step": 37245
    },
    {
      "entropy": 6.0431571960449215,
      "epoch": 4.298903635314484,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00032874616382588085,
      "loss": 5.5231,
      "mean_token_accuracy": 0.21760049909353257,
      "num_tokens": 94427626.0,
      "step": 37250
    },
    {
      "entropy": 6.123611545562744,
      "epoch": 4.299480669359492,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032870609312787624,
      "loss": 5.5756,
      "mean_token_accuracy": 0.20616432279348373,
      "num_tokens": 94439680.0,
      "step": 37255
    },
    {
      "entropy": 6.015365028381348,
      "epoch": 4.300057703404501,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003286660206234939,
      "loss": 5.3759,
      "mean_token_accuracy": 0.22906897366046905,
      "num_tokens": 94451479.0,
      "step": 37260
    },
    {
      "entropy": 6.056613206863403,
      "epoch": 4.30063473744951,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003286259463140816,
      "loss": 5.546,
      "mean_token_accuracy": 0.2036583825945854,
      "num_tokens": 94463944.0,
      "step": 37265
    },
    {
      "entropy": 6.016312980651856,
      "epoch": 4.301211771494518,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003285858702009871,
      "loss": 5.4819,
      "mean_token_accuracy": 0.21575609296560289,
      "num_tokens": 94477758.0,
      "step": 37270
    },
    {
      "entropy": 6.045917320251465,
      "epoch": 4.301788805539527,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032854579228555863,
      "loss": 5.4702,
      "mean_token_accuracy": 0.21647740304470062,
      "num_tokens": 94489882.0,
      "step": 37275
    },
    {
      "entropy": 6.086464548110962,
      "epoch": 4.302365839584535,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032850571256914403,
      "loss": 5.5655,
      "mean_token_accuracy": 0.20549890398979187,
      "num_tokens": 94502617.0,
      "step": 37280
    },
    {
      "entropy": 5.99830732345581,
      "epoch": 4.3029428736295445,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00032846563105309135,
      "loss": 5.5654,
      "mean_token_accuracy": 0.20827167779207229,
      "num_tokens": 94516178.0,
      "step": 37285
    },
    {
      "entropy": 6.055520153045654,
      "epoch": 4.303519907674553,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032842554773874865,
      "loss": 5.4718,
      "mean_token_accuracy": 0.21731895655393602,
      "num_tokens": 94527690.0,
      "step": 37290
    },
    {
      "entropy": 6.126521015167237,
      "epoch": 4.304096941719561,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003283854626274642,
      "loss": 5.6047,
      "mean_token_accuracy": 0.20535658448934554,
      "num_tokens": 94539464.0,
      "step": 37295
    },
    {
      "entropy": 6.153177499771118,
      "epoch": 4.30467397576457,
      "grad_norm": 0.90625,
      "learning_rate": 0.00032834537572058633,
      "loss": 5.5725,
      "mean_token_accuracy": 0.20498090088367463,
      "num_tokens": 94551898.0,
      "step": 37300
    },
    {
      "entropy": 6.116275978088379,
      "epoch": 4.305251009809579,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003283052870194632,
      "loss": 5.5894,
      "mean_token_accuracy": 0.20348978489637376,
      "num_tokens": 94564416.0,
      "step": 37305
    },
    {
      "entropy": 6.052024602890015,
      "epoch": 4.3058280438545875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032826519652544317,
      "loss": 5.5283,
      "mean_token_accuracy": 0.21288257241249084,
      "num_tokens": 94576638.0,
      "step": 37310
    },
    {
      "entropy": 6.108341884613037,
      "epoch": 4.306405077899596,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032822510423987473,
      "loss": 5.574,
      "mean_token_accuracy": 0.21100868731737138,
      "num_tokens": 94588454.0,
      "step": 37315
    },
    {
      "entropy": 6.106811285018921,
      "epoch": 4.306982111944604,
      "grad_norm": 1.0,
      "learning_rate": 0.00032818501016410636,
      "loss": 5.5495,
      "mean_token_accuracy": 0.21424442231655122,
      "num_tokens": 94601052.0,
      "step": 37320
    },
    {
      "entropy": 6.128977060317993,
      "epoch": 4.307559145989614,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003281449142994867,
      "loss": 5.6175,
      "mean_token_accuracy": 0.19523419439792633,
      "num_tokens": 94612953.0,
      "step": 37325
    },
    {
      "entropy": 6.07106499671936,
      "epoch": 4.308136180034622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032810481664736406,
      "loss": 5.5481,
      "mean_token_accuracy": 0.20889983475208282,
      "num_tokens": 94625666.0,
      "step": 37330
    },
    {
      "entropy": 6.078600883483887,
      "epoch": 4.3087132140796305,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032806471720908735,
      "loss": 5.5551,
      "mean_token_accuracy": 0.21396995186805726,
      "num_tokens": 94637575.0,
      "step": 37335
    },
    {
      "entropy": 6.02612624168396,
      "epoch": 4.30929024812464,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003280246159860053,
      "loss": 5.4859,
      "mean_token_accuracy": 0.21429236829280854,
      "num_tokens": 94650551.0,
      "step": 37340
    },
    {
      "entropy": 6.056289434432983,
      "epoch": 4.309867282169648,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003279845129794666,
      "loss": 5.5169,
      "mean_token_accuracy": 0.21121500581502914,
      "num_tokens": 94662398.0,
      "step": 37345
    },
    {
      "entropy": 6.129547834396362,
      "epoch": 4.310444316214657,
      "grad_norm": 0.921875,
      "learning_rate": 0.00032794440819082006,
      "loss": 5.5435,
      "mean_token_accuracy": 0.20888321995735168,
      "num_tokens": 94675477.0,
      "step": 37350
    },
    {
      "entropy": 6.129523658752442,
      "epoch": 4.311021350259665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003279043016214147,
      "loss": 5.573,
      "mean_token_accuracy": 0.20681553930044175,
      "num_tokens": 94688151.0,
      "step": 37355
    },
    {
      "entropy": 6.052450513839721,
      "epoch": 4.311598384304674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032786419327259945,
      "loss": 5.4959,
      "mean_token_accuracy": 0.21740474551916122,
      "num_tokens": 94700147.0,
      "step": 37360
    },
    {
      "entropy": 6.087112283706665,
      "epoch": 4.312175418349683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003278240831457234,
      "loss": 5.572,
      "mean_token_accuracy": 0.20994939059019088,
      "num_tokens": 94712809.0,
      "step": 37365
    },
    {
      "entropy": 6.102428579330445,
      "epoch": 4.312752452394691,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003277839712421354,
      "loss": 5.5566,
      "mean_token_accuracy": 0.20368150621652603,
      "num_tokens": 94725538.0,
      "step": 37370
    },
    {
      "entropy": 6.052475547790527,
      "epoch": 4.3133294864397,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032774385756318477,
      "loss": 5.4931,
      "mean_token_accuracy": 0.20910515785217285,
      "num_tokens": 94739577.0,
      "step": 37375
    },
    {
      "entropy": 6.084396171569824,
      "epoch": 4.313906520484709,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032770374211022076,
      "loss": 5.5507,
      "mean_token_accuracy": 0.20866205245256425,
      "num_tokens": 94750378.0,
      "step": 37380
    },
    {
      "entropy": 6.076275110244751,
      "epoch": 4.314483554529717,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032766362488459254,
      "loss": 5.4887,
      "mean_token_accuracy": 0.21786177009344102,
      "num_tokens": 94762587.0,
      "step": 37385
    },
    {
      "entropy": 6.073229837417602,
      "epoch": 4.315060588574726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003276235058876494,
      "loss": 5.4965,
      "mean_token_accuracy": 0.210693921148777,
      "num_tokens": 94775063.0,
      "step": 37390
    },
    {
      "entropy": 5.964566421508789,
      "epoch": 4.315637622619734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032758338512074085,
      "loss": 5.4678,
      "mean_token_accuracy": 0.21732085794210435,
      "num_tokens": 94786743.0,
      "step": 37395
    },
    {
      "entropy": 6.070464944839477,
      "epoch": 4.3162146566647435,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003275432625852163,
      "loss": 5.5527,
      "mean_token_accuracy": 0.20926856398582458,
      "num_tokens": 94798811.0,
      "step": 37400
    },
    {
      "entropy": 5.931460523605347,
      "epoch": 4.316791690709752,
      "grad_norm": 1.0,
      "learning_rate": 0.0003275031382824251,
      "loss": 5.449,
      "mean_token_accuracy": 0.2196829542517662,
      "num_tokens": 94811908.0,
      "step": 37405
    },
    {
      "entropy": 6.103973007202148,
      "epoch": 4.31736872475476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003274630122137169,
      "loss": 5.5701,
      "mean_token_accuracy": 0.2101457953453064,
      "num_tokens": 94825533.0,
      "step": 37410
    },
    {
      "entropy": 6.026005268096924,
      "epoch": 4.31794575879977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003274228843804415,
      "loss": 5.5175,
      "mean_token_accuracy": 0.2105269104242325,
      "num_tokens": 94837424.0,
      "step": 37415
    },
    {
      "entropy": 6.0348903179168705,
      "epoch": 4.318522792844778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032738275478394837,
      "loss": 5.4989,
      "mean_token_accuracy": 0.21298986226320266,
      "num_tokens": 94851058.0,
      "step": 37420
    },
    {
      "entropy": 6.1331987380981445,
      "epoch": 4.319099826889786,
      "grad_norm": 1.0,
      "learning_rate": 0.00032734262342558725,
      "loss": 5.609,
      "mean_token_accuracy": 0.20900993943214416,
      "num_tokens": 94864389.0,
      "step": 37425
    },
    {
      "entropy": 6.045484399795532,
      "epoch": 4.319676860934795,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000327302490306708,
      "loss": 5.4505,
      "mean_token_accuracy": 0.2170125037431717,
      "num_tokens": 94876384.0,
      "step": 37430
    },
    {
      "entropy": 6.065686511993408,
      "epoch": 4.320253894979804,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032726235542866045,
      "loss": 5.5227,
      "mean_token_accuracy": 0.21027269959449768,
      "num_tokens": 94888238.0,
      "step": 37435
    },
    {
      "entropy": 6.0775025367736815,
      "epoch": 4.320830929024813,
      "grad_norm": 0.921875,
      "learning_rate": 0.00032722221879279454,
      "loss": 5.5496,
      "mean_token_accuracy": 0.21131379008293152,
      "num_tokens": 94901263.0,
      "step": 37440
    },
    {
      "entropy": 6.073624801635742,
      "epoch": 4.321407963069821,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032718208040046035,
      "loss": 5.5098,
      "mean_token_accuracy": 0.21032853722572326,
      "num_tokens": 94914794.0,
      "step": 37445
    },
    {
      "entropy": 6.043347644805908,
      "epoch": 4.321984997114829,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003271419402530077,
      "loss": 5.5043,
      "mean_token_accuracy": 0.20913120806217195,
      "num_tokens": 94927636.0,
      "step": 37450
    },
    {
      "entropy": 6.103030443191528,
      "epoch": 4.322562031159839,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00032710179835178675,
      "loss": 5.5876,
      "mean_token_accuracy": 0.20516337752342223,
      "num_tokens": 94941329.0,
      "step": 37455
    },
    {
      "entropy": 6.13333215713501,
      "epoch": 4.323139065204847,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032706165469814787,
      "loss": 5.6193,
      "mean_token_accuracy": 0.20392811000347139,
      "num_tokens": 94954275.0,
      "step": 37460
    },
    {
      "entropy": 6.04155387878418,
      "epoch": 4.323716099249856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00032702150929344096,
      "loss": 5.5158,
      "mean_token_accuracy": 0.21225043386220932,
      "num_tokens": 94966509.0,
      "step": 37465
    },
    {
      "entropy": 6.032373428344727,
      "epoch": 4.324293133294864,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003269813621390164,
      "loss": 5.4367,
      "mean_token_accuracy": 0.21376957297325133,
      "num_tokens": 94978818.0,
      "step": 37470
    },
    {
      "entropy": 6.028779745101929,
      "epoch": 4.324870167339873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032694121323622455,
      "loss": 5.4517,
      "mean_token_accuracy": 0.21804199367761612,
      "num_tokens": 94991577.0,
      "step": 37475
    },
    {
      "entropy": 6.089264726638794,
      "epoch": 4.325447201384882,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00032690106258641586,
      "loss": 5.6616,
      "mean_token_accuracy": 0.20593415647745134,
      "num_tokens": 95004242.0,
      "step": 37480
    },
    {
      "entropy": 6.0682073593139645,
      "epoch": 4.32602423542989,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032686091019094074,
      "loss": 5.4848,
      "mean_token_accuracy": 0.21447964906692504,
      "num_tokens": 95016673.0,
      "step": 37485
    },
    {
      "entropy": 6.104940938949585,
      "epoch": 4.326601269474899,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003268207560511496,
      "loss": 5.5943,
      "mean_token_accuracy": 0.2044106036424637,
      "num_tokens": 95028453.0,
      "step": 37490
    },
    {
      "entropy": 5.995596551895142,
      "epoch": 4.327178303519908,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032678060016839305,
      "loss": 5.4838,
      "mean_token_accuracy": 0.22168498933315278,
      "num_tokens": 95042668.0,
      "step": 37495
    },
    {
      "entropy": 6.06745228767395,
      "epoch": 4.327755337564916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003267404425440217,
      "loss": 5.4872,
      "mean_token_accuracy": 0.21586869657039642,
      "num_tokens": 95053495.0,
      "step": 37500
    },
    {
      "entropy": 6.064965915679932,
      "epoch": 4.328332371609925,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032670028317938643,
      "loss": 5.5101,
      "mean_token_accuracy": 0.21433160305023194,
      "num_tokens": 95065992.0,
      "step": 37505
    },
    {
      "entropy": 6.019883632659912,
      "epoch": 4.328909405654934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003266601220758377,
      "loss": 5.5106,
      "mean_token_accuracy": 0.2149416223168373,
      "num_tokens": 95078884.0,
      "step": 37510
    },
    {
      "entropy": 6.074158620834351,
      "epoch": 4.329486439699942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003266199592347264,
      "loss": 5.5069,
      "mean_token_accuracy": 0.2107894629240036,
      "num_tokens": 95090487.0,
      "step": 37515
    },
    {
      "entropy": 6.084133529663086,
      "epoch": 4.330063473744951,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003265797946574035,
      "loss": 5.5994,
      "mean_token_accuracy": 0.20072934329509734,
      "num_tokens": 95102842.0,
      "step": 37520
    },
    {
      "entropy": 6.150358629226685,
      "epoch": 4.330640507789959,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003265396283452198,
      "loss": 5.5881,
      "mean_token_accuracy": 0.20757602751255036,
      "num_tokens": 95114523.0,
      "step": 37525
    },
    {
      "entropy": 6.012509536743164,
      "epoch": 4.3312175418349685,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003264994602995263,
      "loss": 5.4371,
      "mean_token_accuracy": 0.22838186472654343,
      "num_tokens": 95127370.0,
      "step": 37530
    },
    {
      "entropy": 6.0408416271209715,
      "epoch": 4.331794575879977,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000326459290521674,
      "loss": 5.5972,
      "mean_token_accuracy": 0.2083314284682274,
      "num_tokens": 95139543.0,
      "step": 37535
    },
    {
      "entropy": 6.137928915023804,
      "epoch": 4.332371609924985,
      "grad_norm": 0.921875,
      "learning_rate": 0.00032641911901301406,
      "loss": 5.5842,
      "mean_token_accuracy": 0.20809554010629655,
      "num_tokens": 95151837.0,
      "step": 37540
    },
    {
      "entropy": 6.0659466743469235,
      "epoch": 4.332948643969994,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003263789457748976,
      "loss": 5.4716,
      "mean_token_accuracy": 0.21733748465776442,
      "num_tokens": 95165188.0,
      "step": 37545
    },
    {
      "entropy": 6.115704154968261,
      "epoch": 4.333525678015003,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032633877080867575,
      "loss": 5.5571,
      "mean_token_accuracy": 0.2118029475212097,
      "num_tokens": 95178311.0,
      "step": 37550
    },
    {
      "entropy": 6.044954442977906,
      "epoch": 4.3341027120600115,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032629859411569987,
      "loss": 5.4755,
      "mean_token_accuracy": 0.2190798595547676,
      "num_tokens": 95191078.0,
      "step": 37555
    },
    {
      "entropy": 6.052467632293701,
      "epoch": 4.33467974610502,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003262584156973213,
      "loss": 5.5035,
      "mean_token_accuracy": 0.21078290045261383,
      "num_tokens": 95205222.0,
      "step": 37560
    },
    {
      "entropy": 5.990278816223144,
      "epoch": 4.335256780150029,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003262182355548913,
      "loss": 5.4387,
      "mean_token_accuracy": 0.2187832310795784,
      "num_tokens": 95218754.0,
      "step": 37565
    },
    {
      "entropy": 6.042447710037232,
      "epoch": 4.335833814195038,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003261780536897615,
      "loss": 5.5031,
      "mean_token_accuracy": 0.2175952523946762,
      "num_tokens": 95232786.0,
      "step": 37570
    },
    {
      "entropy": 6.132240724563599,
      "epoch": 4.336410848240046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032613787010328326,
      "loss": 5.565,
      "mean_token_accuracy": 0.20693478137254714,
      "num_tokens": 95245100.0,
      "step": 37575
    },
    {
      "entropy": 6.020465230941772,
      "epoch": 4.3369878822850545,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003260976847968081,
      "loss": 5.5096,
      "mean_token_accuracy": 0.21758360415697098,
      "num_tokens": 95257161.0,
      "step": 37580
    },
    {
      "entropy": 6.042992639541626,
      "epoch": 4.337564916330064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032605749777168773,
      "loss": 5.4845,
      "mean_token_accuracy": 0.21659012734889985,
      "num_tokens": 95269453.0,
      "step": 37585
    },
    {
      "entropy": 6.132225418090821,
      "epoch": 4.338141950375072,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003260173090292738,
      "loss": 5.5932,
      "mean_token_accuracy": 0.20707600563764572,
      "num_tokens": 95282496.0,
      "step": 37590
    },
    {
      "entropy": 6.075981950759887,
      "epoch": 4.338718984420081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000325977118570918,
      "loss": 5.5628,
      "mean_token_accuracy": 0.2111838400363922,
      "num_tokens": 95295747.0,
      "step": 37595
    },
    {
      "entropy": 6.119698333740234,
      "epoch": 4.339296018465089,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032593692639797215,
      "loss": 5.5387,
      "mean_token_accuracy": 0.2100005567073822,
      "num_tokens": 95307568.0,
      "step": 37600
    },
    {
      "entropy": 5.99921875,
      "epoch": 4.339873052510098,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0003258967325117881,
      "loss": 5.4744,
      "mean_token_accuracy": 0.2237042099237442,
      "num_tokens": 95320988.0,
      "step": 37605
    },
    {
      "entropy": 6.082865810394287,
      "epoch": 4.340450086555107,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032585653691371777,
      "loss": 5.5275,
      "mean_token_accuracy": 0.21049854159355164,
      "num_tokens": 95334606.0,
      "step": 37610
    },
    {
      "entropy": 6.121924543380738,
      "epoch": 4.341027120600115,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032581633960511295,
      "loss": 5.5383,
      "mean_token_accuracy": 0.20518902093172073,
      "num_tokens": 95346421.0,
      "step": 37615
    },
    {
      "entropy": 6.167167091369629,
      "epoch": 4.341604154645124,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003257761405873259,
      "loss": 5.5941,
      "mean_token_accuracy": 0.20753082036972045,
      "num_tokens": 95360052.0,
      "step": 37620
    },
    {
      "entropy": 6.095099592208863,
      "epoch": 4.342181188690133,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032573593986170854,
      "loss": 5.5704,
      "mean_token_accuracy": 0.2097340151667595,
      "num_tokens": 95373290.0,
      "step": 37625
    },
    {
      "entropy": 6.0287370681762695,
      "epoch": 4.342758222735141,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000325695737429613,
      "loss": 5.4569,
      "mean_token_accuracy": 0.2172936052083969,
      "num_tokens": 95385278.0,
      "step": 37630
    },
    {
      "entropy": 6.121437311172485,
      "epoch": 4.34333525678015,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032565553329239164,
      "loss": 5.574,
      "mean_token_accuracy": 0.21242649257183074,
      "num_tokens": 95398012.0,
      "step": 37635
    },
    {
      "entropy": 6.110891342163086,
      "epoch": 4.343912290825159,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003256153274513964,
      "loss": 5.5289,
      "mean_token_accuracy": 0.21502980589866638,
      "num_tokens": 95411172.0,
      "step": 37640
    },
    {
      "entropy": 5.982211256027222,
      "epoch": 4.3444893248701675,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003255751199079799,
      "loss": 5.4508,
      "mean_token_accuracy": 0.22060692608356475,
      "num_tokens": 95423025.0,
      "step": 37645
    },
    {
      "entropy": 6.051953506469727,
      "epoch": 4.345066358915176,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003255349106634943,
      "loss": 5.5126,
      "mean_token_accuracy": 0.21430380195379256,
      "num_tokens": 95435111.0,
      "step": 37650
    },
    {
      "entropy": 6.029353332519531,
      "epoch": 4.345643392960184,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003254946997192922,
      "loss": 5.4873,
      "mean_token_accuracy": 0.21657829582691193,
      "num_tokens": 95448047.0,
      "step": 37655
    },
    {
      "entropy": 6.0411865234375,
      "epoch": 4.346220427005194,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003254544870767259,
      "loss": 5.4748,
      "mean_token_accuracy": 0.2143942892551422,
      "num_tokens": 95459806.0,
      "step": 37660
    },
    {
      "entropy": 6.038434934616089,
      "epoch": 4.346797461050202,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003254142727371479,
      "loss": 5.5186,
      "mean_token_accuracy": 0.20796762257814408,
      "num_tokens": 95471759.0,
      "step": 37665
    },
    {
      "entropy": 6.034385538101196,
      "epoch": 4.3473744950952105,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003253740567019109,
      "loss": 5.5043,
      "mean_token_accuracy": 0.21245481818914413,
      "num_tokens": 95483371.0,
      "step": 37670
    },
    {
      "entropy": 6.06047887802124,
      "epoch": 4.347951529140219,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00032533383897236757,
      "loss": 5.5773,
      "mean_token_accuracy": 0.21378377974033355,
      "num_tokens": 95496348.0,
      "step": 37675
    },
    {
      "entropy": 6.057560873031616,
      "epoch": 4.348528563185228,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003252936195498706,
      "loss": 5.4769,
      "mean_token_accuracy": 0.2145786166191101,
      "num_tokens": 95508349.0,
      "step": 37680
    },
    {
      "entropy": 5.965884494781494,
      "epoch": 4.349105597230237,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032525339843577267,
      "loss": 5.3978,
      "mean_token_accuracy": 0.2242712140083313,
      "num_tokens": 95520389.0,
      "step": 37685
    },
    {
      "entropy": 6.0365142822265625,
      "epoch": 4.349682631275245,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003252131756314266,
      "loss": 5.5416,
      "mean_token_accuracy": 0.20944399535655975,
      "num_tokens": 95533962.0,
      "step": 37690
    },
    {
      "entropy": 5.943994951248169,
      "epoch": 4.350259665320253,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003251729511381854,
      "loss": 5.4332,
      "mean_token_accuracy": 0.22141103595495223,
      "num_tokens": 95545830.0,
      "step": 37695
    },
    {
      "entropy": 6.007016611099243,
      "epoch": 4.350836699365263,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032513272495740195,
      "loss": 5.4606,
      "mean_token_accuracy": 0.21607156842947006,
      "num_tokens": 95558988.0,
      "step": 37700
    },
    {
      "entropy": 6.157297563552857,
      "epoch": 4.351413733410271,
      "grad_norm": 1.4375,
      "learning_rate": 0.00032509249709042906,
      "loss": 5.6447,
      "mean_token_accuracy": 0.197378209233284,
      "num_tokens": 95572379.0,
      "step": 37705
    },
    {
      "entropy": 6.088064861297608,
      "epoch": 4.35199076745528,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032505226753862,
      "loss": 5.5042,
      "mean_token_accuracy": 0.21784281432628633,
      "num_tokens": 95586219.0,
      "step": 37710
    },
    {
      "entropy": 6.17486252784729,
      "epoch": 4.352567801500289,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032501203630332776,
      "loss": 5.6351,
      "mean_token_accuracy": 0.20502230226993562,
      "num_tokens": 95598604.0,
      "step": 37715
    },
    {
      "entropy": 6.1081318855285645,
      "epoch": 4.353144835545297,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032497180338590556,
      "loss": 5.627,
      "mean_token_accuracy": 0.20080945938825606,
      "num_tokens": 95611995.0,
      "step": 37720
    },
    {
      "entropy": 6.045156335830688,
      "epoch": 4.353721869590306,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003249315687877065,
      "loss": 5.5462,
      "mean_token_accuracy": 0.21181490570306777,
      "num_tokens": 95624719.0,
      "step": 37725
    },
    {
      "entropy": 6.0576849460601805,
      "epoch": 4.354298903635314,
      "grad_norm": 1.015625,
      "learning_rate": 0.000324891332510084,
      "loss": 5.4982,
      "mean_token_accuracy": 0.21005320698022842,
      "num_tokens": 95637982.0,
      "step": 37730
    },
    {
      "entropy": 6.117294263839722,
      "epoch": 4.354875937680323,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032485109455439125,
      "loss": 5.6164,
      "mean_token_accuracy": 0.20527080744504927,
      "num_tokens": 95650708.0,
      "step": 37735
    },
    {
      "entropy": 6.148378324508667,
      "epoch": 4.355452971725332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032481085492198176,
      "loss": 5.5758,
      "mean_token_accuracy": 0.21194101870059967,
      "num_tokens": 95663118.0,
      "step": 37740
    },
    {
      "entropy": 6.10278639793396,
      "epoch": 4.35603000577034,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003247706136142088,
      "loss": 5.5961,
      "mean_token_accuracy": 0.20882778018712997,
      "num_tokens": 95675736.0,
      "step": 37745
    },
    {
      "entropy": 6.0739710330963135,
      "epoch": 4.356607039815349,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003247303706324261,
      "loss": 5.5504,
      "mean_token_accuracy": 0.19995128065347673,
      "num_tokens": 95688194.0,
      "step": 37750
    },
    {
      "entropy": 6.119931745529175,
      "epoch": 4.357184073860358,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00032469012597798703,
      "loss": 5.5444,
      "mean_token_accuracy": 0.21032788455486298,
      "num_tokens": 95700524.0,
      "step": 37755
    },
    {
      "entropy": 6.092657804489136,
      "epoch": 4.357761107905366,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003246498796522453,
      "loss": 5.566,
      "mean_token_accuracy": 0.20645192563533782,
      "num_tokens": 95712510.0,
      "step": 37760
    },
    {
      "entropy": 6.0990458011627195,
      "epoch": 4.358338141950375,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032460963165655444,
      "loss": 5.5202,
      "mean_token_accuracy": 0.21628334522247314,
      "num_tokens": 95724728.0,
      "step": 37765
    },
    {
      "entropy": 6.068203449249268,
      "epoch": 4.358915175995384,
      "grad_norm": 1.171875,
      "learning_rate": 0.00032456938199226825,
      "loss": 5.5288,
      "mean_token_accuracy": 0.2114872395992279,
      "num_tokens": 95736212.0,
      "step": 37770
    },
    {
      "entropy": 5.9820150375366214,
      "epoch": 4.359492210040393,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032452913066074063,
      "loss": 5.4487,
      "mean_token_accuracy": 0.21543620824813842,
      "num_tokens": 95747789.0,
      "step": 37775
    },
    {
      "entropy": 6.053436279296875,
      "epoch": 4.360069244085401,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032448887766332526,
      "loss": 5.4542,
      "mean_token_accuracy": 0.2163073256611824,
      "num_tokens": 95759628.0,
      "step": 37780
    },
    {
      "entropy": 6.036184358596802,
      "epoch": 4.360646278130409,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000324448623001376,
      "loss": 5.4966,
      "mean_token_accuracy": 0.21180918216705322,
      "num_tokens": 95773164.0,
      "step": 37785
    },
    {
      "entropy": 6.04491958618164,
      "epoch": 4.361223312175419,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032440836667624697,
      "loss": 5.5101,
      "mean_token_accuracy": 0.2127455785870552,
      "num_tokens": 95786315.0,
      "step": 37790
    },
    {
      "entropy": 5.965170669555664,
      "epoch": 4.361800346220427,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032436810868929206,
      "loss": 5.393,
      "mean_token_accuracy": 0.2238956719636917,
      "num_tokens": 95798839.0,
      "step": 37795
    },
    {
      "entropy": 5.98937840461731,
      "epoch": 4.3623773802654355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003243278490418653,
      "loss": 5.3851,
      "mean_token_accuracy": 0.22631475627422332,
      "num_tokens": 95812980.0,
      "step": 37800
    },
    {
      "entropy": 5.96692304611206,
      "epoch": 4.362954414310444,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032428758773532084,
      "loss": 5.4303,
      "mean_token_accuracy": 0.22445688396692276,
      "num_tokens": 95825112.0,
      "step": 37805
    },
    {
      "entropy": 6.096612215042114,
      "epoch": 4.363531448355453,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003242473247710128,
      "loss": 5.5401,
      "mean_token_accuracy": 0.20766531974077224,
      "num_tokens": 95837395.0,
      "step": 37810
    },
    {
      "entropy": 6.103021430969238,
      "epoch": 4.364108482400462,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003242070601502956,
      "loss": 5.5651,
      "mean_token_accuracy": 0.2003348335623741,
      "num_tokens": 95850351.0,
      "step": 37815
    },
    {
      "entropy": 6.139268398284912,
      "epoch": 4.36468551644547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003241667938745232,
      "loss": 5.573,
      "mean_token_accuracy": 0.20128797441720964,
      "num_tokens": 95863582.0,
      "step": 37820
    },
    {
      "entropy": 6.099933528900147,
      "epoch": 4.365262550490479,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032412652594505025,
      "loss": 5.54,
      "mean_token_accuracy": 0.21333780586719514,
      "num_tokens": 95876496.0,
      "step": 37825
    },
    {
      "entropy": 6.059851789474488,
      "epoch": 4.365839584535488,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003240862563632309,
      "loss": 5.4486,
      "mean_token_accuracy": 0.21509270071983339,
      "num_tokens": 95889921.0,
      "step": 37830
    },
    {
      "entropy": 6.044889259338379,
      "epoch": 4.366416618580496,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032404598513041975,
      "loss": 5.4398,
      "mean_token_accuracy": 0.22177815437316895,
      "num_tokens": 95901783.0,
      "step": 37835
    },
    {
      "entropy": 6.010998487472534,
      "epoch": 4.366993652625505,
      "grad_norm": 0.84375,
      "learning_rate": 0.0003240057122479713,
      "loss": 5.4839,
      "mean_token_accuracy": 0.22154007703065873,
      "num_tokens": 95915640.0,
      "step": 37840
    },
    {
      "entropy": 6.075091457366943,
      "epoch": 4.367570686670514,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032396543771724,
      "loss": 5.5393,
      "mean_token_accuracy": 0.21626498252153398,
      "num_tokens": 95928851.0,
      "step": 37845
    },
    {
      "entropy": 6.083856821060181,
      "epoch": 4.368147720715522,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003239251615395805,
      "loss": 5.4851,
      "mean_token_accuracy": 0.22281526774168015,
      "num_tokens": 95941497.0,
      "step": 37850
    },
    {
      "entropy": 6.082077121734619,
      "epoch": 4.368724754760531,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003238848837163476,
      "loss": 5.4706,
      "mean_token_accuracy": 0.21153527200222016,
      "num_tokens": 95952389.0,
      "step": 37855
    },
    {
      "entropy": 5.9567546367645265,
      "epoch": 4.369301788805539,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003238446042488958,
      "loss": 5.4521,
      "mean_token_accuracy": 0.2161433592438698,
      "num_tokens": 95965043.0,
      "step": 37860
    },
    {
      "entropy": 6.032352590560913,
      "epoch": 4.3698788228505485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003238043231385801,
      "loss": 5.5346,
      "mean_token_accuracy": 0.22179989218711854,
      "num_tokens": 95979478.0,
      "step": 37865
    },
    {
      "entropy": 6.053231048583984,
      "epoch": 4.370455856895557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003237640403867551,
      "loss": 5.5172,
      "mean_token_accuracy": 0.21583304256200792,
      "num_tokens": 95992502.0,
      "step": 37870
    },
    {
      "entropy": 6.107642269134521,
      "epoch": 4.371032890940565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032372375599477594,
      "loss": 5.5126,
      "mean_token_accuracy": 0.20702233612537385,
      "num_tokens": 96006433.0,
      "step": 37875
    },
    {
      "entropy": 6.076223134994507,
      "epoch": 4.371609924985574,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003236834699639974,
      "loss": 5.5239,
      "mean_token_accuracy": 0.21021799594163895,
      "num_tokens": 96019525.0,
      "step": 37880
    },
    {
      "entropy": 6.031029224395752,
      "epoch": 4.372186959030583,
      "grad_norm": 1.359375,
      "learning_rate": 0.00032364318229577454,
      "loss": 5.5524,
      "mean_token_accuracy": 0.20655781179666519,
      "num_tokens": 96031342.0,
      "step": 37885
    },
    {
      "entropy": 6.115832948684693,
      "epoch": 4.3727639930755915,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003236028929914624,
      "loss": 5.5427,
      "mean_token_accuracy": 0.20653196424245834,
      "num_tokens": 96043370.0,
      "step": 37890
    },
    {
      "entropy": 6.024349355697632,
      "epoch": 4.3733410271206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003235626020524161,
      "loss": 5.4632,
      "mean_token_accuracy": 0.22202130407094955,
      "num_tokens": 96057617.0,
      "step": 37895
    },
    {
      "entropy": 6.036150121688843,
      "epoch": 4.373918061165609,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032352230947999087,
      "loss": 5.4563,
      "mean_token_accuracy": 0.21486581563949586,
      "num_tokens": 96069579.0,
      "step": 37900
    },
    {
      "entropy": 5.970242595672607,
      "epoch": 4.374495095210618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032348201527554184,
      "loss": 5.4826,
      "mean_token_accuracy": 0.22254715412855147,
      "num_tokens": 96083167.0,
      "step": 37905
    },
    {
      "entropy": 6.062537431716919,
      "epoch": 4.375072129255626,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032344171944042417,
      "loss": 5.5705,
      "mean_token_accuracy": 0.20287540256977082,
      "num_tokens": 96095697.0,
      "step": 37910
    },
    {
      "entropy": 6.047841596603393,
      "epoch": 4.3756491633006345,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00032340142197599344,
      "loss": 5.558,
      "mean_token_accuracy": 0.21379594653844833,
      "num_tokens": 96110000.0,
      "step": 37915
    },
    {
      "entropy": 5.990476369857788,
      "epoch": 4.376226197345644,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003233611228836049,
      "loss": 5.4135,
      "mean_token_accuracy": 0.21861625760793685,
      "num_tokens": 96122187.0,
      "step": 37920
    },
    {
      "entropy": 6.128258514404297,
      "epoch": 4.376803231390652,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032332082216461405,
      "loss": 5.4945,
      "mean_token_accuracy": 0.21603007912635802,
      "num_tokens": 96134314.0,
      "step": 37925
    },
    {
      "entropy": 5.960884237289429,
      "epoch": 4.377380265435661,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003232805198203763,
      "loss": 5.4882,
      "mean_token_accuracy": 0.21946474611759187,
      "num_tokens": 96146497.0,
      "step": 37930
    },
    {
      "entropy": 6.106948614120483,
      "epoch": 4.377957299480669,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032324021585224735,
      "loss": 5.5557,
      "mean_token_accuracy": 0.20852634757757188,
      "num_tokens": 96160291.0,
      "step": 37935
    },
    {
      "entropy": 6.15196795463562,
      "epoch": 4.378534333525678,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00032319991026158256,
      "loss": 5.6604,
      "mean_token_accuracy": 0.20207667350769043,
      "num_tokens": 96173240.0,
      "step": 37940
    },
    {
      "entropy": 6.164811038970948,
      "epoch": 4.379111367570687,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003231596030497378,
      "loss": 5.6154,
      "mean_token_accuracy": 0.208014814555645,
      "num_tokens": 96184313.0,
      "step": 37945
    },
    {
      "entropy": 6.102305793762207,
      "epoch": 4.379688401615695,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032311929421806866,
      "loss": 5.5079,
      "mean_token_accuracy": 0.21211728304624558,
      "num_tokens": 96195953.0,
      "step": 37950
    },
    {
      "entropy": 6.138415002822876,
      "epoch": 4.380265435660704,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000323078983767931,
      "loss": 5.6069,
      "mean_token_accuracy": 0.20362795889377594,
      "num_tokens": 96207956.0,
      "step": 37955
    },
    {
      "entropy": 6.0936614036560055,
      "epoch": 4.380842469705713,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00032303867170068064,
      "loss": 5.5568,
      "mean_token_accuracy": 0.2089806854724884,
      "num_tokens": 96220017.0,
      "step": 37960
    },
    {
      "entropy": 6.160920810699463,
      "epoch": 4.381419503750721,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003229983580176734,
      "loss": 5.6256,
      "mean_token_accuracy": 0.20560528337955475,
      "num_tokens": 96232549.0,
      "step": 37965
    },
    {
      "entropy": 6.018291378021241,
      "epoch": 4.38199653779573,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032295804272026524,
      "loss": 5.4568,
      "mean_token_accuracy": 0.21562202125787736,
      "num_tokens": 96244778.0,
      "step": 37970
    },
    {
      "entropy": 6.039153575897217,
      "epoch": 4.382573571840739,
      "grad_norm": 0.890625,
      "learning_rate": 0.00032291772580981203,
      "loss": 5.5494,
      "mean_token_accuracy": 0.21221984922885895,
      "num_tokens": 96258087.0,
      "step": 37975
    },
    {
      "entropy": 6.224298477172852,
      "epoch": 4.3831506058857475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003228774072876701,
      "loss": 5.6551,
      "mean_token_accuracy": 0.20120372623205185,
      "num_tokens": 96271101.0,
      "step": 37980
    },
    {
      "entropy": 6.091609954833984,
      "epoch": 4.383727639930756,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032283708715519525,
      "loss": 5.5934,
      "mean_token_accuracy": 0.2053753286600113,
      "num_tokens": 96282734.0,
      "step": 37985
    },
    {
      "entropy": 6.0761994361877445,
      "epoch": 4.384304673975764,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032279676541374376,
      "loss": 5.5731,
      "mean_token_accuracy": 0.20398081392049788,
      "num_tokens": 96295955.0,
      "step": 37990
    },
    {
      "entropy": 6.138612127304077,
      "epoch": 4.384881708020774,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032275644206467176,
      "loss": 5.5531,
      "mean_token_accuracy": 0.21061747372150422,
      "num_tokens": 96308618.0,
      "step": 37995
    },
    {
      "entropy": 6.090914058685303,
      "epoch": 4.385458742065782,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003227161171093356,
      "loss": 5.5878,
      "mean_token_accuracy": 0.21030182540416717,
      "num_tokens": 96321933.0,
      "step": 38000
    },
    {
      "entropy": 6.0297308444976805,
      "epoch": 4.38603577611079,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032267579054909154,
      "loss": 5.5512,
      "mean_token_accuracy": 0.21415201127529143,
      "num_tokens": 96333970.0,
      "step": 38005
    },
    {
      "entropy": 6.014867639541626,
      "epoch": 4.386612810155799,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032263546238529595,
      "loss": 5.4573,
      "mean_token_accuracy": 0.21842311024665834,
      "num_tokens": 96346227.0,
      "step": 38010
    },
    {
      "entropy": 6.102007913589477,
      "epoch": 4.387189844200808,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032259513261930525,
      "loss": 5.5856,
      "mean_token_accuracy": 0.20976892113685608,
      "num_tokens": 96358585.0,
      "step": 38015
    },
    {
      "entropy": 6.118472290039063,
      "epoch": 4.387766878245817,
      "grad_norm": 1.0,
      "learning_rate": 0.0003225548012524759,
      "loss": 5.5585,
      "mean_token_accuracy": 0.21225176006555557,
      "num_tokens": 96372693.0,
      "step": 38020
    },
    {
      "entropy": 6.0720155239105225,
      "epoch": 4.388343912290825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003225144682861644,
      "loss": 5.4539,
      "mean_token_accuracy": 0.20865799635648727,
      "num_tokens": 96384001.0,
      "step": 38025
    },
    {
      "entropy": 6.080781078338623,
      "epoch": 4.388920946335833,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003224741337217273,
      "loss": 5.5239,
      "mean_token_accuracy": 0.212342569231987,
      "num_tokens": 96396281.0,
      "step": 38030
    },
    {
      "entropy": 6.054772806167603,
      "epoch": 4.389497980380843,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032243379756052143,
      "loss": 5.508,
      "mean_token_accuracy": 0.21675640046596528,
      "num_tokens": 96409820.0,
      "step": 38035
    },
    {
      "entropy": 6.037151956558228,
      "epoch": 4.390075014425851,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032239345980390326,
      "loss": 5.5013,
      "mean_token_accuracy": 0.2079442709684372,
      "num_tokens": 96423038.0,
      "step": 38040
    },
    {
      "entropy": 6.022721242904663,
      "epoch": 4.3906520484708595,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003223531204532297,
      "loss": 5.4769,
      "mean_token_accuracy": 0.21164907813072203,
      "num_tokens": 96435395.0,
      "step": 38045
    },
    {
      "entropy": 6.020797252655029,
      "epoch": 4.391229082515869,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003223127795098574,
      "loss": 5.5019,
      "mean_token_accuracy": 0.20482718646526338,
      "num_tokens": 96449094.0,
      "step": 38050
    },
    {
      "entropy": 6.095019388198852,
      "epoch": 4.391806116560877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003222724369751432,
      "loss": 5.4689,
      "mean_token_accuracy": 0.22095294296741486,
      "num_tokens": 96461754.0,
      "step": 38055
    },
    {
      "entropy": 5.9862569808959964,
      "epoch": 4.392383150605886,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003222320928504441,
      "loss": 5.4059,
      "mean_token_accuracy": 0.22218168526887894,
      "num_tokens": 96475510.0,
      "step": 38060
    },
    {
      "entropy": 6.0490086555480955,
      "epoch": 4.392960184650894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032219174713711703,
      "loss": 5.5129,
      "mean_token_accuracy": 0.2090813234448433,
      "num_tokens": 96488056.0,
      "step": 38065
    },
    {
      "entropy": 6.173478651046753,
      "epoch": 4.393537218695903,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032215139983651905,
      "loss": 5.6758,
      "mean_token_accuracy": 0.20194595605134963,
      "num_tokens": 96501406.0,
      "step": 38070
    },
    {
      "entropy": 6.115729808807373,
      "epoch": 4.394114252740912,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003221110509500071,
      "loss": 5.5605,
      "mean_token_accuracy": 0.20267103463411332,
      "num_tokens": 96513715.0,
      "step": 38075
    },
    {
      "entropy": 6.104687643051148,
      "epoch": 4.39469128678592,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032207070047893835,
      "loss": 5.5955,
      "mean_token_accuracy": 0.21220980882644652,
      "num_tokens": 96525819.0,
      "step": 38080
    },
    {
      "entropy": 6.0767608165740965,
      "epoch": 4.395268320830929,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032203034842467,
      "loss": 5.4925,
      "mean_token_accuracy": 0.22021742165088654,
      "num_tokens": 96538329.0,
      "step": 38085
    },
    {
      "entropy": 6.052501964569092,
      "epoch": 4.395845354875938,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032198999478855915,
      "loss": 5.5037,
      "mean_token_accuracy": 0.21344990581274031,
      "num_tokens": 96552037.0,
      "step": 38090
    },
    {
      "entropy": 6.113233661651611,
      "epoch": 4.396422388920946,
      "grad_norm": 1.0,
      "learning_rate": 0.0003219496395719633,
      "loss": 5.518,
      "mean_token_accuracy": 0.20924248099327086,
      "num_tokens": 96564058.0,
      "step": 38095
    },
    {
      "entropy": 6.057490348815918,
      "epoch": 4.396999422965955,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003219092827762395,
      "loss": 5.44,
      "mean_token_accuracy": 0.22290195226669313,
      "num_tokens": 96576676.0,
      "step": 38100
    },
    {
      "entropy": 6.118477058410645,
      "epoch": 4.397576457010963,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003218689244027453,
      "loss": 5.5708,
      "mean_token_accuracy": 0.209315487742424,
      "num_tokens": 96589815.0,
      "step": 38105
    },
    {
      "entropy": 6.0406107902526855,
      "epoch": 4.3981534910559725,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003218285644528383,
      "loss": 5.5138,
      "mean_token_accuracy": 0.2156430035829544,
      "num_tokens": 96601880.0,
      "step": 38110
    },
    {
      "entropy": 6.046216106414795,
      "epoch": 4.398730525100981,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032178820292787566,
      "loss": 5.4971,
      "mean_token_accuracy": 0.21180468052625656,
      "num_tokens": 96615180.0,
      "step": 38115
    },
    {
      "entropy": 5.99285569190979,
      "epoch": 4.399307559145989,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032174783982921495,
      "loss": 5.435,
      "mean_token_accuracy": 0.2249556228518486,
      "num_tokens": 96627236.0,
      "step": 38120
    },
    {
      "entropy": 6.028314638137817,
      "epoch": 4.399884593190999,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000321707475158214,
      "loss": 5.5398,
      "mean_token_accuracy": 0.21656513065099717,
      "num_tokens": 96639928.0,
      "step": 38125
    },
    {
      "entropy": 6.048867321014404,
      "epoch": 4.400461627236007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003216671089162303,
      "loss": 5.4641,
      "mean_token_accuracy": 0.2193652793765068,
      "num_tokens": 96651262.0,
      "step": 38130
    },
    {
      "entropy": 6.079772043228149,
      "epoch": 4.4010386612810155,
      "grad_norm": 1.0,
      "learning_rate": 0.0003216267411046216,
      "loss": 5.5717,
      "mean_token_accuracy": 0.20986044257879258,
      "num_tokens": 96663570.0,
      "step": 38135
    },
    {
      "entropy": 6.106496143341064,
      "epoch": 4.401615695326024,
      "grad_norm": 1.0,
      "learning_rate": 0.00032158637172474556,
      "loss": 5.5906,
      "mean_token_accuracy": 0.20118582844734192,
      "num_tokens": 96675121.0,
      "step": 38140
    },
    {
      "entropy": 6.033359622955322,
      "epoch": 4.402192729371033,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003215460007779601,
      "loss": 5.5123,
      "mean_token_accuracy": 0.21977071613073348,
      "num_tokens": 96688228.0,
      "step": 38145
    },
    {
      "entropy": 6.151213884353638,
      "epoch": 4.402769763416042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032150562826562294,
      "loss": 5.5662,
      "mean_token_accuracy": 0.2116177797317505,
      "num_tokens": 96701902.0,
      "step": 38150
    },
    {
      "entropy": 6.0421003818511965,
      "epoch": 4.40334679746105,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003214652541890922,
      "loss": 5.5119,
      "mean_token_accuracy": 0.21780835837125778,
      "num_tokens": 96714322.0,
      "step": 38155
    },
    {
      "entropy": 6.055783224105835,
      "epoch": 4.4039238315060585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003214248785497256,
      "loss": 5.5263,
      "mean_token_accuracy": 0.21067825853824615,
      "num_tokens": 96728137.0,
      "step": 38160
    },
    {
      "entropy": 6.055116367340088,
      "epoch": 4.404500865551068,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003213845013488813,
      "loss": 5.5202,
      "mean_token_accuracy": 0.2144872322678566,
      "num_tokens": 96741200.0,
      "step": 38165
    },
    {
      "entropy": 5.973556470870972,
      "epoch": 4.405077899596076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003213441225879174,
      "loss": 5.4432,
      "mean_token_accuracy": 0.22491636425256728,
      "num_tokens": 96753530.0,
      "step": 38170
    },
    {
      "entropy": 5.817392778396607,
      "epoch": 4.405654933641085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032130374226819195,
      "loss": 5.2288,
      "mean_token_accuracy": 0.23348209708929063,
      "num_tokens": 96765487.0,
      "step": 38175
    },
    {
      "entropy": 6.150087451934814,
      "epoch": 4.406231967686093,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003212633603910631,
      "loss": 5.5724,
      "mean_token_accuracy": 0.20882961004972458,
      "num_tokens": 96777480.0,
      "step": 38180
    },
    {
      "entropy": 6.076144504547119,
      "epoch": 4.406809001731102,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032122297695788907,
      "loss": 5.5624,
      "mean_token_accuracy": 0.21048825085163117,
      "num_tokens": 96790028.0,
      "step": 38185
    },
    {
      "entropy": 6.079936838150024,
      "epoch": 4.407386035776111,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003211825919700282,
      "loss": 5.5713,
      "mean_token_accuracy": 0.20939162522554397,
      "num_tokens": 96803678.0,
      "step": 38190
    },
    {
      "entropy": 6.196528768539428,
      "epoch": 4.407963069821119,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032114220542883877,
      "loss": 5.6776,
      "mean_token_accuracy": 0.20235282182693481,
      "num_tokens": 96817116.0,
      "step": 38195
    },
    {
      "entropy": 6.044260406494141,
      "epoch": 4.4085401038661285,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00032110181733567924,
      "loss": 5.4394,
      "mean_token_accuracy": 0.2215522438287735,
      "num_tokens": 96828640.0,
      "step": 38200
    },
    {
      "entropy": 6.041435384750367,
      "epoch": 4.409117137911137,
      "grad_norm": 1.0,
      "learning_rate": 0.00032106142769190786,
      "loss": 5.568,
      "mean_token_accuracy": 0.20948518961668014,
      "num_tokens": 96840859.0,
      "step": 38205
    },
    {
      "entropy": 6.105862188339233,
      "epoch": 4.409694171956145,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00032102103649888334,
      "loss": 5.564,
      "mean_token_accuracy": 0.21221884191036225,
      "num_tokens": 96853073.0,
      "step": 38210
    },
    {
      "entropy": 6.09613823890686,
      "epoch": 4.410271206001154,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032098064375796415,
      "loss": 5.6518,
      "mean_token_accuracy": 0.2037271037697792,
      "num_tokens": 96866139.0,
      "step": 38215
    },
    {
      "entropy": 6.044867134094238,
      "epoch": 4.410848240046163,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032094024947050884,
      "loss": 5.5423,
      "mean_token_accuracy": 0.21265282928943635,
      "num_tokens": 96878203.0,
      "step": 38220
    },
    {
      "entropy": 6.104527139663697,
      "epoch": 4.4114252740911715,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003208998536378759,
      "loss": 5.5841,
      "mean_token_accuracy": 0.20305479019880296,
      "num_tokens": 96891132.0,
      "step": 38225
    },
    {
      "entropy": 6.06794056892395,
      "epoch": 4.41200230813618,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032085945626142437,
      "loss": 5.4644,
      "mean_token_accuracy": 0.2225763276219368,
      "num_tokens": 96903713.0,
      "step": 38230
    },
    {
      "entropy": 6.054664421081543,
      "epoch": 4.412579342181188,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032081905734251276,
      "loss": 5.5274,
      "mean_token_accuracy": 0.2090901955962181,
      "num_tokens": 96916654.0,
      "step": 38235
    },
    {
      "entropy": 6.17134804725647,
      "epoch": 4.413156376226198,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003207786568824999,
      "loss": 5.6572,
      "mean_token_accuracy": 0.19880637973546983,
      "num_tokens": 96929960.0,
      "step": 38240
    },
    {
      "entropy": 6.082594776153565,
      "epoch": 4.413733410271206,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003207382548827446,
      "loss": 5.5599,
      "mean_token_accuracy": 0.21472093909978868,
      "num_tokens": 96942529.0,
      "step": 38245
    },
    {
      "entropy": 6.145193481445313,
      "epoch": 4.414310444316214,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032069785134460587,
      "loss": 5.5676,
      "mean_token_accuracy": 0.20699049532413483,
      "num_tokens": 96955561.0,
      "step": 38250
    },
    {
      "entropy": 6.000232267379761,
      "epoch": 4.414887478361223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003206574462694426,
      "loss": 5.4674,
      "mean_token_accuracy": 0.2158682733774185,
      "num_tokens": 96967955.0,
      "step": 38255
    },
    {
      "entropy": 6.037531232833862,
      "epoch": 4.415464512406232,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032061703965861383,
      "loss": 5.5685,
      "mean_token_accuracy": 0.21087002158164977,
      "num_tokens": 96980858.0,
      "step": 38260
    },
    {
      "entropy": 6.073329925537109,
      "epoch": 4.416041546451241,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032057663151347854,
      "loss": 5.6017,
      "mean_token_accuracy": 0.20727828443050383,
      "num_tokens": 96995292.0,
      "step": 38265
    },
    {
      "entropy": 6.064919710159302,
      "epoch": 4.416618580496249,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032053622183539593,
      "loss": 5.4899,
      "mean_token_accuracy": 0.21314300298690797,
      "num_tokens": 97007968.0,
      "step": 38270
    },
    {
      "entropy": 6.102100944519043,
      "epoch": 4.417195614541258,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00032049581062572514,
      "loss": 5.5244,
      "mean_token_accuracy": 0.21712785661220552,
      "num_tokens": 97021600.0,
      "step": 38275
    },
    {
      "entropy": 5.9598664283752445,
      "epoch": 4.417772648586267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032045539788582534,
      "loss": 5.3264,
      "mean_token_accuracy": 0.22414463758468628,
      "num_tokens": 97032529.0,
      "step": 38280
    },
    {
      "entropy": 6.038284587860107,
      "epoch": 4.418349682631275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032041498361705574,
      "loss": 5.5662,
      "mean_token_accuracy": 0.20591207593679428,
      "num_tokens": 97045046.0,
      "step": 38285
    },
    {
      "entropy": 6.0646473407745365,
      "epoch": 4.418926716676284,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003203745678207758,
      "loss": 5.4344,
      "mean_token_accuracy": 0.21799015998840332,
      "num_tokens": 97058844.0,
      "step": 38290
    },
    {
      "entropy": 5.99240460395813,
      "epoch": 4.419503750721293,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032033415049834477,
      "loss": 5.466,
      "mean_token_accuracy": 0.21117955297231675,
      "num_tokens": 97072002.0,
      "step": 38295
    },
    {
      "entropy": 5.937837028503418,
      "epoch": 4.420080784766301,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032029373165112215,
      "loss": 5.4274,
      "mean_token_accuracy": 0.22193049788475036,
      "num_tokens": 97083861.0,
      "step": 38300
    },
    {
      "entropy": 6.031952571868897,
      "epoch": 4.42065781881131,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032025331128046733,
      "loss": 5.479,
      "mean_token_accuracy": 0.22137761861085892,
      "num_tokens": 97096039.0,
      "step": 38305
    },
    {
      "entropy": 6.109057283401489,
      "epoch": 4.421234852856318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032021288938773984,
      "loss": 5.59,
      "mean_token_accuracy": 0.20899964421987532,
      "num_tokens": 97108306.0,
      "step": 38310
    },
    {
      "entropy": 6.1162055969238285,
      "epoch": 4.421811886901327,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003201724659742993,
      "loss": 5.4904,
      "mean_token_accuracy": 0.21170012652873993,
      "num_tokens": 97120028.0,
      "step": 38315
    },
    {
      "entropy": 6.043722438812256,
      "epoch": 4.422388920946336,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032013204104150536,
      "loss": 5.4081,
      "mean_token_accuracy": 0.22663921117782593,
      "num_tokens": 97133741.0,
      "step": 38320
    },
    {
      "entropy": 5.949843168258667,
      "epoch": 4.422965954991344,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032009161459071756,
      "loss": 5.4432,
      "mean_token_accuracy": 0.22553859651088715,
      "num_tokens": 97145853.0,
      "step": 38325
    },
    {
      "entropy": 5.97238392829895,
      "epoch": 4.423542989036353,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003200511866232957,
      "loss": 5.4444,
      "mean_token_accuracy": 0.2190261095762253,
      "num_tokens": 97159771.0,
      "step": 38330
    },
    {
      "entropy": 5.973358106613159,
      "epoch": 4.424120023081362,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032001075714059964,
      "loss": 5.541,
      "mean_token_accuracy": 0.20798029750585556,
      "num_tokens": 97173415.0,
      "step": 38335
    },
    {
      "entropy": 6.00677375793457,
      "epoch": 4.42469705712637,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003199703261439891,
      "loss": 5.4068,
      "mean_token_accuracy": 0.22706763744354247,
      "num_tokens": 97186742.0,
      "step": 38340
    },
    {
      "entropy": 6.056438970565796,
      "epoch": 4.425274091171379,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003199298936348239,
      "loss": 5.5453,
      "mean_token_accuracy": 0.21169002652168273,
      "num_tokens": 97199553.0,
      "step": 38345
    },
    {
      "entropy": 6.038198804855346,
      "epoch": 4.425851125216388,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000319889459614464,
      "loss": 5.5512,
      "mean_token_accuracy": 0.21199320554733275,
      "num_tokens": 97212603.0,
      "step": 38350
    },
    {
      "entropy": 6.041984128952026,
      "epoch": 4.4264281592613965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003198490240842696,
      "loss": 5.4768,
      "mean_token_accuracy": 0.21151960492134095,
      "num_tokens": 97226247.0,
      "step": 38355
    },
    {
      "entropy": 6.065248250961304,
      "epoch": 4.427005193306405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031980858704560043,
      "loss": 5.5227,
      "mean_token_accuracy": 0.21060474812984467,
      "num_tokens": 97237875.0,
      "step": 38360
    },
    {
      "entropy": 6.028769207000733,
      "epoch": 4.427582227351413,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003197681484998168,
      "loss": 5.5282,
      "mean_token_accuracy": 0.20810441821813583,
      "num_tokens": 97251988.0,
      "step": 38365
    },
    {
      "entropy": 6.140949821472168,
      "epoch": 4.428159261396423,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031972770844827857,
      "loss": 5.575,
      "mean_token_accuracy": 0.2073147252202034,
      "num_tokens": 97265002.0,
      "step": 38370
    },
    {
      "entropy": 6.047188758850098,
      "epoch": 4.428736295441431,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031968726689234625,
      "loss": 5.5182,
      "mean_token_accuracy": 0.21059317141771317,
      "num_tokens": 97278527.0,
      "step": 38375
    },
    {
      "entropy": 6.038583421707154,
      "epoch": 4.4293133294864395,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031964682383337986,
      "loss": 5.5956,
      "mean_token_accuracy": 0.20662633329629898,
      "num_tokens": 97290813.0,
      "step": 38380
    },
    {
      "entropy": 6.067373323440552,
      "epoch": 4.429890363531449,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003196063792727397,
      "loss": 5.4897,
      "mean_token_accuracy": 0.2151740238070488,
      "num_tokens": 97303389.0,
      "step": 38385
    },
    {
      "entropy": 6.114503574371338,
      "epoch": 4.430467397576457,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003195659332117861,
      "loss": 5.5621,
      "mean_token_accuracy": 0.2133955329656601,
      "num_tokens": 97315058.0,
      "step": 38390
    },
    {
      "entropy": 6.043418598175049,
      "epoch": 4.431044431621466,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003195254856518795,
      "loss": 5.4771,
      "mean_token_accuracy": 0.21980277001857756,
      "num_tokens": 97328005.0,
      "step": 38395
    },
    {
      "entropy": 6.108169317245483,
      "epoch": 4.431621465666474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003194850365943803,
      "loss": 5.5733,
      "mean_token_accuracy": 0.20297864824533463,
      "num_tokens": 97340044.0,
      "step": 38400
    },
    {
      "entropy": 6.052076005935669,
      "epoch": 4.432198499711483,
      "grad_norm": 1.03125,
      "learning_rate": 0.000319444586040649,
      "loss": 5.4874,
      "mean_token_accuracy": 0.22086975425481797,
      "num_tokens": 97352011.0,
      "step": 38405
    },
    {
      "entropy": 6.056350898742676,
      "epoch": 4.432775533756492,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031940413399204616,
      "loss": 5.6007,
      "mean_token_accuracy": 0.20261027812957763,
      "num_tokens": 97364493.0,
      "step": 38410
    },
    {
      "entropy": 6.08653416633606,
      "epoch": 4.4333525678015,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031936368044993227,
      "loss": 5.5044,
      "mean_token_accuracy": 0.22014954835176467,
      "num_tokens": 97377797.0,
      "step": 38415
    },
    {
      "entropy": 6.049614381790161,
      "epoch": 4.433929601846509,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031932322541566806,
      "loss": 5.4614,
      "mean_token_accuracy": 0.22082065790891647,
      "num_tokens": 97389928.0,
      "step": 38420
    },
    {
      "entropy": 6.061914253234863,
      "epoch": 4.434506635891518,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003192827688906142,
      "loss": 5.5037,
      "mean_token_accuracy": 0.22080715000629425,
      "num_tokens": 97402978.0,
      "step": 38425
    },
    {
      "entropy": 6.032214164733887,
      "epoch": 4.435083669936526,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003192423108761314,
      "loss": 5.5194,
      "mean_token_accuracy": 0.20826325714588165,
      "num_tokens": 97415084.0,
      "step": 38430
    },
    {
      "entropy": 5.987477779388428,
      "epoch": 4.435660703981535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003192018513735804,
      "loss": 5.4147,
      "mean_token_accuracy": 0.22825562506914138,
      "num_tokens": 97426831.0,
      "step": 38435
    },
    {
      "entropy": 6.010431146621704,
      "epoch": 4.436237738026543,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00031916139038432213,
      "loss": 5.4509,
      "mean_token_accuracy": 0.22023300528526307,
      "num_tokens": 97439148.0,
      "step": 38440
    },
    {
      "entropy": 6.116638994216919,
      "epoch": 4.4368147720715525,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003191209279097174,
      "loss": 5.6753,
      "mean_token_accuracy": 0.20192406475543975,
      "num_tokens": 97453106.0,
      "step": 38445
    },
    {
      "entropy": 6.128074216842651,
      "epoch": 4.437391806116561,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031908046395112715,
      "loss": 5.6097,
      "mean_token_accuracy": 0.20352471321821214,
      "num_tokens": 97465735.0,
      "step": 38450
    },
    {
      "entropy": 6.0342264652252195,
      "epoch": 4.437968840161569,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031903999850991247,
      "loss": 5.4692,
      "mean_token_accuracy": 0.21654055118560792,
      "num_tokens": 97478584.0,
      "step": 38455
    },
    {
      "entropy": 6.051268625259399,
      "epoch": 4.438545874206579,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031899953158743427,
      "loss": 5.4289,
      "mean_token_accuracy": 0.22036666721105574,
      "num_tokens": 97491940.0,
      "step": 38460
    },
    {
      "entropy": 6.020132160186767,
      "epoch": 4.439122908251587,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003189590631850537,
      "loss": 5.4934,
      "mean_token_accuracy": 0.21575690656900406,
      "num_tokens": 97505625.0,
      "step": 38465
    },
    {
      "entropy": 6.147014331817627,
      "epoch": 4.4396999422965955,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031891859330413186,
      "loss": 5.6382,
      "mean_token_accuracy": 0.19720063507556915,
      "num_tokens": 97518074.0,
      "step": 38470
    },
    {
      "entropy": 6.106230306625366,
      "epoch": 4.440276976341604,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003188781219460299,
      "loss": 5.5278,
      "mean_token_accuracy": 0.21203760504722596,
      "num_tokens": 97530466.0,
      "step": 38475
    },
    {
      "entropy": 6.131121587753296,
      "epoch": 4.440854010386613,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031883764911210915,
      "loss": 5.6276,
      "mean_token_accuracy": 0.20531610250473023,
      "num_tokens": 97543658.0,
      "step": 38480
    },
    {
      "entropy": 5.995190858840942,
      "epoch": 4.441431044431622,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003187971748037309,
      "loss": 5.4657,
      "mean_token_accuracy": 0.21458000391721727,
      "num_tokens": 97557243.0,
      "step": 38485
    },
    {
      "entropy": 6.0263184070587155,
      "epoch": 4.44200807847663,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031875669902225643,
      "loss": 5.4051,
      "mean_token_accuracy": 0.2193518966436386,
      "num_tokens": 97569230.0,
      "step": 38490
    },
    {
      "entropy": 6.127245664596558,
      "epoch": 4.4425851125216385,
      "grad_norm": 1.0,
      "learning_rate": 0.000318716221769047,
      "loss": 5.641,
      "mean_token_accuracy": 0.20181012749671937,
      "num_tokens": 97582068.0,
      "step": 38495
    },
    {
      "entropy": 6.013142681121826,
      "epoch": 4.443162146566648,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003186757430454643,
      "loss": 5.4241,
      "mean_token_accuracy": 0.22312005013227462,
      "num_tokens": 97595039.0,
      "step": 38500
    },
    {
      "entropy": 5.983185434341431,
      "epoch": 4.443739180611656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031863526285286957,
      "loss": 5.4516,
      "mean_token_accuracy": 0.2144797459244728,
      "num_tokens": 97606703.0,
      "step": 38505
    },
    {
      "entropy": 6.091512298583984,
      "epoch": 4.444316214656665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003185947811926246,
      "loss": 5.5426,
      "mean_token_accuracy": 0.2147781506180763,
      "num_tokens": 97618401.0,
      "step": 38510
    },
    {
      "entropy": 6.073824739456176,
      "epoch": 4.444893248701673,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003185542980660906,
      "loss": 5.4908,
      "mean_token_accuracy": 0.21741002053022385,
      "num_tokens": 97630650.0,
      "step": 38515
    },
    {
      "entropy": 5.92982702255249,
      "epoch": 4.445470282746682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003185138134746296,
      "loss": 5.3744,
      "mean_token_accuracy": 0.21576071232557298,
      "num_tokens": 97643192.0,
      "step": 38520
    },
    {
      "entropy": 6.038323068618775,
      "epoch": 4.446047316791691,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000318473327419603,
      "loss": 5.525,
      "mean_token_accuracy": 0.2184448093175888,
      "num_tokens": 97656559.0,
      "step": 38525
    },
    {
      "entropy": 6.114233589172363,
      "epoch": 4.446624350836699,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003184328399023727,
      "loss": 5.5604,
      "mean_token_accuracy": 0.20933341085910798,
      "num_tokens": 97669778.0,
      "step": 38530
    },
    {
      "entropy": 6.106561136245728,
      "epoch": 4.4472013848817085,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003183923509243003,
      "loss": 5.5675,
      "mean_token_accuracy": 0.21007578521966935,
      "num_tokens": 97682260.0,
      "step": 38535
    },
    {
      "entropy": 6.082838439941407,
      "epoch": 4.447778418926717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031835186048674786,
      "loss": 5.6133,
      "mean_token_accuracy": 0.20035274624824523,
      "num_tokens": 97693812.0,
      "step": 38540
    },
    {
      "entropy": 6.056507349014282,
      "epoch": 4.448355452971725,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003183113685910771,
      "loss": 5.5248,
      "mean_token_accuracy": 0.2132405236363411,
      "num_tokens": 97707149.0,
      "step": 38545
    },
    {
      "entropy": 6.113850736618042,
      "epoch": 4.448932487016734,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003182708752386499,
      "loss": 5.548,
      "mean_token_accuracy": 0.20951681733131408,
      "num_tokens": 97719202.0,
      "step": 38550
    },
    {
      "entropy": 6.106725692749023,
      "epoch": 4.449509521061743,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003182303804308283,
      "loss": 5.5999,
      "mean_token_accuracy": 0.20964802503585817,
      "num_tokens": 97730958.0,
      "step": 38555
    },
    {
      "entropy": 6.147523260116577,
      "epoch": 4.450086555106751,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031818988416897444,
      "loss": 5.5938,
      "mean_token_accuracy": 0.2121189519762993,
      "num_tokens": 97743366.0,
      "step": 38560
    },
    {
      "entropy": 6.031114006042481,
      "epoch": 4.45066358915176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003181493864544502,
      "loss": 5.4968,
      "mean_token_accuracy": 0.2140096455812454,
      "num_tokens": 97756140.0,
      "step": 38565
    },
    {
      "entropy": 6.026924562454224,
      "epoch": 4.451240623196768,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003181088872886178,
      "loss": 5.5173,
      "mean_token_accuracy": 0.2145009696483612,
      "num_tokens": 97770092.0,
      "step": 38570
    },
    {
      "entropy": 6.0982198238372805,
      "epoch": 4.451817657241778,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003180683866728394,
      "loss": 5.5159,
      "mean_token_accuracy": 0.21397635191679001,
      "num_tokens": 97783815.0,
      "step": 38575
    },
    {
      "entropy": 6.046237373352051,
      "epoch": 4.452394691286786,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031802788460847715,
      "loss": 5.527,
      "mean_token_accuracy": 0.2051728844642639,
      "num_tokens": 97796149.0,
      "step": 38580
    },
    {
      "entropy": 6.053700733184814,
      "epoch": 4.452971725331794,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003179873810968934,
      "loss": 5.5285,
      "mean_token_accuracy": 0.2062440723180771,
      "num_tokens": 97808310.0,
      "step": 38585
    },
    {
      "entropy": 6.0198523044586185,
      "epoch": 4.453548759376803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003179468761394504,
      "loss": 5.4264,
      "mean_token_accuracy": 0.21690552532672883,
      "num_tokens": 97820840.0,
      "step": 38590
    },
    {
      "entropy": 6.087871742248535,
      "epoch": 4.454125793421812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003179063697375107,
      "loss": 5.5598,
      "mean_token_accuracy": 0.21034959852695465,
      "num_tokens": 97834359.0,
      "step": 38595
    },
    {
      "entropy": 6.097955799102783,
      "epoch": 4.454702827466821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031786586189243646,
      "loss": 5.5552,
      "mean_token_accuracy": 0.20411264896392822,
      "num_tokens": 97848018.0,
      "step": 38600
    },
    {
      "entropy": 6.0513042449951175,
      "epoch": 4.455279861511829,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003178253526055902,
      "loss": 5.5245,
      "mean_token_accuracy": 0.21323073655366898,
      "num_tokens": 97861598.0,
      "step": 38605
    },
    {
      "entropy": 6.0582154273986815,
      "epoch": 4.455856895556838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031778484187833456,
      "loss": 5.5435,
      "mean_token_accuracy": 0.21287302672863007,
      "num_tokens": 97873280.0,
      "step": 38610
    },
    {
      "entropy": 6.063132095336914,
      "epoch": 4.456433929601847,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031774432971203205,
      "loss": 5.5142,
      "mean_token_accuracy": 0.20476548224687577,
      "num_tokens": 97886713.0,
      "step": 38615
    },
    {
      "entropy": 5.940737724304199,
      "epoch": 4.457010963646855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003177038161080451,
      "loss": 5.4044,
      "mean_token_accuracy": 0.22060102969408035,
      "num_tokens": 97899233.0,
      "step": 38620
    },
    {
      "entropy": 5.959931564331055,
      "epoch": 4.4575879976918635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003176633010677366,
      "loss": 5.4166,
      "mean_token_accuracy": 0.21378268748521806,
      "num_tokens": 97912069.0,
      "step": 38625
    },
    {
      "entropy": 6.086202383041382,
      "epoch": 4.458165031736873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031762278459246916,
      "loss": 5.5286,
      "mean_token_accuracy": 0.21167900115251542,
      "num_tokens": 97925095.0,
      "step": 38630
    },
    {
      "entropy": 6.127517127990723,
      "epoch": 4.458742065781881,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031758226668360555,
      "loss": 5.5918,
      "mean_token_accuracy": 0.20698390305042266,
      "num_tokens": 97937648.0,
      "step": 38635
    },
    {
      "entropy": 6.07396388053894,
      "epoch": 4.45931909982689,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031754174734250853,
      "loss": 5.5258,
      "mean_token_accuracy": 0.21063694059848787,
      "num_tokens": 97950182.0,
      "step": 38640
    },
    {
      "entropy": 6.0780500888824465,
      "epoch": 4.459896133871898,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003175012265705409,
      "loss": 5.5295,
      "mean_token_accuracy": 0.20857050865888596,
      "num_tokens": 97963099.0,
      "step": 38645
    },
    {
      "entropy": 6.1557262420654295,
      "epoch": 4.460473167916907,
      "grad_norm": 0.890625,
      "learning_rate": 0.00031746070436906577,
      "loss": 5.601,
      "mean_token_accuracy": 0.2056301310658455,
      "num_tokens": 97976061.0,
      "step": 38650
    },
    {
      "entropy": 6.099689579010009,
      "epoch": 4.461050201961916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003174201807394459,
      "loss": 5.5265,
      "mean_token_accuracy": 0.20930353850126265,
      "num_tokens": 97988096.0,
      "step": 38655
    },
    {
      "entropy": 6.026842308044434,
      "epoch": 4.461627236006924,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003173796556830442,
      "loss": 5.5061,
      "mean_token_accuracy": 0.2063517689704895,
      "num_tokens": 98000408.0,
      "step": 38660
    },
    {
      "entropy": 6.055496644973755,
      "epoch": 4.462204270051933,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00031733912920122397,
      "loss": 5.5559,
      "mean_token_accuracy": 0.21148732602596282,
      "num_tokens": 98014132.0,
      "step": 38665
    },
    {
      "entropy": 6.104442691802978,
      "epoch": 4.462781304096942,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003172986012953481,
      "loss": 5.5082,
      "mean_token_accuracy": 0.20375020205974578,
      "num_tokens": 98026987.0,
      "step": 38670
    },
    {
      "entropy": 6.015497159957886,
      "epoch": 4.46335833814195,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003172580719667798,
      "loss": 5.4273,
      "mean_token_accuracy": 0.21826139390468596,
      "num_tokens": 98039480.0,
      "step": 38675
    },
    {
      "entropy": 5.990102529525757,
      "epoch": 4.463935372186959,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031721754121688225,
      "loss": 5.4396,
      "mean_token_accuracy": 0.2210967093706131,
      "num_tokens": 98051633.0,
      "step": 38680
    },
    {
      "entropy": 6.13384165763855,
      "epoch": 4.464512406231968,
      "grad_norm": 0.921875,
      "learning_rate": 0.00031717700904701864,
      "loss": 5.6509,
      "mean_token_accuracy": 0.20338100343942642,
      "num_tokens": 98065704.0,
      "step": 38685
    },
    {
      "entropy": 6.004337930679322,
      "epoch": 4.4650894402769765,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003171364754585523,
      "loss": 5.4479,
      "mean_token_accuracy": 0.21626182943582534,
      "num_tokens": 98079091.0,
      "step": 38690
    },
    {
      "entropy": 6.060029745101929,
      "epoch": 4.465666474321985,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031709594045284655,
      "loss": 5.5218,
      "mean_token_accuracy": 0.2135891318321228,
      "num_tokens": 98091735.0,
      "step": 38695
    },
    {
      "entropy": 6.064213514328003,
      "epoch": 4.466243508366993,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003170554040312647,
      "loss": 5.5455,
      "mean_token_accuracy": 0.2111101806163788,
      "num_tokens": 98105490.0,
      "step": 38700
    },
    {
      "entropy": 6.0144736766815186,
      "epoch": 4.466820542412003,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003170148661951703,
      "loss": 5.4559,
      "mean_token_accuracy": 0.21411906331777572,
      "num_tokens": 98118936.0,
      "step": 38705
    },
    {
      "entropy": 6.052533435821533,
      "epoch": 4.467397576457011,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003169743269459267,
      "loss": 5.5169,
      "mean_token_accuracy": 0.22226731330156327,
      "num_tokens": 98130627.0,
      "step": 38710
    },
    {
      "entropy": 6.134881830215454,
      "epoch": 4.4679746105020195,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031693378628489747,
      "loss": 5.5054,
      "mean_token_accuracy": 0.21126098185777664,
      "num_tokens": 98142408.0,
      "step": 38715
    },
    {
      "entropy": 6.083096122741699,
      "epoch": 4.468551644547028,
      "grad_norm": 1.125,
      "learning_rate": 0.0003168932442134461,
      "loss": 5.5642,
      "mean_token_accuracy": 0.21005493253469468,
      "num_tokens": 98155158.0,
      "step": 38720
    },
    {
      "entropy": 5.962113094329834,
      "epoch": 4.469128678592037,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031685270073293637,
      "loss": 5.5078,
      "mean_token_accuracy": 0.21187253594398497,
      "num_tokens": 98167981.0,
      "step": 38725
    },
    {
      "entropy": 6.01308856010437,
      "epoch": 4.469705712637046,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031681215584473184,
      "loss": 5.4618,
      "mean_token_accuracy": 0.2126472532749176,
      "num_tokens": 98180974.0,
      "step": 38730
    },
    {
      "entropy": 5.95109748840332,
      "epoch": 4.470282746682054,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031677160955019616,
      "loss": 5.3795,
      "mean_token_accuracy": 0.23298151344060897,
      "num_tokens": 98194857.0,
      "step": 38735
    },
    {
      "entropy": 6.053857326507568,
      "epoch": 4.4708597807270625,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031673106185069324,
      "loss": 5.524,
      "mean_token_accuracy": 0.21251145154237747,
      "num_tokens": 98206695.0,
      "step": 38740
    },
    {
      "entropy": 6.074065923690796,
      "epoch": 4.471436814772072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003166905127475866,
      "loss": 5.5205,
      "mean_token_accuracy": 0.2104508951306343,
      "num_tokens": 98219132.0,
      "step": 38745
    },
    {
      "entropy": 6.0658153057098385,
      "epoch": 4.47201384881708,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031664996224224043,
      "loss": 5.5287,
      "mean_token_accuracy": 0.20999520421028137,
      "num_tokens": 98230074.0,
      "step": 38750
    },
    {
      "entropy": 5.989320278167725,
      "epoch": 4.472590882862089,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003166094103360184,
      "loss": 5.487,
      "mean_token_accuracy": 0.21548508554697038,
      "num_tokens": 98242503.0,
      "step": 38755
    },
    {
      "entropy": 6.089497661590576,
      "epoch": 4.473167916907098,
      "grad_norm": 0.875,
      "learning_rate": 0.0003165688570302846,
      "loss": 5.5763,
      "mean_token_accuracy": 0.207134610414505,
      "num_tokens": 98254874.0,
      "step": 38760
    },
    {
      "entropy": 6.054527139663696,
      "epoch": 4.473744950952106,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031652830232640285,
      "loss": 5.4548,
      "mean_token_accuracy": 0.20933595001697541,
      "num_tokens": 98266847.0,
      "step": 38765
    },
    {
      "entropy": 6.097812557220459,
      "epoch": 4.474321984997115,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031648774622573735,
      "loss": 5.6257,
      "mean_token_accuracy": 0.20380248576402665,
      "num_tokens": 98279818.0,
      "step": 38770
    },
    {
      "entropy": 5.966616678237915,
      "epoch": 4.474899019042123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031644718872965205,
      "loss": 5.3859,
      "mean_token_accuracy": 0.23135214000940324,
      "num_tokens": 98292648.0,
      "step": 38775
    },
    {
      "entropy": 6.097874021530151,
      "epoch": 4.4754760530871325,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003164066298395112,
      "loss": 5.5779,
      "mean_token_accuracy": 0.2002714082598686,
      "num_tokens": 98305553.0,
      "step": 38780
    },
    {
      "entropy": 6.0839455127716064,
      "epoch": 4.476053087132141,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003163660695566789,
      "loss": 5.4562,
      "mean_token_accuracy": 0.2160343959927559,
      "num_tokens": 98317236.0,
      "step": 38785
    },
    {
      "entropy": 5.966993093490601,
      "epoch": 4.476630121177149,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031632550788251936,
      "loss": 5.4073,
      "mean_token_accuracy": 0.22457791566848756,
      "num_tokens": 98329991.0,
      "step": 38790
    },
    {
      "entropy": 6.011926174163818,
      "epoch": 4.477207155222158,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003162849448183969,
      "loss": 5.5234,
      "mean_token_accuracy": 0.21728626638650894,
      "num_tokens": 98342930.0,
      "step": 38795
    },
    {
      "entropy": 6.074259042739868,
      "epoch": 4.477784189267167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003162443803656759,
      "loss": 5.5529,
      "mean_token_accuracy": 0.21781429499387742,
      "num_tokens": 98355358.0,
      "step": 38800
    },
    {
      "entropy": 6.026161241531372,
      "epoch": 4.4783612233121755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003162038145257206,
      "loss": 5.5152,
      "mean_token_accuracy": 0.21540616303682328,
      "num_tokens": 98368639.0,
      "step": 38805
    },
    {
      "entropy": 5.98667688369751,
      "epoch": 4.478938257357184,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003161632472998954,
      "loss": 5.459,
      "mean_token_accuracy": 0.21354019790887832,
      "num_tokens": 98381042.0,
      "step": 38810
    },
    {
      "entropy": 6.086393404006958,
      "epoch": 4.479515291402192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031612267868956495,
      "loss": 5.5599,
      "mean_token_accuracy": 0.2123895213007927,
      "num_tokens": 98393987.0,
      "step": 38815
    },
    {
      "entropy": 5.98841667175293,
      "epoch": 4.480092325447202,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003160821086960936,
      "loss": 5.3876,
      "mean_token_accuracy": 0.22227954119443893,
      "num_tokens": 98406369.0,
      "step": 38820
    },
    {
      "entropy": 6.043635034561158,
      "epoch": 4.48066935949221,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031604153732084594,
      "loss": 5.485,
      "mean_token_accuracy": 0.21418891698122025,
      "num_tokens": 98419175.0,
      "step": 38825
    },
    {
      "entropy": 6.076495361328125,
      "epoch": 4.481246393537218,
      "grad_norm": 1.0,
      "learning_rate": 0.0003160009645651865,
      "loss": 5.5823,
      "mean_token_accuracy": 0.203786039352417,
      "num_tokens": 98431422.0,
      "step": 38830
    },
    {
      "entropy": 6.102094507217407,
      "epoch": 4.481823427582228,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00031596039043048005,
      "loss": 5.5608,
      "mean_token_accuracy": 0.20975770354270934,
      "num_tokens": 98444610.0,
      "step": 38835
    },
    {
      "entropy": 6.111818552017212,
      "epoch": 4.482400461627236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031591981491809113,
      "loss": 5.4964,
      "mean_token_accuracy": 0.21085863262414933,
      "num_tokens": 98457090.0,
      "step": 38840
    },
    {
      "entropy": 6.059669399261475,
      "epoch": 4.482977495672245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031587923802938464,
      "loss": 5.5375,
      "mean_token_accuracy": 0.20962127447128295,
      "num_tokens": 98469530.0,
      "step": 38845
    },
    {
      "entropy": 6.090096759796142,
      "epoch": 4.483554529717253,
      "grad_norm": 3.28125,
      "learning_rate": 0.0003158386597657252,
      "loss": 5.4784,
      "mean_token_accuracy": 0.20562033653259276,
      "num_tokens": 98482378.0,
      "step": 38850
    },
    {
      "entropy": 6.155495977401733,
      "epoch": 4.484131563762262,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031579808012847775,
      "loss": 5.603,
      "mean_token_accuracy": 0.2058682084083557,
      "num_tokens": 98495958.0,
      "step": 38855
    },
    {
      "entropy": 5.990966415405273,
      "epoch": 4.484708597807271,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003157574991190072,
      "loss": 5.3948,
      "mean_token_accuracy": 0.22254487723112107,
      "num_tokens": 98508665.0,
      "step": 38860
    },
    {
      "entropy": 5.987441062927246,
      "epoch": 4.485285631852279,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003157169167386784,
      "loss": 5.501,
      "mean_token_accuracy": 0.2165486231446266,
      "num_tokens": 98521309.0,
      "step": 38865
    },
    {
      "entropy": 6.012189197540283,
      "epoch": 4.4858626658972875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003156763329888563,
      "loss": 5.4295,
      "mean_token_accuracy": 0.2175240471959114,
      "num_tokens": 98534160.0,
      "step": 38870
    },
    {
      "entropy": 6.043825292587281,
      "epoch": 4.486439699942297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003156357478709059,
      "loss": 5.5631,
      "mean_token_accuracy": 0.20687729567289354,
      "num_tokens": 98545770.0,
      "step": 38875
    },
    {
      "entropy": 6.077350521087647,
      "epoch": 4.487016733987305,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00031559516138619244,
      "loss": 5.5743,
      "mean_token_accuracy": 0.21374419331550598,
      "num_tokens": 98557667.0,
      "step": 38880
    },
    {
      "entropy": 6.131553983688354,
      "epoch": 4.487593768032314,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003155545735360808,
      "loss": 5.5127,
      "mean_token_accuracy": 0.21122974753379822,
      "num_tokens": 98569222.0,
      "step": 38885
    },
    {
      "entropy": 6.042634010314941,
      "epoch": 4.488170802077322,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003155139843219362,
      "loss": 5.5282,
      "mean_token_accuracy": 0.20985184907913207,
      "num_tokens": 98582771.0,
      "step": 38890
    },
    {
      "entropy": 6.042370176315307,
      "epoch": 4.488747836122331,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003154733937451239,
      "loss": 5.5473,
      "mean_token_accuracy": 0.21339697390794754,
      "num_tokens": 98595219.0,
      "step": 38895
    },
    {
      "entropy": 5.927421379089355,
      "epoch": 4.48932487016734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003154328018070091,
      "loss": 5.301,
      "mean_token_accuracy": 0.2331530526280403,
      "num_tokens": 98607449.0,
      "step": 38900
    },
    {
      "entropy": 6.0977026462554935,
      "epoch": 4.489901904212348,
      "grad_norm": 1.0,
      "learning_rate": 0.0003153922085089571,
      "loss": 5.5622,
      "mean_token_accuracy": 0.2080818682909012,
      "num_tokens": 98618702.0,
      "step": 38905
    },
    {
      "entropy": 6.082994508743286,
      "epoch": 4.4904789382573576,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031535161385233323,
      "loss": 5.5004,
      "mean_token_accuracy": 0.21401818841695786,
      "num_tokens": 98630797.0,
      "step": 38910
    },
    {
      "entropy": 6.07491660118103,
      "epoch": 4.491055972302366,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031531101783850284,
      "loss": 5.5688,
      "mean_token_accuracy": 0.20805783271789552,
      "num_tokens": 98642464.0,
      "step": 38915
    },
    {
      "entropy": 6.033314418792725,
      "epoch": 4.491633006347374,
      "grad_norm": 1.0,
      "learning_rate": 0.0003152704204688315,
      "loss": 5.4752,
      "mean_token_accuracy": 0.21202450543642043,
      "num_tokens": 98654451.0,
      "step": 38920
    },
    {
      "entropy": 6.072428512573242,
      "epoch": 4.492210040392383,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031522982174468444,
      "loss": 5.5286,
      "mean_token_accuracy": 0.2166917309165001,
      "num_tokens": 98667765.0,
      "step": 38925
    },
    {
      "entropy": 5.999984931945801,
      "epoch": 4.492787074437392,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031518922166742737,
      "loss": 5.4683,
      "mean_token_accuracy": 0.22029939591884612,
      "num_tokens": 98681996.0,
      "step": 38930
    },
    {
      "entropy": 6.0265637874603275,
      "epoch": 4.4933641084824005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003151486202384258,
      "loss": 5.4606,
      "mean_token_accuracy": 0.21285815984010698,
      "num_tokens": 98694394.0,
      "step": 38935
    },
    {
      "entropy": 6.0405174732208256,
      "epoch": 4.493941142527409,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00031510801745904526,
      "loss": 5.5394,
      "mean_token_accuracy": 0.2090293988585472,
      "num_tokens": 98706655.0,
      "step": 38940
    },
    {
      "entropy": 6.091063070297241,
      "epoch": 4.494518176572417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031506741333065154,
      "loss": 5.5394,
      "mean_token_accuracy": 0.21001248061656952,
      "num_tokens": 98718623.0,
      "step": 38945
    },
    {
      "entropy": 6.003238582611084,
      "epoch": 4.495095210617427,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003150268078546103,
      "loss": 5.5091,
      "mean_token_accuracy": 0.2139189913868904,
      "num_tokens": 98732002.0,
      "step": 38950
    },
    {
      "entropy": 6.030997848510742,
      "epoch": 4.495672244662435,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003149862010322872,
      "loss": 5.4784,
      "mean_token_accuracy": 0.21097345203161239,
      "num_tokens": 98744216.0,
      "step": 38955
    },
    {
      "entropy": 6.019388675689697,
      "epoch": 4.4962492787074435,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003149455928650481,
      "loss": 5.4755,
      "mean_token_accuracy": 0.2156911388039589,
      "num_tokens": 98755629.0,
      "step": 38960
    },
    {
      "entropy": 6.0697589874267575,
      "epoch": 4.496826312752452,
      "grad_norm": 1.0,
      "learning_rate": 0.0003149049833542589,
      "loss": 5.5341,
      "mean_token_accuracy": 0.20544422268867493,
      "num_tokens": 98767619.0,
      "step": 38965
    },
    {
      "entropy": 6.106377124786377,
      "epoch": 4.497403346797461,
      "grad_norm": 0.9375,
      "learning_rate": 0.00031486437250128534,
      "loss": 5.5258,
      "mean_token_accuracy": 0.20506151169538497,
      "num_tokens": 98779445.0,
      "step": 38970
    },
    {
      "entropy": 6.020062971115112,
      "epoch": 4.49798038084247,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003148237603074934,
      "loss": 5.4661,
      "mean_token_accuracy": 0.2236568883061409,
      "num_tokens": 98793433.0,
      "step": 38975
    },
    {
      "entropy": 6.076809215545654,
      "epoch": 4.498557414887478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000314783146774249,
      "loss": 5.5951,
      "mean_token_accuracy": 0.20681009739637374,
      "num_tokens": 98804870.0,
      "step": 38980
    },
    {
      "entropy": 6.022223472595215,
      "epoch": 4.499134448932487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00031474253190291843,
      "loss": 5.4706,
      "mean_token_accuracy": 0.21976792812347412,
      "num_tokens": 98818015.0,
      "step": 38985
    },
    {
      "entropy": 6.02209849357605,
      "epoch": 4.499711482977496,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003147019156948674,
      "loss": 5.5424,
      "mean_token_accuracy": 0.21284134536981583,
      "num_tokens": 98830496.0,
      "step": 38990
    },
    {
      "entropy": 6.084675073623657,
      "epoch": 4.500288517022504,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003146612981514621,
      "loss": 5.4818,
      "mean_token_accuracy": 0.22049425095319747,
      "num_tokens": 98843926.0,
      "step": 38995
    },
    {
      "entropy": 6.031761932373047,
      "epoch": 4.500865551067513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003146206792740688,
      "loss": 5.483,
      "mean_token_accuracy": 0.21585797816514968,
      "num_tokens": 98855792.0,
      "step": 39000
    },
    {
      "epoch": 4.500865551067513,
      "eval_entropy": 5.8790941585489795,
      "eval_loss": 5.683927059173584,
      "eval_mean_token_accuracy": 0.21207002264584945,
      "eval_num_tokens": 98855792.0,
      "eval_runtime": 22.8507,
      "eval_samples_per_second": 1231.298,
      "eval_steps_per_second": 153.912,
      "step": 39000
    },
    {
      "entropy": 6.058261871337891,
      "epoch": 4.501442585112522,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003145800590640537,
      "loss": 5.5797,
      "mean_token_accuracy": 0.20696779042482377,
      "num_tokens": 98866485.0,
      "step": 39005
    },
    {
      "entropy": 6.0534892082214355,
      "epoch": 4.50201961915753,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0003145394375227829,
      "loss": 5.4257,
      "mean_token_accuracy": 0.22497599571943283,
      "num_tokens": 98878786.0,
      "step": 39010
    },
    {
      "entropy": 6.135503911972046,
      "epoch": 4.502596653202539,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003144988146516227,
      "loss": 5.6134,
      "mean_token_accuracy": 0.212575863301754,
      "num_tokens": 98891882.0,
      "step": 39015
    },
    {
      "entropy": 6.118965768814087,
      "epoch": 4.503173687247548,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003144581904519396,
      "loss": 5.5803,
      "mean_token_accuracy": 0.20736913681030272,
      "num_tokens": 98904532.0,
      "step": 39020
    },
    {
      "entropy": 6.067539358139038,
      "epoch": 4.5037507212925565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003144175649250998,
      "loss": 5.4889,
      "mean_token_accuracy": 0.21242594122886657,
      "num_tokens": 98916378.0,
      "step": 39025
    },
    {
      "entropy": 6.091108798980713,
      "epoch": 4.504327755337565,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031437693807246985,
      "loss": 5.5891,
      "mean_token_accuracy": 0.20822285413742064,
      "num_tokens": 98929968.0,
      "step": 39030
    },
    {
      "entropy": 6.0288303852081295,
      "epoch": 4.504904789382573,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031433630989541604,
      "loss": 5.5164,
      "mean_token_accuracy": 0.20927223563194275,
      "num_tokens": 98943830.0,
      "step": 39035
    },
    {
      "entropy": 6.1323905944824215,
      "epoch": 4.505481823427582,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003142956803953051,
      "loss": 5.5545,
      "mean_token_accuracy": 0.21695072203874588,
      "num_tokens": 98956853.0,
      "step": 39040
    },
    {
      "entropy": 6.064871072769165,
      "epoch": 4.506058857472591,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00031425504957350344,
      "loss": 5.4957,
      "mean_token_accuracy": 0.21694860905408858,
      "num_tokens": 98968422.0,
      "step": 39045
    },
    {
      "entropy": 6.029691934585571,
      "epoch": 4.5066358915175995,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003142144174313777,
      "loss": 5.4738,
      "mean_token_accuracy": 0.21867966502904893,
      "num_tokens": 98981153.0,
      "step": 39050
    },
    {
      "entropy": 6.0517956733703615,
      "epoch": 4.507212925562608,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031417378397029445,
      "loss": 5.5162,
      "mean_token_accuracy": 0.21556487530469895,
      "num_tokens": 98994016.0,
      "step": 39055
    },
    {
      "entropy": 6.080186271667481,
      "epoch": 4.507789959607617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031413314919162045,
      "loss": 5.456,
      "mean_token_accuracy": 0.21728504151105882,
      "num_tokens": 99006564.0,
      "step": 39060
    },
    {
      "entropy": 6.032404899597168,
      "epoch": 4.508366993652626,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003140925130967224,
      "loss": 5.5638,
      "mean_token_accuracy": 0.21023759692907334,
      "num_tokens": 99020202.0,
      "step": 39065
    },
    {
      "entropy": 5.961019611358642,
      "epoch": 4.508944027697634,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003140518756869672,
      "loss": 5.4544,
      "mean_token_accuracy": 0.22588376253843306,
      "num_tokens": 99032965.0,
      "step": 39070
    },
    {
      "entropy": 6.044476556777954,
      "epoch": 4.5095210617426424,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00031401123696372143,
      "loss": 5.5551,
      "mean_token_accuracy": 0.2122431591153145,
      "num_tokens": 99047618.0,
      "step": 39075
    },
    {
      "entropy": 6.174966859817505,
      "epoch": 4.510098095787652,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031397059692835213,
      "loss": 5.6228,
      "mean_token_accuracy": 0.20695795565843583,
      "num_tokens": 99059866.0,
      "step": 39080
    },
    {
      "entropy": 6.129109907150268,
      "epoch": 4.51067512983266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00031392995558222615,
      "loss": 5.5893,
      "mean_token_accuracy": 0.20488440096378327,
      "num_tokens": 99071729.0,
      "step": 39085
    },
    {
      "entropy": 6.047139596939087,
      "epoch": 4.511252163877669,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031388931292671045,
      "loss": 5.4937,
      "mean_token_accuracy": 0.21503215581178664,
      "num_tokens": 99083766.0,
      "step": 39090
    },
    {
      "entropy": 6.013152599334717,
      "epoch": 4.511829197922678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000313848668963172,
      "loss": 5.4782,
      "mean_token_accuracy": 0.21693350970745087,
      "num_tokens": 99095110.0,
      "step": 39095
    },
    {
      "entropy": 5.990360593795776,
      "epoch": 4.512406231967686,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000313808023692978,
      "loss": 5.5355,
      "mean_token_accuracy": 0.21664632707834244,
      "num_tokens": 99107510.0,
      "step": 39100
    },
    {
      "entropy": 6.0199596881866455,
      "epoch": 4.512983266012695,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003137673771174953,
      "loss": 5.4882,
      "mean_token_accuracy": 0.2168717622756958,
      "num_tokens": 99120598.0,
      "step": 39105
    },
    {
      "entropy": 6.093764591217041,
      "epoch": 4.513560300057703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031372672923809115,
      "loss": 5.5957,
      "mean_token_accuracy": 0.20528324395418168,
      "num_tokens": 99133964.0,
      "step": 39110
    },
    {
      "entropy": 5.997952890396118,
      "epoch": 4.514137334102712,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031368608005613265,
      "loss": 5.4237,
      "mean_token_accuracy": 0.21976717859506606,
      "num_tokens": 99146739.0,
      "step": 39115
    },
    {
      "entropy": 6.052892732620239,
      "epoch": 4.514714368147721,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003136454295729871,
      "loss": 5.5149,
      "mean_token_accuracy": 0.20980237126350404,
      "num_tokens": 99158766.0,
      "step": 39120
    },
    {
      "entropy": 6.080827903747559,
      "epoch": 4.515291402192729,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003136047777900217,
      "loss": 5.5192,
      "mean_token_accuracy": 0.20828806459903718,
      "num_tokens": 99171849.0,
      "step": 39125
    },
    {
      "entropy": 6.003246450424195,
      "epoch": 4.515868436237738,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003135641247086038,
      "loss": 5.4754,
      "mean_token_accuracy": 0.21498167961835862,
      "num_tokens": 99183182.0,
      "step": 39130
    },
    {
      "entropy": 6.072053337097168,
      "epoch": 4.516445470282747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003135234703301008,
      "loss": 5.5074,
      "mean_token_accuracy": 0.21812089383602143,
      "num_tokens": 99195833.0,
      "step": 39135
    },
    {
      "entropy": 6.06141300201416,
      "epoch": 4.517022504327755,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003134828146558798,
      "loss": 5.4977,
      "mean_token_accuracy": 0.21854747235774993,
      "num_tokens": 99208546.0,
      "step": 39140
    },
    {
      "entropy": 6.058836126327515,
      "epoch": 4.517599538372764,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031344215768730856,
      "loss": 5.4879,
      "mean_token_accuracy": 0.2205967754125595,
      "num_tokens": 99220930.0,
      "step": 39145
    },
    {
      "entropy": 6.061733627319336,
      "epoch": 4.518176572417772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031340149942575443,
      "loss": 5.5251,
      "mean_token_accuracy": 0.2117103949189186,
      "num_tokens": 99234578.0,
      "step": 39150
    },
    {
      "entropy": 6.038629341125488,
      "epoch": 4.518753606462782,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000313360839872585,
      "loss": 5.5235,
      "mean_token_accuracy": 0.20940779894590378,
      "num_tokens": 99245823.0,
      "step": 39155
    },
    {
      "entropy": 6.0264537811279295,
      "epoch": 4.51933064050779,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031332017902916775,
      "loss": 5.4488,
      "mean_token_accuracy": 0.21725354045629502,
      "num_tokens": 99258386.0,
      "step": 39160
    },
    {
      "entropy": 6.009280586242676,
      "epoch": 4.519907674552798,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031327951689687027,
      "loss": 5.4926,
      "mean_token_accuracy": 0.21567383110523225,
      "num_tokens": 99270648.0,
      "step": 39165
    },
    {
      "entropy": 6.0526360988616945,
      "epoch": 4.520484708597808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031323885347706026,
      "loss": 5.5735,
      "mean_token_accuracy": 0.2092253088951111,
      "num_tokens": 99282678.0,
      "step": 39170
    },
    {
      "entropy": 6.074481534957886,
      "epoch": 4.521061742642816,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003131981887711055,
      "loss": 5.4815,
      "mean_token_accuracy": 0.21493065655231475,
      "num_tokens": 99295031.0,
      "step": 39175
    },
    {
      "entropy": 6.078620672225952,
      "epoch": 4.5216387766878245,
      "grad_norm": 0.90625,
      "learning_rate": 0.00031315752278037357,
      "loss": 5.5235,
      "mean_token_accuracy": 0.21108047366142274,
      "num_tokens": 99307144.0,
      "step": 39180
    },
    {
      "entropy": 6.004184246063232,
      "epoch": 4.522215810732833,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00031311685550623226,
      "loss": 5.4431,
      "mean_token_accuracy": 0.21694666445255278,
      "num_tokens": 99320763.0,
      "step": 39185
    },
    {
      "entropy": 6.00265097618103,
      "epoch": 4.522792844777842,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003130761869500495,
      "loss": 5.5141,
      "mean_token_accuracy": 0.21193920820951462,
      "num_tokens": 99332833.0,
      "step": 39190
    },
    {
      "entropy": 6.1322413921356205,
      "epoch": 4.523369878822851,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031303551711319317,
      "loss": 5.5824,
      "mean_token_accuracy": 0.20419093817472458,
      "num_tokens": 99345126.0,
      "step": 39195
    },
    {
      "entropy": 6.111878299713135,
      "epoch": 4.523946912867859,
      "grad_norm": 0.90625,
      "learning_rate": 0.00031299484599703103,
      "loss": 5.5811,
      "mean_token_accuracy": 0.21282886266708373,
      "num_tokens": 99358746.0,
      "step": 39200
    },
    {
      "entropy": 6.079991292953491,
      "epoch": 4.5245239469128675,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031295417360293113,
      "loss": 5.5893,
      "mean_token_accuracy": 0.2062515303492546,
      "num_tokens": 99370496.0,
      "step": 39205
    },
    {
      "entropy": 6.060204935073853,
      "epoch": 4.525100980957877,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003129134999322615,
      "loss": 5.5006,
      "mean_token_accuracy": 0.21328499019145966,
      "num_tokens": 99381817.0,
      "step": 39210
    },
    {
      "entropy": 6.092281866073608,
      "epoch": 4.525678015002885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003128728249863901,
      "loss": 5.5182,
      "mean_token_accuracy": 0.2151888594031334,
      "num_tokens": 99394141.0,
      "step": 39215
    },
    {
      "entropy": 6.0464883804321286,
      "epoch": 4.526255049047894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031283214876668504,
      "loss": 5.5979,
      "mean_token_accuracy": 0.1997549131512642,
      "num_tokens": 99405900.0,
      "step": 39220
    },
    {
      "entropy": 6.1588362693786625,
      "epoch": 4.526832083092902,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0003127914712745144,
      "loss": 5.5994,
      "mean_token_accuracy": 0.20104878395795822,
      "num_tokens": 99418621.0,
      "step": 39225
    },
    {
      "entropy": 6.04631814956665,
      "epoch": 4.527409117137911,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031275079251124645,
      "loss": 5.4854,
      "mean_token_accuracy": 0.2172614499926567,
      "num_tokens": 99430276.0,
      "step": 39230
    },
    {
      "entropy": 6.004769897460937,
      "epoch": 4.52798615118292,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003127101124782494,
      "loss": 5.4442,
      "mean_token_accuracy": 0.21340155154466628,
      "num_tokens": 99443320.0,
      "step": 39235
    },
    {
      "entropy": 6.0623210906982425,
      "epoch": 4.528563185227928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031266943117689135,
      "loss": 5.5454,
      "mean_token_accuracy": 0.2072186976671219,
      "num_tokens": 99455222.0,
      "step": 39240
    },
    {
      "entropy": 6.15461049079895,
      "epoch": 4.5291402192729375,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003126287486085407,
      "loss": 5.6168,
      "mean_token_accuracy": 0.20534174293279647,
      "num_tokens": 99467930.0,
      "step": 39245
    },
    {
      "entropy": 6.07463264465332,
      "epoch": 4.529717253317946,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003125880647745658,
      "loss": 5.4782,
      "mean_token_accuracy": 0.21575211435556413,
      "num_tokens": 99479752.0,
      "step": 39250
    },
    {
      "entropy": 6.016558647155762,
      "epoch": 4.530294287362954,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031254737967633506,
      "loss": 5.4606,
      "mean_token_accuracy": 0.21810299158096313,
      "num_tokens": 99491195.0,
      "step": 39255
    },
    {
      "entropy": 6.070026922225952,
      "epoch": 4.530871321407963,
      "grad_norm": 0.9375,
      "learning_rate": 0.00031250669331521686,
      "loss": 5.6088,
      "mean_token_accuracy": 0.2064017176628113,
      "num_tokens": 99504159.0,
      "step": 39260
    },
    {
      "entropy": 6.069992971420288,
      "epoch": 4.531448355452972,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031246600569257956,
      "loss": 5.4041,
      "mean_token_accuracy": 0.22241573184728622,
      "num_tokens": 99517395.0,
      "step": 39265
    },
    {
      "entropy": 6.070858955383301,
      "epoch": 4.5320253894979805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003124253168097918,
      "loss": 5.5183,
      "mean_token_accuracy": 0.2170992448925972,
      "num_tokens": 99530740.0,
      "step": 39270
    },
    {
      "entropy": 5.980552768707275,
      "epoch": 4.532602423542989,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00031238462666822217,
      "loss": 5.4307,
      "mean_token_accuracy": 0.22284302711486817,
      "num_tokens": 99542642.0,
      "step": 39275
    },
    {
      "entropy": 6.010508251190186,
      "epoch": 4.533179457587997,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003123439352692392,
      "loss": 5.4878,
      "mean_token_accuracy": 0.22187323421239852,
      "num_tokens": 99554584.0,
      "step": 39280
    },
    {
      "entropy": 6.0084432601928714,
      "epoch": 4.533756491633007,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003123032426142114,
      "loss": 5.5269,
      "mean_token_accuracy": 0.21809304058551787,
      "num_tokens": 99567546.0,
      "step": 39285
    },
    {
      "entropy": 6.043466949462891,
      "epoch": 4.534333525678015,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031226254870450767,
      "loss": 5.494,
      "mean_token_accuracy": 0.22058373391628266,
      "num_tokens": 99581154.0,
      "step": 39290
    },
    {
      "entropy": 6.087696218490601,
      "epoch": 4.5349105597230235,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031222185354149655,
      "loss": 5.568,
      "mean_token_accuracy": 0.2059185966849327,
      "num_tokens": 99594019.0,
      "step": 39295
    },
    {
      "entropy": 5.973210859298706,
      "epoch": 4.535487593768032,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031218115712654693,
      "loss": 5.5273,
      "mean_token_accuracy": 0.20920064747333528,
      "num_tokens": 99607529.0,
      "step": 39300
    },
    {
      "entropy": 6.0646647930145265,
      "epoch": 4.536064627813041,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003121404594610275,
      "loss": 5.5093,
      "mean_token_accuracy": 0.20972030907869338,
      "num_tokens": 99619853.0,
      "step": 39305
    },
    {
      "entropy": 6.095301580429077,
      "epoch": 4.53664166185805,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003120997605463073,
      "loss": 5.5133,
      "mean_token_accuracy": 0.2138122648000717,
      "num_tokens": 99632232.0,
      "step": 39310
    },
    {
      "entropy": 6.04727578163147,
      "epoch": 4.537218695903058,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031205906038375496,
      "loss": 5.5435,
      "mean_token_accuracy": 0.2152028948068619,
      "num_tokens": 99644680.0,
      "step": 39315
    },
    {
      "entropy": 6.027457904815674,
      "epoch": 4.537795729948067,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003120183589747396,
      "loss": 5.4443,
      "mean_token_accuracy": 0.21704814881086348,
      "num_tokens": 99657829.0,
      "step": 39320
    },
    {
      "entropy": 6.050464868545532,
      "epoch": 4.538372763993076,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00031197765632063007,
      "loss": 5.4536,
      "mean_token_accuracy": 0.21014560610055924,
      "num_tokens": 99669932.0,
      "step": 39325
    },
    {
      "entropy": 6.045565462112426,
      "epoch": 4.538949798038084,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031193695242279543,
      "loss": 5.464,
      "mean_token_accuracy": 0.22082182466983796,
      "num_tokens": 99683984.0,
      "step": 39330
    },
    {
      "entropy": 5.969738626480103,
      "epoch": 4.539526832083093,
      "grad_norm": 0.9375,
      "learning_rate": 0.00031189624728260475,
      "loss": 5.4174,
      "mean_token_accuracy": 0.22220737487077713,
      "num_tokens": 99697309.0,
      "step": 39335
    },
    {
      "entropy": 6.024098491668701,
      "epoch": 4.540103866128102,
      "grad_norm": 0.890625,
      "learning_rate": 0.00031185554090142715,
      "loss": 5.5312,
      "mean_token_accuracy": 0.21594950556755066,
      "num_tokens": 99711906.0,
      "step": 39340
    },
    {
      "entropy": 5.967256116867065,
      "epoch": 4.54068090017311,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00031181483328063177,
      "loss": 5.3855,
      "mean_token_accuracy": 0.21761398911476135,
      "num_tokens": 99725181.0,
      "step": 39345
    },
    {
      "entropy": 6.097359275817871,
      "epoch": 4.541257934218119,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00031177412442158763,
      "loss": 5.5091,
      "mean_token_accuracy": 0.20836860239505767,
      "num_tokens": 99737554.0,
      "step": 39350
    },
    {
      "entropy": 6.031353330612182,
      "epoch": 4.541834968263127,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031173341432566424,
      "loss": 5.5121,
      "mean_token_accuracy": 0.21167874336242676,
      "num_tokens": 99750490.0,
      "step": 39355
    },
    {
      "entropy": 6.024143934249878,
      "epoch": 4.5424120023081365,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031169270299423056,
      "loss": 5.477,
      "mean_token_accuracy": 0.211247019469738,
      "num_tokens": 99761954.0,
      "step": 39360
    },
    {
      "entropy": 5.9345029354095455,
      "epoch": 4.542989036353145,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003116519904286562,
      "loss": 5.403,
      "mean_token_accuracy": 0.22890340983867646,
      "num_tokens": 99775013.0,
      "step": 39365
    },
    {
      "entropy": 6.060037660598755,
      "epoch": 4.543566070398153,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003116112766303102,
      "loss": 5.4674,
      "mean_token_accuracy": 0.21734015196561812,
      "num_tokens": 99788109.0,
      "step": 39370
    },
    {
      "entropy": 6.007438993453979,
      "epoch": 4.544143104443162,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031157056160056217,
      "loss": 5.4587,
      "mean_token_accuracy": 0.21650406569242478,
      "num_tokens": 99802384.0,
      "step": 39375
    },
    {
      "entropy": 5.9942090034484865,
      "epoch": 4.544720138488171,
      "grad_norm": 1.0,
      "learning_rate": 0.0003115298453407815,
      "loss": 5.4231,
      "mean_token_accuracy": 0.21298206150531768,
      "num_tokens": 99814274.0,
      "step": 39380
    },
    {
      "entropy": 6.051714658737183,
      "epoch": 4.545297172533179,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003114891278523376,
      "loss": 5.5524,
      "mean_token_accuracy": 0.21761103272438048,
      "num_tokens": 99827202.0,
      "step": 39385
    },
    {
      "entropy": 5.937704229354859,
      "epoch": 4.545874206578188,
      "grad_norm": 1.765625,
      "learning_rate": 0.00031144840913659996,
      "loss": 5.3301,
      "mean_token_accuracy": 0.2386062115430832,
      "num_tokens": 99839985.0,
      "step": 39390
    },
    {
      "entropy": 6.057811594009399,
      "epoch": 4.546451240623197,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003114076891949382,
      "loss": 5.5903,
      "mean_token_accuracy": 0.19804765582084655,
      "num_tokens": 99853399.0,
      "step": 39395
    },
    {
      "entropy": 5.987566184997559,
      "epoch": 4.547028274668206,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00031136696802872195,
      "loss": 5.4652,
      "mean_token_accuracy": 0.22220586836338044,
      "num_tokens": 99866540.0,
      "step": 39400
    },
    {
      "entropy": 6.096220874786377,
      "epoch": 4.547605308713214,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031132624563932074,
      "loss": 5.5067,
      "mean_token_accuracy": 0.21738308072090148,
      "num_tokens": 99877708.0,
      "step": 39405
    },
    {
      "entropy": 5.959886932373047,
      "epoch": 4.548182342758222,
      "grad_norm": 1.0,
      "learning_rate": 0.00031128552202810424,
      "loss": 5.4683,
      "mean_token_accuracy": 0.2102986454963684,
      "num_tokens": 99890468.0,
      "step": 39410
    },
    {
      "entropy": 5.961117744445801,
      "epoch": 4.548759376803232,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031124479719644234,
      "loss": 5.4061,
      "mean_token_accuracy": 0.22255504727363587,
      "num_tokens": 99903158.0,
      "step": 39415
    },
    {
      "entropy": 6.066654396057129,
      "epoch": 4.54933641084824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003112040711457047,
      "loss": 5.5403,
      "mean_token_accuracy": 0.20981570929288865,
      "num_tokens": 99914826.0,
      "step": 39420
    },
    {
      "entropy": 6.071909236907959,
      "epoch": 4.549913444893249,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003111633438772611,
      "loss": 5.5671,
      "mean_token_accuracy": 0.20716703981161116,
      "num_tokens": 99927403.0,
      "step": 39425
    },
    {
      "entropy": 5.929892730712891,
      "epoch": 4.550490478938257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003111226153924813,
      "loss": 5.3692,
      "mean_token_accuracy": 0.22776197493076325,
      "num_tokens": 99939559.0,
      "step": 39430
    },
    {
      "entropy": 6.178140020370483,
      "epoch": 4.551067512983266,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003110818856927353,
      "loss": 5.5659,
      "mean_token_accuracy": 0.20862917900085448,
      "num_tokens": 99950944.0,
      "step": 39435
    },
    {
      "entropy": 6.027177047729492,
      "epoch": 4.551644547028275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003110411547793931,
      "loss": 5.4491,
      "mean_token_accuracy": 0.21976558715105057,
      "num_tokens": 99963608.0,
      "step": 39440
    },
    {
      "entropy": 6.057816457748413,
      "epoch": 4.552221581073283,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003110004226538245,
      "loss": 5.6431,
      "mean_token_accuracy": 0.20254571586847306,
      "num_tokens": 99976188.0,
      "step": 39445
    },
    {
      "entropy": 6.089889287948608,
      "epoch": 4.5527986151182915,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003109596893173995,
      "loss": 5.5653,
      "mean_token_accuracy": 0.20962436199188234,
      "num_tokens": 99989166.0,
      "step": 39450
    },
    {
      "entropy": 6.0351152420043945,
      "epoch": 4.553375649163301,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003109189547714883,
      "loss": 5.4547,
      "mean_token_accuracy": 0.2187407210469246,
      "num_tokens": 100002545.0,
      "step": 39455
    },
    {
      "entropy": 6.0733642578125,
      "epoch": 4.553952683208309,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003108782190174609,
      "loss": 5.5309,
      "mean_token_accuracy": 0.21088054776191711,
      "num_tokens": 100016268.0,
      "step": 39460
    },
    {
      "entropy": 6.134630966186523,
      "epoch": 4.554529717253318,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003108374820566874,
      "loss": 5.5925,
      "mean_token_accuracy": 0.20737059116363527,
      "num_tokens": 100029594.0,
      "step": 39465
    },
    {
      "entropy": 6.086150646209717,
      "epoch": 4.555106751298327,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000310796743890538,
      "loss": 5.4804,
      "mean_token_accuracy": 0.21270056962966918,
      "num_tokens": 100042207.0,
      "step": 39470
    },
    {
      "entropy": 6.070754623413086,
      "epoch": 4.555683785343335,
      "grad_norm": 1.0,
      "learning_rate": 0.00031075600452038285,
      "loss": 5.4744,
      "mean_token_accuracy": 0.22236666977405548,
      "num_tokens": 100055033.0,
      "step": 39475
    },
    {
      "entropy": 5.938803243637085,
      "epoch": 4.556260819388344,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00031071526394759233,
      "loss": 5.4662,
      "mean_token_accuracy": 0.21735987663269044,
      "num_tokens": 100067595.0,
      "step": 39480
    },
    {
      "entropy": 6.049497938156128,
      "epoch": 4.556837853433352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003106745221735366,
      "loss": 5.495,
      "mean_token_accuracy": 0.22013844400644303,
      "num_tokens": 100079945.0,
      "step": 39485
    },
    {
      "entropy": 6.106379795074463,
      "epoch": 4.5574148874783615,
      "grad_norm": 0.984375,
      "learning_rate": 0.000310633779199586,
      "loss": 5.5671,
      "mean_token_accuracy": 0.20960116982460023,
      "num_tokens": 100091942.0,
      "step": 39490
    },
    {
      "entropy": 6.172228717803955,
      "epoch": 4.55799192152337,
      "grad_norm": 0.96875,
      "learning_rate": 0.000310593035027111,
      "loss": 5.6157,
      "mean_token_accuracy": 0.2081824481487274,
      "num_tokens": 100104285.0,
      "step": 39495
    },
    {
      "entropy": 6.009035396575928,
      "epoch": 4.558568955568378,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031055228965748194,
      "loss": 5.5328,
      "mean_token_accuracy": 0.21313058584928513,
      "num_tokens": 100117798.0,
      "step": 39500
    },
    {
      "entropy": 6.073286437988282,
      "epoch": 4.559145989613388,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0003105115430920693,
      "loss": 5.518,
      "mean_token_accuracy": 0.21984363794326783,
      "num_tokens": 100131338.0,
      "step": 39505
    },
    {
      "entropy": 6.075874900817871,
      "epoch": 4.559723023658396,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031047079533224356,
      "loss": 5.5307,
      "mean_token_accuracy": 0.21363597214221955,
      "num_tokens": 100143978.0,
      "step": 39510
    },
    {
      "entropy": 6.082835721969604,
      "epoch": 4.5603000577034045,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003104300463793751,
      "loss": 5.4747,
      "mean_token_accuracy": 0.21285403668880462,
      "num_tokens": 100157529.0,
      "step": 39515
    },
    {
      "entropy": 6.051505661010742,
      "epoch": 4.560877091748413,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031038929623483473,
      "loss": 5.5091,
      "mean_token_accuracy": 0.22028353661298752,
      "num_tokens": 100170502.0,
      "step": 39520
    },
    {
      "entropy": 6.0016210079193115,
      "epoch": 4.561454125793421,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031034854489999297,
      "loss": 5.4024,
      "mean_token_accuracy": 0.2216365307569504,
      "num_tokens": 100183086.0,
      "step": 39525
    },
    {
      "entropy": 6.046122407913208,
      "epoch": 4.562031159838431,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003103077923762204,
      "loss": 5.5438,
      "mean_token_accuracy": 0.21107827872037888,
      "num_tokens": 100195237.0,
      "step": 39530
    },
    {
      "entropy": 6.033243989944458,
      "epoch": 4.562608193883439,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00031026703866488784,
      "loss": 5.545,
      "mean_token_accuracy": 0.21325092166662216,
      "num_tokens": 100208586.0,
      "step": 39535
    },
    {
      "entropy": 6.03102068901062,
      "epoch": 4.5631852279284475,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031022628376736587,
      "loss": 5.4608,
      "mean_token_accuracy": 0.221030892431736,
      "num_tokens": 100221588.0,
      "step": 39540
    },
    {
      "entropy": 6.089952039718628,
      "epoch": 4.563762261973457,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003101855276850254,
      "loss": 5.5212,
      "mean_token_accuracy": 0.21016396284103395,
      "num_tokens": 100234978.0,
      "step": 39545
    },
    {
      "entropy": 6.003946018218994,
      "epoch": 4.564339296018465,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031014477041923724,
      "loss": 5.4236,
      "mean_token_accuracy": 0.22133288383483887,
      "num_tokens": 100249248.0,
      "step": 39550
    },
    {
      "entropy": 6.041824102401733,
      "epoch": 4.564916330063474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003101040119713721,
      "loss": 5.5975,
      "mean_token_accuracy": 0.2022898405790329,
      "num_tokens": 100263047.0,
      "step": 39555
    },
    {
      "entropy": 6.171460723876953,
      "epoch": 4.565493364108482,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031006325234280097,
      "loss": 5.6387,
      "mean_token_accuracy": 0.20130623281002044,
      "num_tokens": 100276220.0,
      "step": 39560
    },
    {
      "entropy": 6.132236003875732,
      "epoch": 4.566070398153491,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031002249153489473,
      "loss": 5.5486,
      "mean_token_accuracy": 0.2062343180179596,
      "num_tokens": 100287946.0,
      "step": 39565
    },
    {
      "entropy": 5.989096307754517,
      "epoch": 4.5666474321985,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00030998172954902446,
      "loss": 5.4275,
      "mean_token_accuracy": 0.22464457750320435,
      "num_tokens": 100300023.0,
      "step": 39570
    },
    {
      "entropy": 5.866420125961303,
      "epoch": 4.567224466243508,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003099409663865611,
      "loss": 5.3013,
      "mean_token_accuracy": 0.23131477534770967,
      "num_tokens": 100312686.0,
      "step": 39575
    },
    {
      "entropy": 6.02342586517334,
      "epoch": 4.5678015002885175,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00030990020204887557,
      "loss": 5.5141,
      "mean_token_accuracy": 0.21159286201000213,
      "num_tokens": 100325833.0,
      "step": 39580
    },
    {
      "entropy": 5.975371789932251,
      "epoch": 4.568378534333526,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003098594365373392,
      "loss": 5.4306,
      "mean_token_accuracy": 0.22996634244918823,
      "num_tokens": 100339386.0,
      "step": 39585
    },
    {
      "entropy": 6.05181360244751,
      "epoch": 4.568955568378534,
      "grad_norm": 1.0,
      "learning_rate": 0.00030981866985332305,
      "loss": 5.4862,
      "mean_token_accuracy": 0.20718948692083358,
      "num_tokens": 100350812.0,
      "step": 39590
    },
    {
      "entropy": 6.129055595397949,
      "epoch": 4.569532602423543,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003097779019981981,
      "loss": 5.5787,
      "mean_token_accuracy": 0.20898860692977905,
      "num_tokens": 100363576.0,
      "step": 39595
    },
    {
      "entropy": 5.9428795337677,
      "epoch": 4.570109636468551,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003097371329733357,
      "loss": 5.4685,
      "mean_token_accuracy": 0.21806725561618806,
      "num_tokens": 100376137.0,
      "step": 39600
    },
    {
      "entropy": 5.9458106517791744,
      "epoch": 4.5706866705135605,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00030969636278010715,
      "loss": 5.3976,
      "mean_token_accuracy": 0.2217326581478119,
      "num_tokens": 100389474.0,
      "step": 39605
    },
    {
      "entropy": 6.052435636520386,
      "epoch": 4.571263704558569,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003096555914198837,
      "loss": 5.5398,
      "mean_token_accuracy": 0.2088010311126709,
      "num_tokens": 100401562.0,
      "step": 39610
    },
    {
      "entropy": 6.070488452911377,
      "epoch": 4.571840738603577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003096148188940366,
      "loss": 5.4801,
      "mean_token_accuracy": 0.21927041113376616,
      "num_tokens": 100413711.0,
      "step": 39615
    },
    {
      "entropy": 6.017584371566772,
      "epoch": 4.572417772648587,
      "grad_norm": 1.0,
      "learning_rate": 0.0003095740452039372,
      "loss": 5.4622,
      "mean_token_accuracy": 0.21482844203710555,
      "num_tokens": 100425977.0,
      "step": 39620
    },
    {
      "entropy": 6.029469776153564,
      "epoch": 4.572994806693595,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000309533270350957,
      "loss": 5.5211,
      "mean_token_accuracy": 0.2164745584130287,
      "num_tokens": 100439167.0,
      "step": 39625
    },
    {
      "entropy": 6.083765697479248,
      "epoch": 4.5735718407386035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030949249433646746,
      "loss": 5.663,
      "mean_token_accuracy": 0.20389586240053176,
      "num_tokens": 100451336.0,
      "step": 39630
    },
    {
      "entropy": 6.1409605026245115,
      "epoch": 4.574148874783612,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030945171716184003,
      "loss": 5.5238,
      "mean_token_accuracy": 0.21577880829572677,
      "num_tokens": 100463401.0,
      "step": 39635
    },
    {
      "entropy": 6.055810546875,
      "epoch": 4.574725908828621,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00030941093882844606,
      "loss": 5.5762,
      "mean_token_accuracy": 0.21321677565574645,
      "num_tokens": 100476594.0,
      "step": 39640
    },
    {
      "entropy": 6.093710708618164,
      "epoch": 4.57530294287363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003093701593376574,
      "loss": 5.5835,
      "mean_token_accuracy": 0.2159482792019844,
      "num_tokens": 100489140.0,
      "step": 39645
    },
    {
      "entropy": 6.105528020858765,
      "epoch": 4.575879976918638,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030932937869084545,
      "loss": 5.5576,
      "mean_token_accuracy": 0.21341225802898406,
      "num_tokens": 100501831.0,
      "step": 39650
    },
    {
      "entropy": 6.091807174682617,
      "epoch": 4.576457010963647,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003092885968893819,
      "loss": 5.4983,
      "mean_token_accuracy": 0.2166688397526741,
      "num_tokens": 100513814.0,
      "step": 39655
    },
    {
      "entropy": 6.077985763549805,
      "epoch": 4.577034045008656,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030924781393463835,
      "loss": 5.5578,
      "mean_token_accuracy": 0.20960844308137894,
      "num_tokens": 100525913.0,
      "step": 39660
    },
    {
      "entropy": 6.158643293380737,
      "epoch": 4.577611079053664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030920702982798663,
      "loss": 5.6545,
      "mean_token_accuracy": 0.20176791548728942,
      "num_tokens": 100538802.0,
      "step": 39665
    },
    {
      "entropy": 6.098426246643067,
      "epoch": 4.578188113098673,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00030916624457079855,
      "loss": 5.5654,
      "mean_token_accuracy": 0.20966840535402298,
      "num_tokens": 100552839.0,
      "step": 39670
    },
    {
      "entropy": 6.100569343566894,
      "epoch": 4.578765147143681,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003091254581644458,
      "loss": 5.5344,
      "mean_token_accuracy": 0.21703532934188843,
      "num_tokens": 100565642.0,
      "step": 39675
    },
    {
      "entropy": 6.055329179763794,
      "epoch": 4.57934218118869,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030908467061030004,
      "loss": 5.4573,
      "mean_token_accuracy": 0.21789049059152604,
      "num_tokens": 100577217.0,
      "step": 39680
    },
    {
      "entropy": 6.088983106613159,
      "epoch": 4.579919215233699,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003090438819097335,
      "loss": 5.5893,
      "mean_token_accuracy": 0.21278050541877747,
      "num_tokens": 100589405.0,
      "step": 39685
    },
    {
      "entropy": 6.055174207687378,
      "epoch": 4.580496249278707,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003090030920641178,
      "loss": 5.5525,
      "mean_token_accuracy": 0.20813312977552414,
      "num_tokens": 100601739.0,
      "step": 39690
    },
    {
      "entropy": 6.162087392807007,
      "epoch": 4.581073283323716,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030896230107482504,
      "loss": 5.6462,
      "mean_token_accuracy": 0.20202667713165284,
      "num_tokens": 100615077.0,
      "step": 39695
    },
    {
      "entropy": 5.974533557891846,
      "epoch": 4.581650317368725,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00030892150894322717,
      "loss": 5.4887,
      "mean_token_accuracy": 0.2161990523338318,
      "num_tokens": 100628352.0,
      "step": 39700
    },
    {
      "entropy": 6.023526668548584,
      "epoch": 4.582227351413733,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030888071567069614,
      "loss": 5.5737,
      "mean_token_accuracy": 0.2099379912018776,
      "num_tokens": 100640826.0,
      "step": 39705
    },
    {
      "entropy": 6.041655206680298,
      "epoch": 4.582804385458742,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030883992125860406,
      "loss": 5.5127,
      "mean_token_accuracy": 0.21921360045671462,
      "num_tokens": 100652102.0,
      "step": 39710
    },
    {
      "entropy": 6.172430992126465,
      "epoch": 4.583381419503751,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003087991257083231,
      "loss": 5.6361,
      "mean_token_accuracy": 0.20719803720712662,
      "num_tokens": 100665722.0,
      "step": 39715
    },
    {
      "entropy": 6.005380058288575,
      "epoch": 4.583958453548759,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003087583290212253,
      "loss": 5.4569,
      "mean_token_accuracy": 0.2170966923236847,
      "num_tokens": 100677910.0,
      "step": 39720
    },
    {
      "entropy": 6.049957370758056,
      "epoch": 4.584535487593768,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003087175311986828,
      "loss": 5.5447,
      "mean_token_accuracy": 0.21826043725013733,
      "num_tokens": 100690187.0,
      "step": 39725
    },
    {
      "entropy": 5.901098299026489,
      "epoch": 4.585112521638777,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030867673224206806,
      "loss": 5.365,
      "mean_token_accuracy": 0.23597146272659303,
      "num_tokens": 100702942.0,
      "step": 39730
    },
    {
      "entropy": 6.058048057556152,
      "epoch": 4.585689555683786,
      "grad_norm": 1.0,
      "learning_rate": 0.000308635932152753,
      "loss": 5.5214,
      "mean_token_accuracy": 0.21089319437742232,
      "num_tokens": 100716716.0,
      "step": 39735
    },
    {
      "entropy": 6.122119188308716,
      "epoch": 4.586266589728794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00030859513093211015,
      "loss": 5.6285,
      "mean_token_accuracy": 0.21020228713750838,
      "num_tokens": 100731017.0,
      "step": 39740
    },
    {
      "entropy": 6.066273880004883,
      "epoch": 4.586843623773802,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003085543285815117,
      "loss": 5.4738,
      "mean_token_accuracy": 0.21074859648942948,
      "num_tokens": 100743546.0,
      "step": 39745
    },
    {
      "entropy": 5.9989426612854,
      "epoch": 4.587420657818812,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00030851352510233017,
      "loss": 5.4501,
      "mean_token_accuracy": 0.21823455393314362,
      "num_tokens": 100756668.0,
      "step": 39750
    },
    {
      "entropy": 6.023809289932251,
      "epoch": 4.58799769186382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003084727204959378,
      "loss": 5.525,
      "mean_token_accuracy": 0.21621302813291549,
      "num_tokens": 100769491.0,
      "step": 39755
    },
    {
      "entropy": 6.068006563186645,
      "epoch": 4.5885747259088285,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003084319147637072,
      "loss": 5.531,
      "mean_token_accuracy": 0.20983080416917801,
      "num_tokens": 100784434.0,
      "step": 39760
    },
    {
      "entropy": 6.05798454284668,
      "epoch": 4.589151759953837,
      "grad_norm": 0.875,
      "learning_rate": 0.00030839110790701057,
      "loss": 5.5055,
      "mean_token_accuracy": 0.20944105833768845,
      "num_tokens": 100797638.0,
      "step": 39765
    },
    {
      "entropy": 6.047077178955078,
      "epoch": 4.589728793998846,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030835029992722074,
      "loss": 5.505,
      "mean_token_accuracy": 0.21706412136554717,
      "num_tokens": 100810542.0,
      "step": 39770
    },
    {
      "entropy": 6.017255973815918,
      "epoch": 4.590305828043855,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030830949082571015,
      "loss": 5.4978,
      "mean_token_accuracy": 0.21799633353948594,
      "num_tokens": 100822476.0,
      "step": 39775
    },
    {
      "entropy": 6.054379940032959,
      "epoch": 4.590882862088863,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003082686806038513,
      "loss": 5.4226,
      "mean_token_accuracy": 0.2249370813369751,
      "num_tokens": 100834990.0,
      "step": 39780
    },
    {
      "entropy": 6.054190635681152,
      "epoch": 4.5914598961338715,
      "grad_norm": 1.0,
      "learning_rate": 0.000308227869263017,
      "loss": 5.5095,
      "mean_token_accuracy": 0.21247874498367308,
      "num_tokens": 100846308.0,
      "step": 39785
    },
    {
      "entropy": 6.019290447235107,
      "epoch": 4.592036930178881,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003081870568045798,
      "loss": 5.5334,
      "mean_token_accuracy": 0.20970378667116166,
      "num_tokens": 100860326.0,
      "step": 39790
    },
    {
      "entropy": 5.990709209442139,
      "epoch": 4.592613964223889,
      "grad_norm": 0.921875,
      "learning_rate": 0.00030814624322991237,
      "loss": 5.5024,
      "mean_token_accuracy": 0.21360168159008025,
      "num_tokens": 100872304.0,
      "step": 39795
    },
    {
      "entropy": 6.073654365539551,
      "epoch": 4.593190998268898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003081054285403875,
      "loss": 5.4861,
      "mean_token_accuracy": 0.21650294661521913,
      "num_tokens": 100885387.0,
      "step": 39800
    },
    {
      "entropy": 6.08500337600708,
      "epoch": 4.593768032313907,
      "grad_norm": 0.96875,
      "learning_rate": 0.000308064612737378,
      "loss": 5.5446,
      "mean_token_accuracy": 0.21620493829250337,
      "num_tokens": 100898457.0,
      "step": 39805
    },
    {
      "entropy": 6.044233942031861,
      "epoch": 4.594345066358915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003080237958222567,
      "loss": 5.459,
      "mean_token_accuracy": 0.2133478432893753,
      "num_tokens": 100911419.0,
      "step": 39810
    },
    {
      "entropy": 5.9902441024780275,
      "epoch": 4.594922100403924,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003079829777963965,
      "loss": 5.4638,
      "mean_token_accuracy": 0.21956752687692643,
      "num_tokens": 100924195.0,
      "step": 39815
    },
    {
      "entropy": 6.01695237159729,
      "epoch": 4.595499134448932,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030794215866117015,
      "loss": 5.4651,
      "mean_token_accuracy": 0.2226399302482605,
      "num_tokens": 100936699.0,
      "step": 39820
    },
    {
      "entropy": 5.903493928909302,
      "epoch": 4.5960761684939415,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00030790133841795064,
      "loss": 5.36,
      "mean_token_accuracy": 0.23130894601345062,
      "num_tokens": 100949381.0,
      "step": 39825
    },
    {
      "entropy": 6.0256048202514645,
      "epoch": 4.59665320253895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030786051706811095,
      "loss": 5.4577,
      "mean_token_accuracy": 0.21627974212169648,
      "num_tokens": 100963139.0,
      "step": 39830
    },
    {
      "entropy": 6.092518377304077,
      "epoch": 4.597230236583958,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003078196946130242,
      "loss": 5.5805,
      "mean_token_accuracy": 0.20136864483356476,
      "num_tokens": 100976723.0,
      "step": 39835
    },
    {
      "entropy": 6.006163597106934,
      "epoch": 4.597807270628967,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003077788710540632,
      "loss": 5.5018,
      "mean_token_accuracy": 0.21340110898017883,
      "num_tokens": 100989563.0,
      "step": 39840
    },
    {
      "entropy": 6.088361072540283,
      "epoch": 4.598384304673976,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030773804639260124,
      "loss": 5.5261,
      "mean_token_accuracy": 0.2132933869957924,
      "num_tokens": 101002182.0,
      "step": 39845
    },
    {
      "entropy": 6.10531702041626,
      "epoch": 4.5989613387189845,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0003076972206300113,
      "loss": 5.539,
      "mean_token_accuracy": 0.20818218737840652,
      "num_tokens": 101015332.0,
      "step": 39850
    },
    {
      "entropy": 5.935640001296997,
      "epoch": 4.599538372763993,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003076563937676666,
      "loss": 5.3636,
      "mean_token_accuracy": 0.21997026354074478,
      "num_tokens": 101027138.0,
      "step": 39855
    },
    {
      "entropy": 5.979724788665772,
      "epoch": 4.600115406809001,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003076155658069403,
      "loss": 5.4803,
      "mean_token_accuracy": 0.21426504403352736,
      "num_tokens": 101039843.0,
      "step": 39860
    },
    {
      "entropy": 6.104697799682617,
      "epoch": 4.600692440854011,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003075747367492056,
      "loss": 5.5811,
      "mean_token_accuracy": 0.21186555922031403,
      "num_tokens": 101052123.0,
      "step": 39865
    },
    {
      "entropy": 5.999017763137817,
      "epoch": 4.601269474899019,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003075339065958359,
      "loss": 5.4436,
      "mean_token_accuracy": 0.21938565522432327,
      "num_tokens": 101065391.0,
      "step": 39870
    },
    {
      "entropy": 6.067542982101441,
      "epoch": 4.6018465089440275,
      "grad_norm": 1.0,
      "learning_rate": 0.00030749307534820445,
      "loss": 5.4892,
      "mean_token_accuracy": 0.21351806819438934,
      "num_tokens": 101076499.0,
      "step": 39875
    },
    {
      "entropy": 6.085097885131836,
      "epoch": 4.602423542989037,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030745224300768445,
      "loss": 5.5694,
      "mean_token_accuracy": 0.21838047802448274,
      "num_tokens": 101090272.0,
      "step": 39880
    },
    {
      "entropy": 6.068652582168579,
      "epoch": 4.603000577034045,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030741140957564944,
      "loss": 5.5351,
      "mean_token_accuracy": 0.2065625548362732,
      "num_tokens": 101102656.0,
      "step": 39885
    },
    {
      "entropy": 5.951930809020996,
      "epoch": 4.603577611079054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003073705750534727,
      "loss": 5.3901,
      "mean_token_accuracy": 0.22713368386030197,
      "num_tokens": 101115081.0,
      "step": 39890
    },
    {
      "entropy": 6.0828031539917,
      "epoch": 4.604154645124062,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003073297394425278,
      "loss": 5.5927,
      "mean_token_accuracy": 0.21093661934137345,
      "num_tokens": 101126959.0,
      "step": 39895
    },
    {
      "entropy": 6.097988939285278,
      "epoch": 4.604731679169071,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003072889027441881,
      "loss": 5.4927,
      "mean_token_accuracy": 0.21615091115236282,
      "num_tokens": 101139483.0,
      "step": 39900
    },
    {
      "entropy": 6.056278038024902,
      "epoch": 4.60530871321408,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003072480649598274,
      "loss": 5.5813,
      "mean_token_accuracy": 0.20850392282009125,
      "num_tokens": 101152248.0,
      "step": 39905
    },
    {
      "entropy": 6.082311487197876,
      "epoch": 4.605885747259088,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003072072260908188,
      "loss": 5.5373,
      "mean_token_accuracy": 0.2132813647389412,
      "num_tokens": 101164969.0,
      "step": 39910
    },
    {
      "entropy": 5.982616806030274,
      "epoch": 4.606462781304097,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030716638613853623,
      "loss": 5.3876,
      "mean_token_accuracy": 0.22186468690633773,
      "num_tokens": 101179328.0,
      "step": 39915
    },
    {
      "entropy": 6.002566337585449,
      "epoch": 4.607039815349106,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003071255451043533,
      "loss": 5.5026,
      "mean_token_accuracy": 0.21773656755685805,
      "num_tokens": 101193068.0,
      "step": 39920
    },
    {
      "entropy": 6.042377567291259,
      "epoch": 4.607616849394114,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003070847029896436,
      "loss": 5.5407,
      "mean_token_accuracy": 0.2098624274134636,
      "num_tokens": 101204619.0,
      "step": 39925
    },
    {
      "entropy": 6.002839469909668,
      "epoch": 4.608193883439123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003070438597957808,
      "loss": 5.4674,
      "mean_token_accuracy": 0.22082670032978058,
      "num_tokens": 101217105.0,
      "step": 39930
    },
    {
      "entropy": 5.99722728729248,
      "epoch": 4.608770917484131,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030700301552413875,
      "loss": 5.432,
      "mean_token_accuracy": 0.22525749504566192,
      "num_tokens": 101229354.0,
      "step": 39935
    },
    {
      "entropy": 5.989676141738892,
      "epoch": 4.6093479515291405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003069621701760912,
      "loss": 5.4841,
      "mean_token_accuracy": 0.2130702406167984,
      "num_tokens": 101242027.0,
      "step": 39940
    },
    {
      "entropy": 6.0434986591339115,
      "epoch": 4.609924985574149,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003069213237530118,
      "loss": 5.485,
      "mean_token_accuracy": 0.21681522727012634,
      "num_tokens": 101254494.0,
      "step": 39945
    },
    {
      "entropy": 6.088327884674072,
      "epoch": 4.610502019619157,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030688047625627464,
      "loss": 5.4885,
      "mean_token_accuracy": 0.21699120700359345,
      "num_tokens": 101266951.0,
      "step": 39950
    },
    {
      "entropy": 6.107674312591553,
      "epoch": 4.611079053664167,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003068396276872534,
      "loss": 5.5382,
      "mean_token_accuracy": 0.21126474142074586,
      "num_tokens": 101280983.0,
      "step": 39955
    },
    {
      "entropy": 6.027893781661987,
      "epoch": 4.611656087709175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003067987780473222,
      "loss": 5.4854,
      "mean_token_accuracy": 0.21899836659431457,
      "num_tokens": 101294107.0,
      "step": 39960
    },
    {
      "entropy": 6.105498504638672,
      "epoch": 4.612233121754183,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030675792733785484,
      "loss": 5.5955,
      "mean_token_accuracy": 0.20831957459449768,
      "num_tokens": 101307313.0,
      "step": 39965
    },
    {
      "entropy": 6.05607237815857,
      "epoch": 4.612810155799192,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003067170755602254,
      "loss": 5.4696,
      "mean_token_accuracy": 0.20868636220693587,
      "num_tokens": 101320437.0,
      "step": 39970
    },
    {
      "entropy": 6.040258884429932,
      "epoch": 4.613387189844201,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00030667622271580785,
      "loss": 5.5135,
      "mean_token_accuracy": 0.21233072876930237,
      "num_tokens": 101333798.0,
      "step": 39975
    },
    {
      "entropy": 5.98102560043335,
      "epoch": 4.61396422388921,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003066353688059763,
      "loss": 5.5177,
      "mean_token_accuracy": 0.2155057445168495,
      "num_tokens": 101346355.0,
      "step": 39980
    },
    {
      "entropy": 6.022253847122192,
      "epoch": 4.614541257934218,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003065945138321049,
      "loss": 5.4791,
      "mean_token_accuracy": 0.22370898723602295,
      "num_tokens": 101359882.0,
      "step": 39985
    },
    {
      "entropy": 6.06699538230896,
      "epoch": 4.615118291979226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030655365779556773,
      "loss": 5.5047,
      "mean_token_accuracy": 0.2193661227822304,
      "num_tokens": 101371009.0,
      "step": 39990
    },
    {
      "entropy": 6.093859767913818,
      "epoch": 4.615695326024236,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003065128006977388,
      "loss": 5.5473,
      "mean_token_accuracy": 0.2165049284696579,
      "num_tokens": 101384570.0,
      "step": 39995
    },
    {
      "entropy": 5.958289337158203,
      "epoch": 4.616272360069244,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030647194253999256,
      "loss": 5.4093,
      "mean_token_accuracy": 0.2202360600233078,
      "num_tokens": 101398104.0,
      "step": 40000
    },
    {
      "entropy": 6.111950206756592,
      "epoch": 4.6168493941142525,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030643108332370316,
      "loss": 5.5213,
      "mean_token_accuracy": 0.21204749047756194,
      "num_tokens": 101409790.0,
      "step": 40005
    },
    {
      "entropy": 6.075001811981201,
      "epoch": 4.617426428159261,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003063902230502449,
      "loss": 5.566,
      "mean_token_accuracy": 0.20609041303396225,
      "num_tokens": 101421350.0,
      "step": 40010
    },
    {
      "entropy": 6.0637177467346195,
      "epoch": 4.61800346220427,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030634936172099206,
      "loss": 5.5843,
      "mean_token_accuracy": 0.20862679928541183,
      "num_tokens": 101432423.0,
      "step": 40015
    },
    {
      "entropy": 6.063412141799927,
      "epoch": 4.618580496249279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030630849933731903,
      "loss": 5.5406,
      "mean_token_accuracy": 0.21451467275619507,
      "num_tokens": 101443706.0,
      "step": 40020
    },
    {
      "entropy": 5.93739128112793,
      "epoch": 4.619157530294287,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003062676359006001,
      "loss": 5.3475,
      "mean_token_accuracy": 0.22942973524332047,
      "num_tokens": 101456524.0,
      "step": 40025
    },
    {
      "entropy": 6.048275899887085,
      "epoch": 4.619734564339296,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030622677141220987,
      "loss": 5.4492,
      "mean_token_accuracy": 0.22029660642147064,
      "num_tokens": 101469623.0,
      "step": 40030
    },
    {
      "entropy": 6.051504135131836,
      "epoch": 4.620311598384305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003061859058735227,
      "loss": 5.5085,
      "mean_token_accuracy": 0.211344350874424,
      "num_tokens": 101483109.0,
      "step": 40035
    },
    {
      "entropy": 6.009643936157227,
      "epoch": 4.620888632429313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000306145039285913,
      "loss": 5.4181,
      "mean_token_accuracy": 0.21842905580997468,
      "num_tokens": 101494181.0,
      "step": 40040
    },
    {
      "entropy": 6.048507738113403,
      "epoch": 4.621465666474322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030610417165075537,
      "loss": 5.4857,
      "mean_token_accuracy": 0.2160505011677742,
      "num_tokens": 101506221.0,
      "step": 40045
    },
    {
      "entropy": 6.060896921157837,
      "epoch": 4.622042700519331,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003060633029694245,
      "loss": 5.5689,
      "mean_token_accuracy": 0.21058838963508605,
      "num_tokens": 101519953.0,
      "step": 40050
    },
    {
      "entropy": 6.114691209793091,
      "epoch": 4.622619734564339,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030602243324329473,
      "loss": 5.5082,
      "mean_token_accuracy": 0.21910332888364792,
      "num_tokens": 101532720.0,
      "step": 40055
    },
    {
      "entropy": 6.052140951156616,
      "epoch": 4.623196768609348,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030598156247374085,
      "loss": 5.5625,
      "mean_token_accuracy": 0.20924121141433716,
      "num_tokens": 101545205.0,
      "step": 40060
    },
    {
      "entropy": 6.13967981338501,
      "epoch": 4.623773802654356,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030594069066213755,
      "loss": 5.6257,
      "mean_token_accuracy": 0.20198575556278228,
      "num_tokens": 101559373.0,
      "step": 40065
    },
    {
      "entropy": 6.181623268127441,
      "epoch": 4.6243508366993655,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003058998178098595,
      "loss": 5.6248,
      "mean_token_accuracy": 0.20573222637176514,
      "num_tokens": 101571162.0,
      "step": 40070
    },
    {
      "entropy": 6.122658681869507,
      "epoch": 4.624927870744374,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003058589439182813,
      "loss": 5.6034,
      "mean_token_accuracy": 0.20781558603048325,
      "num_tokens": 101583182.0,
      "step": 40075
    },
    {
      "entropy": 6.0718262672424315,
      "epoch": 4.625504904789382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030581806898877803,
      "loss": 5.5186,
      "mean_token_accuracy": 0.2113463893532753,
      "num_tokens": 101596332.0,
      "step": 40080
    },
    {
      "entropy": 6.070315790176392,
      "epoch": 4.626081938834391,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030577719302272424,
      "loss": 5.4949,
      "mean_token_accuracy": 0.21204714328050614,
      "num_tokens": 101607455.0,
      "step": 40085
    },
    {
      "entropy": 5.961989545822144,
      "epoch": 4.6266589728794,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003057363160214948,
      "loss": 5.4431,
      "mean_token_accuracy": 0.22714711725711823,
      "num_tokens": 101620331.0,
      "step": 40090
    },
    {
      "entropy": 6.012278985977173,
      "epoch": 4.6272360069244085,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003056954379864648,
      "loss": 5.5276,
      "mean_token_accuracy": 0.20908010005950928,
      "num_tokens": 101632817.0,
      "step": 40095
    },
    {
      "entropy": 5.9790740489959715,
      "epoch": 4.627813040969417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030565455891900877,
      "loss": 5.4614,
      "mean_token_accuracy": 0.21819052398204802,
      "num_tokens": 101644677.0,
      "step": 40100
    },
    {
      "entropy": 6.110869121551514,
      "epoch": 4.628390075014426,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030561367882050197,
      "loss": 5.6214,
      "mean_token_accuracy": 0.2094668596982956,
      "num_tokens": 101655835.0,
      "step": 40105
    },
    {
      "entropy": 6.145654582977295,
      "epoch": 4.628967109059435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003055727976923193,
      "loss": 5.6142,
      "mean_token_accuracy": 0.20257598310709,
      "num_tokens": 101668373.0,
      "step": 40110
    },
    {
      "entropy": 6.056234979629517,
      "epoch": 4.629544143104443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030553191553583584,
      "loss": 5.5188,
      "mean_token_accuracy": 0.20837038308382033,
      "num_tokens": 101680711.0,
      "step": 40115
    },
    {
      "entropy": 6.049994277954101,
      "epoch": 4.6301211771494515,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003054910323524265,
      "loss": 5.4742,
      "mean_token_accuracy": 0.22557793706655502,
      "num_tokens": 101693795.0,
      "step": 40120
    },
    {
      "entropy": 5.994244766235352,
      "epoch": 4.630698211194461,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003054501481434664,
      "loss": 5.4556,
      "mean_token_accuracy": 0.2223817527294159,
      "num_tokens": 101707223.0,
      "step": 40125
    },
    {
      "entropy": 6.081955194473267,
      "epoch": 4.631275245239469,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003054092629103307,
      "loss": 5.5473,
      "mean_token_accuracy": 0.2087576672434807,
      "num_tokens": 101719325.0,
      "step": 40130
    },
    {
      "entropy": 6.048444652557373,
      "epoch": 4.631852279284478,
      "grad_norm": 1.0,
      "learning_rate": 0.0003053683766543946,
      "loss": 5.4777,
      "mean_token_accuracy": 0.22106947898864746,
      "num_tokens": 101733033.0,
      "step": 40135
    },
    {
      "entropy": 6.004821109771728,
      "epoch": 4.632429313329487,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00030532748937703313,
      "loss": 5.4496,
      "mean_token_accuracy": 0.20908743888139725,
      "num_tokens": 101745633.0,
      "step": 40140
    },
    {
      "entropy": 6.111246776580811,
      "epoch": 4.633006347374495,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003052866010796218,
      "loss": 5.5116,
      "mean_token_accuracy": 0.2124009683728218,
      "num_tokens": 101757290.0,
      "step": 40145
    },
    {
      "entropy": 5.997294569015503,
      "epoch": 4.633583381419504,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003052457117635355,
      "loss": 5.3963,
      "mean_token_accuracy": 0.22513843923807145,
      "num_tokens": 101770275.0,
      "step": 40150
    },
    {
      "entropy": 6.032103252410889,
      "epoch": 4.634160415464512,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003052048214301498,
      "loss": 5.5333,
      "mean_token_accuracy": 0.22218757569789888,
      "num_tokens": 101784463.0,
      "step": 40155
    },
    {
      "entropy": 5.9645768165588375,
      "epoch": 4.634737449509521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030516393008083993,
      "loss": 5.3772,
      "mean_token_accuracy": 0.230370232462883,
      "num_tokens": 101798474.0,
      "step": 40160
    },
    {
      "entropy": 6.090016412734985,
      "epoch": 4.63531448355453,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00030512303771698126,
      "loss": 5.5642,
      "mean_token_accuracy": 0.20714891850948333,
      "num_tokens": 101810694.0,
      "step": 40165
    },
    {
      "entropy": 6.04038634300232,
      "epoch": 4.635891517599538,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030508214433994915,
      "loss": 5.5017,
      "mean_token_accuracy": 0.2169783666729927,
      "num_tokens": 101822968.0,
      "step": 40170
    },
    {
      "entropy": 5.979956483840942,
      "epoch": 4.636468551644547,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003050412499511191,
      "loss": 5.5168,
      "mean_token_accuracy": 0.22230055779218674,
      "num_tokens": 101835550.0,
      "step": 40175
    },
    {
      "entropy": 6.030199956893921,
      "epoch": 4.637045585689556,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003050003545518665,
      "loss": 5.564,
      "mean_token_accuracy": 0.20780004262924195,
      "num_tokens": 101847916.0,
      "step": 40180
    },
    {
      "entropy": 6.110112953186035,
      "epoch": 4.6376226197345645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030495945814356686,
      "loss": 5.59,
      "mean_token_accuracy": 0.2112215742468834,
      "num_tokens": 101860357.0,
      "step": 40185
    },
    {
      "entropy": 6.058466577529908,
      "epoch": 4.638199653779573,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030491856072759575,
      "loss": 5.582,
      "mean_token_accuracy": 0.20878954976797104,
      "num_tokens": 101873782.0,
      "step": 40190
    },
    {
      "entropy": 6.066515922546387,
      "epoch": 4.638776687824581,
      "grad_norm": 1.0,
      "learning_rate": 0.0003048776623053287,
      "loss": 5.4386,
      "mean_token_accuracy": 0.21951247304677962,
      "num_tokens": 101886187.0,
      "step": 40195
    },
    {
      "entropy": 6.183600425720215,
      "epoch": 4.639353721869591,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00030483676287814125,
      "loss": 5.6893,
      "mean_token_accuracy": 0.1957939237356186,
      "num_tokens": 101897849.0,
      "step": 40200
    },
    {
      "entropy": 6.015444374084472,
      "epoch": 4.639930755914599,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030479586244740924,
      "loss": 5.4217,
      "mean_token_accuracy": 0.21342692226171495,
      "num_tokens": 101910117.0,
      "step": 40205
    },
    {
      "entropy": 5.990631437301635,
      "epoch": 4.640507789959607,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030475496101450817,
      "loss": 5.4472,
      "mean_token_accuracy": 0.2224772483110428,
      "num_tokens": 101922252.0,
      "step": 40210
    },
    {
      "entropy": 6.104843282699585,
      "epoch": 4.641084824004617,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030471405858081374,
      "loss": 5.552,
      "mean_token_accuracy": 0.20883577018976213,
      "num_tokens": 101933893.0,
      "step": 40215
    },
    {
      "entropy": 6.05994553565979,
      "epoch": 4.641661858049625,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030467315514770164,
      "loss": 5.5226,
      "mean_token_accuracy": 0.21765454560518266,
      "num_tokens": 101947093.0,
      "step": 40220
    },
    {
      "entropy": 6.011163711547852,
      "epoch": 4.642238892094634,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030463225071654783,
      "loss": 5.4545,
      "mean_token_accuracy": 0.21549930721521376,
      "num_tokens": 101958931.0,
      "step": 40225
    },
    {
      "entropy": 6.057373571395874,
      "epoch": 4.642815926139642,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030459134528872795,
      "loss": 5.5204,
      "mean_token_accuracy": 0.21093595325946807,
      "num_tokens": 101971884.0,
      "step": 40230
    },
    {
      "entropy": 6.148956537246704,
      "epoch": 4.64339296018465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003045504388656178,
      "loss": 5.5503,
      "mean_token_accuracy": 0.2084178328514099,
      "num_tokens": 101983528.0,
      "step": 40235
    },
    {
      "entropy": 6.120232677459716,
      "epoch": 4.64396999422966,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003045095314485933,
      "loss": 5.5321,
      "mean_token_accuracy": 0.2111535221338272,
      "num_tokens": 101996206.0,
      "step": 40240
    },
    {
      "entropy": 6.115531969070434,
      "epoch": 4.644547028274668,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003044686230390305,
      "loss": 5.5944,
      "mean_token_accuracy": 0.20840590447187424,
      "num_tokens": 102009429.0,
      "step": 40245
    },
    {
      "entropy": 6.040201234817505,
      "epoch": 4.645124062319677,
      "grad_norm": 1.0,
      "learning_rate": 0.00030442771363830515,
      "loss": 5.4953,
      "mean_token_accuracy": 0.2116700142621994,
      "num_tokens": 102021531.0,
      "step": 40250
    },
    {
      "entropy": 6.148918962478637,
      "epoch": 4.645701096364686,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003043868032477933,
      "loss": 5.616,
      "mean_token_accuracy": 0.20456252992153168,
      "num_tokens": 102034800.0,
      "step": 40255
    },
    {
      "entropy": 5.948317766189575,
      "epoch": 4.646278130409694,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0003043458918688708,
      "loss": 5.3826,
      "mean_token_accuracy": 0.2217560812830925,
      "num_tokens": 102047934.0,
      "step": 40260
    },
    {
      "entropy": 6.167413949966431,
      "epoch": 4.646855164454703,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030430497950291387,
      "loss": 5.6501,
      "mean_token_accuracy": 0.19922988265752792,
      "num_tokens": 102059871.0,
      "step": 40265
    },
    {
      "entropy": 6.111153268814087,
      "epoch": 4.647432198499711,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003042640661512985,
      "loss": 5.5236,
      "mean_token_accuracy": 0.2110740065574646,
      "num_tokens": 102072780.0,
      "step": 40270
    },
    {
      "entropy": 6.108676433563232,
      "epoch": 4.64800923254472,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030422315181540086,
      "loss": 5.5466,
      "mean_token_accuracy": 0.2154386579990387,
      "num_tokens": 102085713.0,
      "step": 40275
    },
    {
      "entropy": 6.21439528465271,
      "epoch": 4.648586266589729,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000304182236496597,
      "loss": 5.6565,
      "mean_token_accuracy": 0.2038946896791458,
      "num_tokens": 102097471.0,
      "step": 40280
    },
    {
      "entropy": 5.991228103637695,
      "epoch": 4.649163300634737,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030414132019626307,
      "loss": 5.4461,
      "mean_token_accuracy": 0.22149051427841188,
      "num_tokens": 102110987.0,
      "step": 40285
    },
    {
      "entropy": 6.04875431060791,
      "epoch": 4.649740334679747,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030410040291577536,
      "loss": 5.4961,
      "mean_token_accuracy": 0.21272838711738587,
      "num_tokens": 102123599.0,
      "step": 40290
    },
    {
      "entropy": 6.092485666275024,
      "epoch": 4.650317368724755,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030405948465651,
      "loss": 5.5541,
      "mean_token_accuracy": 0.20986680835485458,
      "num_tokens": 102135997.0,
      "step": 40295
    },
    {
      "entropy": 6.0224292278289795,
      "epoch": 4.650894402769763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030401856541984335,
      "loss": 5.5239,
      "mean_token_accuracy": 0.213881978392601,
      "num_tokens": 102149290.0,
      "step": 40300
    },
    {
      "entropy": 6.025422430038452,
      "epoch": 4.651471436814772,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003039776452071516,
      "loss": 5.4551,
      "mean_token_accuracy": 0.22320044040679932,
      "num_tokens": 102162124.0,
      "step": 40305
    },
    {
      "entropy": 6.086610794067383,
      "epoch": 4.65204847085978,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003039367240198112,
      "loss": 5.5206,
      "mean_token_accuracy": 0.20983438491821288,
      "num_tokens": 102174173.0,
      "step": 40310
    },
    {
      "entropy": 6.021279144287109,
      "epoch": 4.6526255049047895,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003038958018591985,
      "loss": 5.4167,
      "mean_token_accuracy": 0.21398167312145233,
      "num_tokens": 102186388.0,
      "step": 40315
    },
    {
      "entropy": 6.050440835952759,
      "epoch": 4.653202538949798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003038548787266898,
      "loss": 5.5425,
      "mean_token_accuracy": 0.20860018730163574,
      "num_tokens": 102199860.0,
      "step": 40320
    },
    {
      "entropy": 6.138425636291504,
      "epoch": 4.653779572994806,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003038139546236616,
      "loss": 5.5871,
      "mean_token_accuracy": 0.21366337388753892,
      "num_tokens": 102212810.0,
      "step": 40325
    },
    {
      "entropy": 5.998503065109253,
      "epoch": 4.654356607039816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030377302955149034,
      "loss": 5.4552,
      "mean_token_accuracy": 0.2115766927599907,
      "num_tokens": 102223748.0,
      "step": 40330
    },
    {
      "entropy": 5.924293851852417,
      "epoch": 4.654933641084824,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030373210351155254,
      "loss": 5.4416,
      "mean_token_accuracy": 0.2317650258541107,
      "num_tokens": 102237052.0,
      "step": 40335
    },
    {
      "entropy": 6.101961135864258,
      "epoch": 4.6555106751298325,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030369117650522475,
      "loss": 5.4624,
      "mean_token_accuracy": 0.21635974794626237,
      "num_tokens": 102249298.0,
      "step": 40340
    },
    {
      "entropy": 6.016424560546875,
      "epoch": 4.656087709174841,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0003036502485338834,
      "loss": 5.4485,
      "mean_token_accuracy": 0.21837669163942336,
      "num_tokens": 102263758.0,
      "step": 40345
    },
    {
      "entropy": 6.049523639678955,
      "epoch": 4.65666474321985,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030360931959890516,
      "loss": 5.5368,
      "mean_token_accuracy": 0.21151572167873384,
      "num_tokens": 102276461.0,
      "step": 40350
    },
    {
      "entropy": 6.074767637252807,
      "epoch": 4.657241777264859,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003035683897016667,
      "loss": 5.5448,
      "mean_token_accuracy": 0.21361365020275117,
      "num_tokens": 102289594.0,
      "step": 40355
    },
    {
      "entropy": 6.036602640151978,
      "epoch": 4.657818811309867,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030352745884354475,
      "loss": 5.4473,
      "mean_token_accuracy": 0.21512358337640763,
      "num_tokens": 102300959.0,
      "step": 40360
    },
    {
      "entropy": 5.991143608093262,
      "epoch": 4.658395845354876,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030348652702591586,
      "loss": 5.4517,
      "mean_token_accuracy": 0.22633498162031174,
      "num_tokens": 102313736.0,
      "step": 40365
    },
    {
      "entropy": 6.057518005371094,
      "epoch": 4.658972879399885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003034455942501567,
      "loss": 5.5689,
      "mean_token_accuracy": 0.21355525851249696,
      "num_tokens": 102326302.0,
      "step": 40370
    },
    {
      "entropy": 6.107525587081909,
      "epoch": 4.659549913444893,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003034046605176441,
      "loss": 5.5176,
      "mean_token_accuracy": 0.21311535984277724,
      "num_tokens": 102337682.0,
      "step": 40375
    },
    {
      "entropy": 6.051120042800903,
      "epoch": 4.660126947489902,
      "grad_norm": 1.234375,
      "learning_rate": 0.00030336372582975493,
      "loss": 5.555,
      "mean_token_accuracy": 0.2172890067100525,
      "num_tokens": 102350046.0,
      "step": 40380
    },
    {
      "entropy": 6.078455829620362,
      "epoch": 4.660703981534911,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003033227901878659,
      "loss": 5.5699,
      "mean_token_accuracy": 0.210324989259243,
      "num_tokens": 102362905.0,
      "step": 40385
    },
    {
      "entropy": 6.121299362182617,
      "epoch": 4.661281015579919,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003032818535933538,
      "loss": 5.5045,
      "mean_token_accuracy": 0.21094539612531663,
      "num_tokens": 102374684.0,
      "step": 40390
    },
    {
      "entropy": 6.174676322937012,
      "epoch": 4.661858049624928,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003032409160475957,
      "loss": 5.6106,
      "mean_token_accuracy": 0.20823263078927995,
      "num_tokens": 102387268.0,
      "step": 40395
    },
    {
      "entropy": 6.069587230682373,
      "epoch": 4.662435083669936,
      "grad_norm": 1.0,
      "learning_rate": 0.0003031999775519685,
      "loss": 5.5582,
      "mean_token_accuracy": 0.2067757800221443,
      "num_tokens": 102398985.0,
      "step": 40400
    },
    {
      "entropy": 5.999266147613525,
      "epoch": 4.6630121177149455,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030315903810784896,
      "loss": 5.4253,
      "mean_token_accuracy": 0.22282031178474426,
      "num_tokens": 102411531.0,
      "step": 40405
    },
    {
      "entropy": 6.182189226150513,
      "epoch": 4.663589151759954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003031180977166141,
      "loss": 5.5796,
      "mean_token_accuracy": 0.20635385513305665,
      "num_tokens": 102424805.0,
      "step": 40410
    },
    {
      "entropy": 6.055428743362427,
      "epoch": 4.664166185804962,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003030771563796411,
      "loss": 5.5504,
      "mean_token_accuracy": 0.20864809453487396,
      "num_tokens": 102436132.0,
      "step": 40415
    },
    {
      "entropy": 6.000490760803222,
      "epoch": 4.664743219849971,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030303621409830686,
      "loss": 5.423,
      "mean_token_accuracy": 0.22001756131649017,
      "num_tokens": 102448708.0,
      "step": 40420
    },
    {
      "entropy": 6.062108278274536,
      "epoch": 4.66532025389498,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030299527087398846,
      "loss": 5.5652,
      "mean_token_accuracy": 0.20293550342321395,
      "num_tokens": 102459940.0,
      "step": 40425
    },
    {
      "entropy": 6.08624849319458,
      "epoch": 4.6658972879399885,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030295432670806303,
      "loss": 5.5075,
      "mean_token_accuracy": 0.20980108231306077,
      "num_tokens": 102473429.0,
      "step": 40430
    },
    {
      "entropy": 6.122294282913208,
      "epoch": 4.666474321984997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003029133816019077,
      "loss": 5.537,
      "mean_token_accuracy": 0.21689476519823075,
      "num_tokens": 102486752.0,
      "step": 40435
    },
    {
      "entropy": 5.9142271995544435,
      "epoch": 4.667051356030006,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003028724355568997,
      "loss": 5.3954,
      "mean_token_accuracy": 0.2256261706352234,
      "num_tokens": 102500455.0,
      "step": 40440
    },
    {
      "entropy": 6.075957632064819,
      "epoch": 4.667628390075015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030283148857441616,
      "loss": 5.5213,
      "mean_token_accuracy": 0.21587922871112825,
      "num_tokens": 102512688.0,
      "step": 40445
    },
    {
      "entropy": 5.963104343414306,
      "epoch": 4.668205424120023,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003027905406558343,
      "loss": 5.4315,
      "mean_token_accuracy": 0.22278572916984557,
      "num_tokens": 102525525.0,
      "step": 40450
    },
    {
      "entropy": 6.112188053131104,
      "epoch": 4.6687824581650315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003027495918025314,
      "loss": 5.5313,
      "mean_token_accuracy": 0.20939795672893524,
      "num_tokens": 102537309.0,
      "step": 40455
    },
    {
      "entropy": 6.087237215042114,
      "epoch": 4.669359492210041,
      "grad_norm": 0.921875,
      "learning_rate": 0.00030270864201588485,
      "loss": 5.5387,
      "mean_token_accuracy": 0.20934234112501143,
      "num_tokens": 102549647.0,
      "step": 40460
    },
    {
      "entropy": 6.000871658325195,
      "epoch": 4.669936526255049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030266769129727177,
      "loss": 5.4792,
      "mean_token_accuracy": 0.21280010789632797,
      "num_tokens": 102561736.0,
      "step": 40465
    },
    {
      "entropy": 5.9513002872467045,
      "epoch": 4.670513560300058,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030262673964806975,
      "loss": 5.3783,
      "mean_token_accuracy": 0.23099043667316438,
      "num_tokens": 102575171.0,
      "step": 40470
    },
    {
      "entropy": 6.081803798675537,
      "epoch": 4.671090594345066,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003025857870696559,
      "loss": 5.5691,
      "mean_token_accuracy": 0.21026743352413177,
      "num_tokens": 102587500.0,
      "step": 40475
    },
    {
      "entropy": 6.009596920013427,
      "epoch": 4.671667628390075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030254483356340794,
      "loss": 5.4175,
      "mean_token_accuracy": 0.22356452643871308,
      "num_tokens": 102601674.0,
      "step": 40480
    },
    {
      "entropy": 6.0100830078125,
      "epoch": 4.672244662435084,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030250387913070316,
      "loss": 5.4425,
      "mean_token_accuracy": 0.22131087481975556,
      "num_tokens": 102615370.0,
      "step": 40485
    },
    {
      "entropy": 6.114531230926514,
      "epoch": 4.672821696480092,
      "grad_norm": 0.90625,
      "learning_rate": 0.00030246292377291914,
      "loss": 5.5214,
      "mean_token_accuracy": 0.20989109575748444,
      "num_tokens": 102627755.0,
      "step": 40490
    },
    {
      "entropy": 6.0195910930633545,
      "epoch": 4.673398730525101,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030242196749143314,
      "loss": 5.5442,
      "mean_token_accuracy": 0.20827856957912444,
      "num_tokens": 102641095.0,
      "step": 40495
    },
    {
      "entropy": 5.998216962814331,
      "epoch": 4.67397576457011,
      "grad_norm": 1.015625,
      "learning_rate": 0.000302381010287623,
      "loss": 5.4269,
      "mean_token_accuracy": 0.221379292011261,
      "num_tokens": 102652582.0,
      "step": 40500
    },
    {
      "entropy": 6.112171316146851,
      "epoch": 4.674552798615118,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003023400521628662,
      "loss": 5.5914,
      "mean_token_accuracy": 0.20505908578634263,
      "num_tokens": 102665476.0,
      "step": 40505
    },
    {
      "entropy": 6.202398490905762,
      "epoch": 4.675129832660127,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003022990931185403,
      "loss": 5.6067,
      "mean_token_accuracy": 0.2000592902302742,
      "num_tokens": 102677440.0,
      "step": 40510
    },
    {
      "entropy": 6.161210823059082,
      "epoch": 4.675706866705136,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003022581331560229,
      "loss": 5.555,
      "mean_token_accuracy": 0.2069035455584526,
      "num_tokens": 102690400.0,
      "step": 40515
    },
    {
      "entropy": 6.058077383041382,
      "epoch": 4.676283900750144,
      "grad_norm": 1.0,
      "learning_rate": 0.0003022171722766918,
      "loss": 5.4938,
      "mean_token_accuracy": 0.21703884601593018,
      "num_tokens": 102703758.0,
      "step": 40520
    },
    {
      "entropy": 6.006094980239868,
      "epoch": 4.676860934795153,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030217621048192467,
      "loss": 5.4529,
      "mean_token_accuracy": 0.22131969183683395,
      "num_tokens": 102715237.0,
      "step": 40525
    },
    {
      "entropy": 6.05950813293457,
      "epoch": 4.677437968840161,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00030213524777309916,
      "loss": 5.5388,
      "mean_token_accuracy": 0.20908078253269197,
      "num_tokens": 102728766.0,
      "step": 40530
    },
    {
      "entropy": 6.074459266662598,
      "epoch": 4.678015002885171,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00030209428415159303,
      "loss": 5.5557,
      "mean_token_accuracy": 0.2142320990562439,
      "num_tokens": 102742393.0,
      "step": 40535
    },
    {
      "entropy": 6.168030214309693,
      "epoch": 4.678592036930179,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003020533196187841,
      "loss": 5.6318,
      "mean_token_accuracy": 0.20276992172002792,
      "num_tokens": 102756089.0,
      "step": 40540
    },
    {
      "entropy": 6.106509113311768,
      "epoch": 4.679169070975187,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00030201235417605024,
      "loss": 5.551,
      "mean_token_accuracy": 0.21083916425704957,
      "num_tokens": 102768654.0,
      "step": 40545
    },
    {
      "entropy": 6.0760866641998295,
      "epoch": 4.679746105020196,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003019713878247693,
      "loss": 5.5729,
      "mean_token_accuracy": 0.2167840138077736,
      "num_tokens": 102781443.0,
      "step": 40550
    },
    {
      "entropy": 6.120911026000977,
      "epoch": 4.680323139065205,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000301930420566319,
      "loss": 5.6031,
      "mean_token_accuracy": 0.2029586985707283,
      "num_tokens": 102793028.0,
      "step": 40555
    },
    {
      "entropy": 6.0567920207977295,
      "epoch": 4.680900173110214,
      "grad_norm": 1.453125,
      "learning_rate": 0.00030188945240207756,
      "loss": 5.4807,
      "mean_token_accuracy": 0.22060326784849166,
      "num_tokens": 102805652.0,
      "step": 40560
    },
    {
      "entropy": 6.0971879959106445,
      "epoch": 4.681477207155222,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030184848333342267,
      "loss": 5.5688,
      "mean_token_accuracy": 0.21173155307769775,
      "num_tokens": 102817959.0,
      "step": 40565
    },
    {
      "entropy": 6.079104995727539,
      "epoch": 4.68205424120023,
      "grad_norm": 1.0,
      "learning_rate": 0.0003018075133617324,
      "loss": 5.5244,
      "mean_token_accuracy": 0.21249692440032958,
      "num_tokens": 102830692.0,
      "step": 40570
    },
    {
      "entropy": 6.074726915359497,
      "epoch": 4.68263127524524,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003017665424883846,
      "loss": 5.5355,
      "mean_token_accuracy": 0.21349019408226014,
      "num_tokens": 102843495.0,
      "step": 40575
    },
    {
      "entropy": 6.0668299198150635,
      "epoch": 4.683208309290248,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003017255707147576,
      "loss": 5.4996,
      "mean_token_accuracy": 0.22020189166069032,
      "num_tokens": 102856585.0,
      "step": 40580
    },
    {
      "entropy": 6.111496925354004,
      "epoch": 4.6837853433352565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030168459804222927,
      "loss": 5.5844,
      "mean_token_accuracy": 0.2090723305940628,
      "num_tokens": 102869750.0,
      "step": 40585
    },
    {
      "entropy": 6.044322204589844,
      "epoch": 4.684362377380266,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003016436244721777,
      "loss": 5.4797,
      "mean_token_accuracy": 0.22113304287195207,
      "num_tokens": 102883680.0,
      "step": 40590
    },
    {
      "entropy": 6.044998550415039,
      "epoch": 4.684939411425274,
      "grad_norm": 1.0,
      "learning_rate": 0.0003016026500059811,
      "loss": 5.5064,
      "mean_token_accuracy": 0.214402474462986,
      "num_tokens": 102896869.0,
      "step": 40595
    },
    {
      "entropy": 6.031843852996826,
      "epoch": 4.685516445470283,
      "grad_norm": 1.0,
      "learning_rate": 0.00030156167464501747,
      "loss": 5.4457,
      "mean_token_accuracy": 0.2232113763689995,
      "num_tokens": 102910469.0,
      "step": 40600
    },
    {
      "entropy": 6.112345457077026,
      "epoch": 4.686093479515291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003015206983906652,
      "loss": 5.5375,
      "mean_token_accuracy": 0.20791415274143218,
      "num_tokens": 102922363.0,
      "step": 40605
    },
    {
      "entropy": 6.03321681022644,
      "epoch": 4.6866705135603,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030147972124430246,
      "loss": 5.579,
      "mean_token_accuracy": 0.2131233423948288,
      "num_tokens": 102934855.0,
      "step": 40610
    },
    {
      "entropy": 6.057859754562378,
      "epoch": 4.687247547605309,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030143874320730734,
      "loss": 5.5408,
      "mean_token_accuracy": 0.21599163264036178,
      "num_tokens": 102948088.0,
      "step": 40615
    },
    {
      "entropy": 6.110048151016235,
      "epoch": 4.687824581650317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003013977642810583,
      "loss": 5.5623,
      "mean_token_accuracy": 0.21446237713098526,
      "num_tokens": 102962152.0,
      "step": 40620
    },
    {
      "entropy": 5.978468990325927,
      "epoch": 4.688401615695326,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030135678446693354,
      "loss": 5.4435,
      "mean_token_accuracy": 0.21835525184869767,
      "num_tokens": 102975004.0,
      "step": 40625
    },
    {
      "entropy": 5.973551082611084,
      "epoch": 4.688978649740335,
      "grad_norm": 0.921875,
      "learning_rate": 0.00030131580376631145,
      "loss": 5.3749,
      "mean_token_accuracy": 0.22152878046035768,
      "num_tokens": 102988278.0,
      "step": 40630
    },
    {
      "entropy": 6.049982500076294,
      "epoch": 4.689555683785343,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030127482218057035,
      "loss": 5.4757,
      "mean_token_accuracy": 0.21126235127449036,
      "num_tokens": 103000246.0,
      "step": 40635
    },
    {
      "entropy": 6.013923406600952,
      "epoch": 4.690132717830352,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003012338397110886,
      "loss": 5.4848,
      "mean_token_accuracy": 0.213090743124485,
      "num_tokens": 103014001.0,
      "step": 40640
    },
    {
      "entropy": 6.0590473175048825,
      "epoch": 4.69070975187536,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00030119285635924484,
      "loss": 5.5176,
      "mean_token_accuracy": 0.2110501080751419,
      "num_tokens": 103027539.0,
      "step": 40645
    },
    {
      "entropy": 6.025898361206055,
      "epoch": 4.6912867859203695,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00030115187212641724,
      "loss": 5.435,
      "mean_token_accuracy": 0.2183668226003647,
      "num_tokens": 103040991.0,
      "step": 40650
    },
    {
      "entropy": 5.991097068786621,
      "epoch": 4.691863819965378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003011108870139845,
      "loss": 5.3892,
      "mean_token_accuracy": 0.22508731633424758,
      "num_tokens": 103053671.0,
      "step": 40655
    },
    {
      "entropy": 6.025773811340332,
      "epoch": 4.692440854010386,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003010699010233249,
      "loss": 5.462,
      "mean_token_accuracy": 0.22039386183023452,
      "num_tokens": 103065152.0,
      "step": 40660
    },
    {
      "entropy": 5.994659948348999,
      "epoch": 4.693017888055396,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030102891415581724,
      "loss": 5.4255,
      "mean_token_accuracy": 0.22461961656808854,
      "num_tokens": 103078114.0,
      "step": 40665
    },
    {
      "entropy": 5.98750581741333,
      "epoch": 4.693594922100404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00030098792641283995,
      "loss": 5.4152,
      "mean_token_accuracy": 0.21898276060819627,
      "num_tokens": 103089869.0,
      "step": 40670
    },
    {
      "entropy": 6.028250026702881,
      "epoch": 4.6941719561454125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030094693779577177,
      "loss": 5.4619,
      "mean_token_accuracy": 0.21789764165878295,
      "num_tokens": 103103371.0,
      "step": 40675
    },
    {
      "entropy": 6.020797872543335,
      "epoch": 4.694748990190421,
      "grad_norm": 0.953125,
      "learning_rate": 0.00030090594830599104,
      "loss": 5.4616,
      "mean_token_accuracy": 0.2280898854136467,
      "num_tokens": 103116526.0,
      "step": 40680
    },
    {
      "entropy": 5.92597770690918,
      "epoch": 4.69532602423543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003008649579448767,
      "loss": 5.3479,
      "mean_token_accuracy": 0.23473746031522752,
      "num_tokens": 103129493.0,
      "step": 40685
    },
    {
      "entropy": 5.989361953735352,
      "epoch": 4.695903058280439,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030082396671380734,
      "loss": 5.5248,
      "mean_token_accuracy": 0.2192692995071411,
      "num_tokens": 103143721.0,
      "step": 40690
    },
    {
      "entropy": 6.086354351043701,
      "epoch": 4.696480092325447,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003007829746141617,
      "loss": 5.6002,
      "mean_token_accuracy": 0.21117883920669556,
      "num_tokens": 103157976.0,
      "step": 40695
    },
    {
      "entropy": 6.078159999847412,
      "epoch": 4.6970571263704555,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030074198164731854,
      "loss": 5.4782,
      "mean_token_accuracy": 0.21990445852279664,
      "num_tokens": 103169356.0,
      "step": 40700
    },
    {
      "entropy": 6.11819543838501,
      "epoch": 4.697634160415465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003007009878146566,
      "loss": 5.5506,
      "mean_token_accuracy": 0.2135201498866081,
      "num_tokens": 103181204.0,
      "step": 40705
    },
    {
      "entropy": 6.076857233047486,
      "epoch": 4.698211194460473,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00030065999311755466,
      "loss": 5.5967,
      "mean_token_accuracy": 0.2082436889410019,
      "num_tokens": 103195101.0,
      "step": 40710
    },
    {
      "entropy": 6.055859184265136,
      "epoch": 4.698788228505482,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003006189975573916,
      "loss": 5.4453,
      "mean_token_accuracy": 0.2220652461051941,
      "num_tokens": 103208129.0,
      "step": 40715
    },
    {
      "entropy": 6.0622953414917,
      "epoch": 4.69936526255049,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030057800113554627,
      "loss": 5.4505,
      "mean_token_accuracy": 0.21694075167179108,
      "num_tokens": 103220743.0,
      "step": 40720
    },
    {
      "entropy": 6.059666872024536,
      "epoch": 4.699942296595499,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030053700385339766,
      "loss": 5.5488,
      "mean_token_accuracy": 0.20757769793272018,
      "num_tokens": 103233664.0,
      "step": 40725
    },
    {
      "entropy": 6.071609783172607,
      "epoch": 4.700519330640508,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003004960057123246,
      "loss": 5.4654,
      "mean_token_accuracy": 0.21064720600843428,
      "num_tokens": 103247144.0,
      "step": 40730
    },
    {
      "entropy": 6.116744899749756,
      "epoch": 4.701096364685516,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000300455006713706,
      "loss": 5.5842,
      "mean_token_accuracy": 0.2080810397863388,
      "num_tokens": 103258735.0,
      "step": 40735
    },
    {
      "entropy": 6.037677669525147,
      "epoch": 4.7016733987305255,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00030041400685892086,
      "loss": 5.4948,
      "mean_token_accuracy": 0.21213069558143616,
      "num_tokens": 103272370.0,
      "step": 40740
    },
    {
      "entropy": 5.990386724472046,
      "epoch": 4.702250432775534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003003730061493483,
      "loss": 5.4663,
      "mean_token_accuracy": 0.21988701373338698,
      "num_tokens": 103284215.0,
      "step": 40745
    },
    {
      "entropy": 6.119244003295899,
      "epoch": 4.702827466820542,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00030033200458636714,
      "loss": 5.5562,
      "mean_token_accuracy": 0.21128773987293242,
      "num_tokens": 103297560.0,
      "step": 40750
    },
    {
      "entropy": 5.98164553642273,
      "epoch": 4.703404500865551,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030029100217135674,
      "loss": 5.3948,
      "mean_token_accuracy": 0.2216815397143364,
      "num_tokens": 103309861.0,
      "step": 40755
    },
    {
      "entropy": 5.983503437042236,
      "epoch": 4.70398153491056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030024999890569605,
      "loss": 5.4354,
      "mean_token_accuracy": 0.2225635066628456,
      "num_tokens": 103322244.0,
      "step": 40760
    },
    {
      "entropy": 6.081357336044311,
      "epoch": 4.7045585689555685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003002089947907641,
      "loss": 5.5848,
      "mean_token_accuracy": 0.21002936959266663,
      "num_tokens": 103334943.0,
      "step": 40765
    },
    {
      "entropy": 6.1299222946167,
      "epoch": 4.705135603000577,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003001679898279402,
      "loss": 5.5531,
      "mean_token_accuracy": 0.20946212708950043,
      "num_tokens": 103346851.0,
      "step": 40770
    },
    {
      "entropy": 6.078902244567871,
      "epoch": 4.705712637045586,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003001269840186034,
      "loss": 5.5241,
      "mean_token_accuracy": 0.21239568442106246,
      "num_tokens": 103359434.0,
      "step": 40775
    },
    {
      "entropy": 6.0624631404876705,
      "epoch": 4.706289671090595,
      "grad_norm": 1.171875,
      "learning_rate": 0.00030008597736413305,
      "loss": 5.4825,
      "mean_token_accuracy": 0.20944168716669082,
      "num_tokens": 103371945.0,
      "step": 40780
    },
    {
      "entropy": 6.023629283905029,
      "epoch": 4.706866705135603,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030004496986590825,
      "loss": 5.4984,
      "mean_token_accuracy": 0.21912001669406891,
      "num_tokens": 103385204.0,
      "step": 40785
    },
    {
      "entropy": 6.022859191894531,
      "epoch": 4.707443739180611,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00030000396152530834,
      "loss": 5.477,
      "mean_token_accuracy": 0.218401937186718,
      "num_tokens": 103398024.0,
      "step": 40790
    },
    {
      "entropy": 6.062512922286987,
      "epoch": 4.70802077322562,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002999629523437127,
      "loss": 5.5153,
      "mean_token_accuracy": 0.21566094905138017,
      "num_tokens": 103410877.0,
      "step": 40795
    },
    {
      "entropy": 6.0337498664855955,
      "epoch": 4.708597807270629,
      "grad_norm": 1.0,
      "learning_rate": 0.0002999219423225005,
      "loss": 5.4405,
      "mean_token_accuracy": 0.21865384727716447,
      "num_tokens": 103423329.0,
      "step": 40800
    },
    {
      "entropy": 6.02803111076355,
      "epoch": 4.709174841315638,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029988093146305104,
      "loss": 5.4869,
      "mean_token_accuracy": 0.2196614056825638,
      "num_tokens": 103435434.0,
      "step": 40805
    },
    {
      "entropy": 6.008598518371582,
      "epoch": 4.709751875360646,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000299839919766744,
      "loss": 5.4641,
      "mean_token_accuracy": 0.2177417680621147,
      "num_tokens": 103447696.0,
      "step": 40810
    },
    {
      "entropy": 5.986138439178466,
      "epoch": 4.710328909405655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002997989072349584,
      "loss": 5.4263,
      "mean_token_accuracy": 0.22091514021158218,
      "num_tokens": 103460749.0,
      "step": 40815
    },
    {
      "entropy": 6.046768760681152,
      "epoch": 4.710905943450664,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029975789386907397,
      "loss": 5.5364,
      "mean_token_accuracy": 0.20974406003952026,
      "num_tokens": 103472951.0,
      "step": 40820
    },
    {
      "entropy": 6.050793695449829,
      "epoch": 4.711482977495672,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029971687967047004,
      "loss": 5.5111,
      "mean_token_accuracy": 0.21794998943805693,
      "num_tokens": 103486240.0,
      "step": 40825
    },
    {
      "entropy": 6.173254346847534,
      "epoch": 4.7120600115406805,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00029967586464052623,
      "loss": 5.6145,
      "mean_token_accuracy": 0.2032383620738983,
      "num_tokens": 103498934.0,
      "step": 40830
    },
    {
      "entropy": 6.058687210083008,
      "epoch": 4.71263704558569,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002996348487806219,
      "loss": 5.5129,
      "mean_token_accuracy": 0.21091001927852632,
      "num_tokens": 103513094.0,
      "step": 40835
    },
    {
      "entropy": 6.062951850891113,
      "epoch": 4.713214079630698,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002995938320921367,
      "loss": 5.542,
      "mean_token_accuracy": 0.20624936372041702,
      "num_tokens": 103525705.0,
      "step": 40840
    },
    {
      "entropy": 6.097893285751343,
      "epoch": 4.713791113675707,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029955281457645015,
      "loss": 5.5605,
      "mean_token_accuracy": 0.20943787097930908,
      "num_tokens": 103539160.0,
      "step": 40845
    },
    {
      "entropy": 6.13492488861084,
      "epoch": 4.714368147720716,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002995117962349419,
      "loss": 5.4996,
      "mean_token_accuracy": 0.21205293238162995,
      "num_tokens": 103551602.0,
      "step": 40850
    },
    {
      "entropy": 6.047506093978882,
      "epoch": 4.714945181765724,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002994707770689915,
      "loss": 5.4847,
      "mean_token_accuracy": 0.2184184432029724,
      "num_tokens": 103564281.0,
      "step": 40855
    },
    {
      "entropy": 6.056606674194336,
      "epoch": 4.715522215810733,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002994297570799787,
      "loss": 5.5489,
      "mean_token_accuracy": 0.20512326210737228,
      "num_tokens": 103576533.0,
      "step": 40860
    },
    {
      "entropy": 6.098846197128296,
      "epoch": 4.716099249855741,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029938873626928327,
      "loss": 5.5152,
      "mean_token_accuracy": 0.2128812462091446,
      "num_tokens": 103588916.0,
      "step": 40865
    },
    {
      "entropy": 6.063002061843872,
      "epoch": 4.71667628390075,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029934771463828463,
      "loss": 5.5468,
      "mean_token_accuracy": 0.2154998615384102,
      "num_tokens": 103602224.0,
      "step": 40870
    },
    {
      "entropy": 5.978532123565674,
      "epoch": 4.717253317945759,
      "grad_norm": 1.0,
      "learning_rate": 0.00029930669218836283,
      "loss": 5.4142,
      "mean_token_accuracy": 0.22095377892255783,
      "num_tokens": 103614975.0,
      "step": 40875
    },
    {
      "entropy": 6.088528156280518,
      "epoch": 4.717830351990767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002992656689208974,
      "loss": 5.5481,
      "mean_token_accuracy": 0.2112618550658226,
      "num_tokens": 103626752.0,
      "step": 40880
    },
    {
      "entropy": 6.144932556152344,
      "epoch": 4.718407386035776,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029922464483726833,
      "loss": 5.5768,
      "mean_token_accuracy": 0.20175130665302277,
      "num_tokens": 103639841.0,
      "step": 40885
    },
    {
      "entropy": 6.095277547836304,
      "epoch": 4.718984420080785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029918361993885536,
      "loss": 5.5269,
      "mean_token_accuracy": 0.2107507675886154,
      "num_tokens": 103653064.0,
      "step": 40890
    },
    {
      "entropy": 6.075502204895019,
      "epoch": 4.7195614541257935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002991425942270383,
      "loss": 5.5212,
      "mean_token_accuracy": 0.21893896162509918,
      "num_tokens": 103666109.0,
      "step": 40895
    },
    {
      "entropy": 6.056168460845948,
      "epoch": 4.720138488170802,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029910156770319714,
      "loss": 5.5139,
      "mean_token_accuracy": 0.21284041106700896,
      "num_tokens": 103678728.0,
      "step": 40900
    },
    {
      "entropy": 6.1105872631073,
      "epoch": 4.72071552221581,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029906054036871166,
      "loss": 5.5479,
      "mean_token_accuracy": 0.2111024796962738,
      "num_tokens": 103691488.0,
      "step": 40905
    },
    {
      "entropy": 6.117109107971191,
      "epoch": 4.72129255626082,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029901951222496186,
      "loss": 5.5742,
      "mean_token_accuracy": 0.20292351245880128,
      "num_tokens": 103703954.0,
      "step": 40910
    },
    {
      "entropy": 6.08160662651062,
      "epoch": 4.721869590305828,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029897848327332775,
      "loss": 5.5468,
      "mean_token_accuracy": 0.21015863865613937,
      "num_tokens": 103717128.0,
      "step": 40915
    },
    {
      "entropy": 5.963445043563842,
      "epoch": 4.7224466243508365,
      "grad_norm": 0.875,
      "learning_rate": 0.00029893745351518926,
      "loss": 5.4468,
      "mean_token_accuracy": 0.221268193423748,
      "num_tokens": 103729206.0,
      "step": 40920
    },
    {
      "entropy": 6.0469564437866214,
      "epoch": 4.723023658395846,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002988964229519264,
      "loss": 5.5042,
      "mean_token_accuracy": 0.2128221198916435,
      "num_tokens": 103742156.0,
      "step": 40925
    },
    {
      "entropy": 6.023454236984253,
      "epoch": 4.723600692440854,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029885539158491916,
      "loss": 5.4906,
      "mean_token_accuracy": 0.21182971596717834,
      "num_tokens": 103754123.0,
      "step": 40930
    },
    {
      "entropy": 6.096383190155029,
      "epoch": 4.724177726485863,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002988143594155478,
      "loss": 5.5939,
      "mean_token_accuracy": 0.2097270295023918,
      "num_tokens": 103767001.0,
      "step": 40935
    },
    {
      "entropy": 6.088947677612305,
      "epoch": 4.724754760530871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029877332644519223,
      "loss": 5.5007,
      "mean_token_accuracy": 0.21239416152238846,
      "num_tokens": 103778434.0,
      "step": 40940
    },
    {
      "entropy": 6.058033418655396,
      "epoch": 4.7253317945758795,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002987322926752326,
      "loss": 5.4808,
      "mean_token_accuracy": 0.21226408183574677,
      "num_tokens": 103791200.0,
      "step": 40945
    },
    {
      "entropy": 6.052468347549438,
      "epoch": 4.725908828620889,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029869125810704904,
      "loss": 5.4018,
      "mean_token_accuracy": 0.22025685310363768,
      "num_tokens": 103802698.0,
      "step": 40950
    },
    {
      "entropy": 5.957513666152954,
      "epoch": 4.726485862665897,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002986502227420219,
      "loss": 5.3751,
      "mean_token_accuracy": 0.22531258016824723,
      "num_tokens": 103814952.0,
      "step": 40955
    },
    {
      "entropy": 5.969437313079834,
      "epoch": 4.727062896710906,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029860918658153123,
      "loss": 5.4229,
      "mean_token_accuracy": 0.21807370632886885,
      "num_tokens": 103828006.0,
      "step": 40960
    },
    {
      "entropy": 6.065869426727295,
      "epoch": 4.727639930755915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029856814962695734,
      "loss": 5.6246,
      "mean_token_accuracy": 0.20851611942052842,
      "num_tokens": 103840978.0,
      "step": 40965
    },
    {
      "entropy": 6.05151333808899,
      "epoch": 4.728216964800923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029852711187968036,
      "loss": 5.5559,
      "mean_token_accuracy": 0.20919988602399825,
      "num_tokens": 103852922.0,
      "step": 40970
    },
    {
      "entropy": 6.074489259719849,
      "epoch": 4.728793998845932,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029848607334108065,
      "loss": 5.4347,
      "mean_token_accuracy": 0.22283679842948914,
      "num_tokens": 103865976.0,
      "step": 40975
    },
    {
      "entropy": 6.070368957519531,
      "epoch": 4.72937103289094,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029844503401253866,
      "loss": 5.5209,
      "mean_token_accuracy": 0.2183467611670494,
      "num_tokens": 103879595.0,
      "step": 40980
    },
    {
      "entropy": 6.064594173431397,
      "epoch": 4.7299480669359495,
      "grad_norm": 1.0,
      "learning_rate": 0.0002984039938954344,
      "loss": 5.579,
      "mean_token_accuracy": 0.20583467185497284,
      "num_tokens": 103892799.0,
      "step": 40985
    },
    {
      "entropy": 6.025279760360718,
      "epoch": 4.730525100980958,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029836295299114855,
      "loss": 5.4034,
      "mean_token_accuracy": 0.22447287887334824,
      "num_tokens": 103904962.0,
      "step": 40990
    },
    {
      "entropy": 6.061993885040283,
      "epoch": 4.731102135025966,
      "grad_norm": 1.0,
      "learning_rate": 0.0002983219113010614,
      "loss": 5.4631,
      "mean_token_accuracy": 0.22052764892578125,
      "num_tokens": 103919198.0,
      "step": 40995
    },
    {
      "entropy": 5.9834754943847654,
      "epoch": 4.731679169070976,
      "grad_norm": 0.890625,
      "learning_rate": 0.00029828086882655334,
      "loss": 5.5115,
      "mean_token_accuracy": 0.21856827288866043,
      "num_tokens": 103932955.0,
      "step": 41000
    },
    {
      "entropy": 6.038109064102173,
      "epoch": 4.732256203115984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002982398255690048,
      "loss": 5.5211,
      "mean_token_accuracy": 0.2128079727292061,
      "num_tokens": 103944901.0,
      "step": 41005
    },
    {
      "entropy": 6.036848497390747,
      "epoch": 4.7328332371609925,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029819878152979624,
      "loss": 5.4663,
      "mean_token_accuracy": 0.21948096454143523,
      "num_tokens": 103956798.0,
      "step": 41010
    },
    {
      "entropy": 5.9218159198760985,
      "epoch": 4.733410271206001,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002981577367103082,
      "loss": 5.3434,
      "mean_token_accuracy": 0.2228821337223053,
      "num_tokens": 103970566.0,
      "step": 41015
    },
    {
      "entropy": 6.064480876922607,
      "epoch": 4.73398730525101,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029811669111192123,
      "loss": 5.4991,
      "mean_token_accuracy": 0.21889271438121796,
      "num_tokens": 103982756.0,
      "step": 41020
    },
    {
      "entropy": 6.033596134185791,
      "epoch": 4.734564339296019,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029807564473601585,
      "loss": 5.48,
      "mean_token_accuracy": 0.21389850080013276,
      "num_tokens": 103995414.0,
      "step": 41025
    },
    {
      "entropy": 6.007001876831055,
      "epoch": 4.735141373341027,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002980345975839726,
      "loss": 5.3953,
      "mean_token_accuracy": 0.2226052924990654,
      "num_tokens": 104006656.0,
      "step": 41030
    },
    {
      "entropy": 5.99414291381836,
      "epoch": 4.7357184073860354,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002979935496571721,
      "loss": 5.4481,
      "mean_token_accuracy": 0.22291264981031417,
      "num_tokens": 104019075.0,
      "step": 41035
    },
    {
      "entropy": 6.078718137741089,
      "epoch": 4.736295441431045,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029795250095699494,
      "loss": 5.5429,
      "mean_token_accuracy": 0.21271851360797883,
      "num_tokens": 104032280.0,
      "step": 41040
    },
    {
      "entropy": 6.020309400558472,
      "epoch": 4.736872475476053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002979114514848218,
      "loss": 5.4535,
      "mean_token_accuracy": 0.2141135513782501,
      "num_tokens": 104044783.0,
      "step": 41045
    },
    {
      "entropy": 6.138972997665405,
      "epoch": 4.737449509521062,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029787040124203346,
      "loss": 5.5618,
      "mean_token_accuracy": 0.20506847351789476,
      "num_tokens": 104057224.0,
      "step": 41050
    },
    {
      "entropy": 6.098556852340698,
      "epoch": 4.73802654356607,
      "grad_norm": 1.0,
      "learning_rate": 0.00029782935023001046,
      "loss": 5.5916,
      "mean_token_accuracy": 0.21267506927251817,
      "num_tokens": 104071187.0,
      "step": 41055
    },
    {
      "entropy": 6.095409774780274,
      "epoch": 4.738603577611079,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002977882984501336,
      "loss": 5.5383,
      "mean_token_accuracy": 0.2130522459745407,
      "num_tokens": 104084788.0,
      "step": 41060
    },
    {
      "entropy": 6.064972162246704,
      "epoch": 4.739180611656088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002977472459037837,
      "loss": 5.4513,
      "mean_token_accuracy": 0.2187037467956543,
      "num_tokens": 104097452.0,
      "step": 41065
    },
    {
      "entropy": 5.983641624450684,
      "epoch": 4.739757645701096,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029770619259234145,
      "loss": 5.3842,
      "mean_token_accuracy": 0.2228585109114647,
      "num_tokens": 104111197.0,
      "step": 41070
    },
    {
      "entropy": 6.117957735061646,
      "epoch": 4.7403346797461055,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029766513851718765,
      "loss": 5.5811,
      "mean_token_accuracy": 0.20884908735752106,
      "num_tokens": 104124065.0,
      "step": 41075
    },
    {
      "entropy": 5.982294988632202,
      "epoch": 4.740911713791114,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002976240836797032,
      "loss": 5.4455,
      "mean_token_accuracy": 0.22249816358089447,
      "num_tokens": 104137441.0,
      "step": 41080
    },
    {
      "entropy": 6.037612676620483,
      "epoch": 4.741488747836122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029758302808126895,
      "loss": 5.4973,
      "mean_token_accuracy": 0.21552164405584334,
      "num_tokens": 104149154.0,
      "step": 41085
    },
    {
      "entropy": 6.1168077945709225,
      "epoch": 4.742065781881131,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002975419717232658,
      "loss": 5.6265,
      "mean_token_accuracy": 0.21063170582056046,
      "num_tokens": 104160482.0,
      "step": 41090
    },
    {
      "entropy": 6.088898372650147,
      "epoch": 4.74264281592614,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029750091460707456,
      "loss": 5.5104,
      "mean_token_accuracy": 0.2148719012737274,
      "num_tokens": 104173648.0,
      "step": 41095
    },
    {
      "entropy": 6.096096658706665,
      "epoch": 4.743219849971148,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002974598567340763,
      "loss": 5.478,
      "mean_token_accuracy": 0.21064760237932206,
      "num_tokens": 104185129.0,
      "step": 41100
    },
    {
      "entropy": 5.954454374313355,
      "epoch": 4.743796884016157,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002974187981056519,
      "loss": 5.4121,
      "mean_token_accuracy": 0.22588610649108887,
      "num_tokens": 104198235.0,
      "step": 41105
    },
    {
      "entropy": 6.0615723609924315,
      "epoch": 4.744373918061165,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002973777387231824,
      "loss": 5.536,
      "mean_token_accuracy": 0.21372592449188232,
      "num_tokens": 104212892.0,
      "step": 41110
    },
    {
      "entropy": 6.126613330841065,
      "epoch": 4.744950952106175,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029733667858804864,
      "loss": 5.5666,
      "mean_token_accuracy": 0.20422818511724472,
      "num_tokens": 104225366.0,
      "step": 41115
    },
    {
      "entropy": 6.089966535568237,
      "epoch": 4.745527986151183,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029729561770163186,
      "loss": 5.5443,
      "mean_token_accuracy": 0.2123413488268852,
      "num_tokens": 104236791.0,
      "step": 41120
    },
    {
      "entropy": 5.967685842514038,
      "epoch": 4.746105020196191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002972545560653131,
      "loss": 5.3715,
      "mean_token_accuracy": 0.22523281276226043,
      "num_tokens": 104249003.0,
      "step": 41125
    },
    {
      "entropy": 5.980962705612183,
      "epoch": 4.7466820542412,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002972134936804734,
      "loss": 5.4012,
      "mean_token_accuracy": 0.22716932743787766,
      "num_tokens": 104261773.0,
      "step": 41130
    },
    {
      "entropy": 6.018964385986328,
      "epoch": 4.747259088286209,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002971724305484938,
      "loss": 5.4963,
      "mean_token_accuracy": 0.2135877400636673,
      "num_tokens": 104274847.0,
      "step": 41135
    },
    {
      "entropy": 6.043982410430909,
      "epoch": 4.7478361223312175,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029713136667075556,
      "loss": 5.5009,
      "mean_token_accuracy": 0.2169567123055458,
      "num_tokens": 104287650.0,
      "step": 41140
    },
    {
      "entropy": 6.138433742523193,
      "epoch": 4.748413156376226,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029709030204863975,
      "loss": 5.5291,
      "mean_token_accuracy": 0.2167165294289589,
      "num_tokens": 104300106.0,
      "step": 41145
    },
    {
      "entropy": 6.044185161590576,
      "epoch": 4.748990190421235,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029704923668352767,
      "loss": 5.4632,
      "mean_token_accuracy": 0.21755122989416123,
      "num_tokens": 104314288.0,
      "step": 41150
    },
    {
      "entropy": 5.973044538497925,
      "epoch": 4.749567224466244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002970081705768005,
      "loss": 5.431,
      "mean_token_accuracy": 0.2229637995362282,
      "num_tokens": 104327826.0,
      "step": 41155
    },
    {
      "entropy": 5.958447504043579,
      "epoch": 4.750144258511252,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029696710372983926,
      "loss": 5.3956,
      "mean_token_accuracy": 0.224171444773674,
      "num_tokens": 104340535.0,
      "step": 41160
    },
    {
      "entropy": 6.063758373260498,
      "epoch": 4.7507212925562605,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002969260361440255,
      "loss": 5.517,
      "mean_token_accuracy": 0.21696289032697677,
      "num_tokens": 104352444.0,
      "step": 41165
    },
    {
      "entropy": 6.073453140258789,
      "epoch": 4.75129832660127,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029688496782074046,
      "loss": 5.4681,
      "mean_token_accuracy": 0.21553591191768645,
      "num_tokens": 104364434.0,
      "step": 41170
    },
    {
      "entropy": 6.031896543502808,
      "epoch": 4.751875360646278,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029684389876136544,
      "loss": 5.447,
      "mean_token_accuracy": 0.22167080342769624,
      "num_tokens": 104378352.0,
      "step": 41175
    },
    {
      "entropy": 5.99749526977539,
      "epoch": 4.752452394691287,
      "grad_norm": 1.0,
      "learning_rate": 0.0002968028289672816,
      "loss": 5.4582,
      "mean_token_accuracy": 0.2226710394024849,
      "num_tokens": 104390096.0,
      "step": 41180
    },
    {
      "entropy": 5.988511943817139,
      "epoch": 4.753029428736295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002967617584398705,
      "loss": 5.4613,
      "mean_token_accuracy": 0.21954770833253862,
      "num_tokens": 104403746.0,
      "step": 41185
    },
    {
      "entropy": 5.845155811309814,
      "epoch": 4.753606462781304,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002967206871805135,
      "loss": 5.2796,
      "mean_token_accuracy": 0.23126329481601715,
      "num_tokens": 104416796.0,
      "step": 41190
    },
    {
      "entropy": 6.160752677917481,
      "epoch": 4.754183496826313,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029667961519059193,
      "loss": 5.5588,
      "mean_token_accuracy": 0.20575629770755768,
      "num_tokens": 104428959.0,
      "step": 41195
    },
    {
      "entropy": 6.070642280578613,
      "epoch": 4.754760530871321,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002966385424714872,
      "loss": 5.5414,
      "mean_token_accuracy": 0.21221750676631929,
      "num_tokens": 104442992.0,
      "step": 41200
    },
    {
      "entropy": 6.01796145439148,
      "epoch": 4.75533756491633,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000296597469024581,
      "loss": 5.4557,
      "mean_token_accuracy": 0.21803570091724395,
      "num_tokens": 104454621.0,
      "step": 41205
    },
    {
      "entropy": 6.13784122467041,
      "epoch": 4.755914598961339,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002965563948512545,
      "loss": 5.5406,
      "mean_token_accuracy": 0.20647853314876558,
      "num_tokens": 104465674.0,
      "step": 41210
    },
    {
      "entropy": 6.133686304092407,
      "epoch": 4.756491633006347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002965153199528895,
      "loss": 5.6044,
      "mean_token_accuracy": 0.20142944306135177,
      "num_tokens": 104477197.0,
      "step": 41215
    },
    {
      "entropy": 6.099056720733643,
      "epoch": 4.757068667051356,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002964742443308674,
      "loss": 5.5212,
      "mean_token_accuracy": 0.20995987355709075,
      "num_tokens": 104489720.0,
      "step": 41220
    },
    {
      "entropy": 6.0465374946594235,
      "epoch": 4.757645701096365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029643316798656967,
      "loss": 5.4687,
      "mean_token_accuracy": 0.22009489834308624,
      "num_tokens": 104502279.0,
      "step": 41225
    },
    {
      "entropy": 5.9896721839904785,
      "epoch": 4.7582227351413735,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000296392090921378,
      "loss": 5.4265,
      "mean_token_accuracy": 0.2296479046344757,
      "num_tokens": 104515865.0,
      "step": 41230
    },
    {
      "entropy": 6.043453550338745,
      "epoch": 4.758799769186382,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000296351013136674,
      "loss": 5.6027,
      "mean_token_accuracy": 0.2050420805811882,
      "num_tokens": 104531117.0,
      "step": 41235
    },
    {
      "entropy": 5.996869325637817,
      "epoch": 4.75937680323139,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002963099346338393,
      "loss": 5.4059,
      "mean_token_accuracy": 0.22415702193975448,
      "num_tokens": 104542621.0,
      "step": 41240
    },
    {
      "entropy": 6.0908228874206545,
      "epoch": 4.7599538372764,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029626885541425546,
      "loss": 5.5956,
      "mean_token_accuracy": 0.20917401015758513,
      "num_tokens": 104555859.0,
      "step": 41245
    },
    {
      "entropy": 6.156450605392456,
      "epoch": 4.760530871321408,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002962277754793043,
      "loss": 5.5883,
      "mean_token_accuracy": 0.20979728400707245,
      "num_tokens": 104569322.0,
      "step": 41250
    },
    {
      "entropy": 6.130332136154175,
      "epoch": 4.7611079053664165,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002961866948303675,
      "loss": 5.5935,
      "mean_token_accuracy": 0.20549824088811874,
      "num_tokens": 104582542.0,
      "step": 41255
    },
    {
      "entropy": 6.095255422592163,
      "epoch": 4.761684939411425,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002961456134688266,
      "loss": 5.5586,
      "mean_token_accuracy": 0.20963636487722398,
      "num_tokens": 104594772.0,
      "step": 41260
    },
    {
      "entropy": 6.027664852142334,
      "epoch": 4.762261973456434,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002961045313960636,
      "loss": 5.4349,
      "mean_token_accuracy": 0.21449389308691025,
      "num_tokens": 104608077.0,
      "step": 41265
    },
    {
      "entropy": 6.146485614776611,
      "epoch": 4.762839007501443,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029606344861346013,
      "loss": 5.6589,
      "mean_token_accuracy": 0.20151630342006682,
      "num_tokens": 104620725.0,
      "step": 41270
    },
    {
      "entropy": 6.105652236938477,
      "epoch": 4.763416041546451,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029602236512239803,
      "loss": 5.5305,
      "mean_token_accuracy": 0.20572818368673323,
      "num_tokens": 104633868.0,
      "step": 41275
    },
    {
      "entropy": 6.115200805664062,
      "epoch": 4.7639930755914595,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002959812809242592,
      "loss": 5.535,
      "mean_token_accuracy": 0.2111074447631836,
      "num_tokens": 104646846.0,
      "step": 41280
    },
    {
      "entropy": 6.069597673416138,
      "epoch": 4.764570109636469,
      "grad_norm": 1.0,
      "learning_rate": 0.00029594019602042533,
      "loss": 5.5963,
      "mean_token_accuracy": 0.20269977897405625,
      "num_tokens": 104658957.0,
      "step": 41285
    },
    {
      "entropy": 6.107734823226929,
      "epoch": 4.765147143681477,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029589911041227835,
      "loss": 5.4872,
      "mean_token_accuracy": 0.2133088916540146,
      "num_tokens": 104670707.0,
      "step": 41290
    },
    {
      "entropy": 6.160231351852417,
      "epoch": 4.765724177726486,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029585802410120023,
      "loss": 5.5981,
      "mean_token_accuracy": 0.21329203397035598,
      "num_tokens": 104683689.0,
      "step": 41295
    },
    {
      "entropy": 6.154626893997192,
      "epoch": 4.766301211771495,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029581693708857277,
      "loss": 5.5321,
      "mean_token_accuracy": 0.21114071756601333,
      "num_tokens": 104696398.0,
      "step": 41300
    },
    {
      "entropy": 5.91089243888855,
      "epoch": 4.766878245816503,
      "grad_norm": 0.984375,
      "learning_rate": 0.000295775849375778,
      "loss": 5.3818,
      "mean_token_accuracy": 0.2205381453037262,
      "num_tokens": 104708212.0,
      "step": 41305
    },
    {
      "entropy": 6.053814888000488,
      "epoch": 4.767455279861512,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0002957347609641979,
      "loss": 5.5167,
      "mean_token_accuracy": 0.21384516060352327,
      "num_tokens": 104722215.0,
      "step": 41310
    },
    {
      "entropy": 6.095475149154663,
      "epoch": 4.76803231390652,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029569367185521434,
      "loss": 5.5733,
      "mean_token_accuracy": 0.20726267248392105,
      "num_tokens": 104733585.0,
      "step": 41315
    },
    {
      "entropy": 6.082965755462647,
      "epoch": 4.7686093479515295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029565258205020944,
      "loss": 5.448,
      "mean_token_accuracy": 0.22041648328304292,
      "num_tokens": 104745147.0,
      "step": 41320
    },
    {
      "entropy": 5.899153137207032,
      "epoch": 4.769186381996538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002956114915505652,
      "loss": 5.3914,
      "mean_token_accuracy": 0.2258801519870758,
      "num_tokens": 104758316.0,
      "step": 41325
    },
    {
      "entropy": 6.024810552597046,
      "epoch": 4.769763416041546,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002955704003576637,
      "loss": 5.51,
      "mean_token_accuracy": 0.2152888759970665,
      "num_tokens": 104769838.0,
      "step": 41330
    },
    {
      "entropy": 6.080876398086548,
      "epoch": 4.770340450086556,
      "grad_norm": 1.109375,
      "learning_rate": 0.000295529308472887,
      "loss": 5.4944,
      "mean_token_accuracy": 0.21570412814617157,
      "num_tokens": 104781925.0,
      "step": 41335
    },
    {
      "entropy": 6.083562135696411,
      "epoch": 4.770917484131564,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029548821589761717,
      "loss": 5.5594,
      "mean_token_accuracy": 0.21082534193992614,
      "num_tokens": 104795762.0,
      "step": 41340
    },
    {
      "entropy": 6.031008720397949,
      "epoch": 4.771494518176572,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029544712263323636,
      "loss": 5.3941,
      "mean_token_accuracy": 0.2175571858882904,
      "num_tokens": 104807073.0,
      "step": 41345
    },
    {
      "entropy": 5.932028532028198,
      "epoch": 4.772071552221581,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002954060286811268,
      "loss": 5.3427,
      "mean_token_accuracy": 0.23438269197940825,
      "num_tokens": 104819816.0,
      "step": 41350
    },
    {
      "entropy": 5.9605449676513675,
      "epoch": 4.772648586266589,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002953649340426706,
      "loss": 5.4292,
      "mean_token_accuracy": 0.2205592140555382,
      "num_tokens": 104833067.0,
      "step": 41355
    },
    {
      "entropy": 6.028792190551758,
      "epoch": 4.773225620311599,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002953238387192499,
      "loss": 5.4552,
      "mean_token_accuracy": 0.21971480995416642,
      "num_tokens": 104845844.0,
      "step": 41360
    },
    {
      "entropy": 6.134691286087036,
      "epoch": 4.773802654356607,
      "grad_norm": 0.984375,
      "learning_rate": 0.000295282742712247,
      "loss": 5.5621,
      "mean_token_accuracy": 0.20757120251655578,
      "num_tokens": 104859291.0,
      "step": 41365
    },
    {
      "entropy": 6.0594720363616945,
      "epoch": 4.774379688401615,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002952416460230442,
      "loss": 5.5259,
      "mean_token_accuracy": 0.2145521342754364,
      "num_tokens": 104872761.0,
      "step": 41370
    },
    {
      "entropy": 6.093132591247558,
      "epoch": 4.774956722446625,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002952005486530236,
      "loss": 5.5442,
      "mean_token_accuracy": 0.21537879407405852,
      "num_tokens": 104885268.0,
      "step": 41375
    },
    {
      "entropy": 6.176255130767823,
      "epoch": 4.775533756491633,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002951594506035676,
      "loss": 5.5909,
      "mean_token_accuracy": 0.20619026720523834,
      "num_tokens": 104897443.0,
      "step": 41380
    },
    {
      "entropy": 6.135396432876587,
      "epoch": 4.776110790536642,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002951183518760585,
      "loss": 5.5959,
      "mean_token_accuracy": 0.20602556318044662,
      "num_tokens": 104910199.0,
      "step": 41385
    },
    {
      "entropy": 6.0705383777618405,
      "epoch": 4.77668782458165,
      "grad_norm": 1.0,
      "learning_rate": 0.00029507725247187867,
      "loss": 5.5288,
      "mean_token_accuracy": 0.21240886002779008,
      "num_tokens": 104922713.0,
      "step": 41390
    },
    {
      "entropy": 6.025086879730225,
      "epoch": 4.777264858626659,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029503615239241036,
      "loss": 5.4602,
      "mean_token_accuracy": 0.21857306510210037,
      "num_tokens": 104935399.0,
      "step": 41395
    },
    {
      "entropy": 6.074596881866455,
      "epoch": 4.777841892671668,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002949950516390361,
      "loss": 5.5013,
      "mean_token_accuracy": 0.21360948979854583,
      "num_tokens": 104947595.0,
      "step": 41400
    },
    {
      "entropy": 5.992557144165039,
      "epoch": 4.778418926716676,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029495395021313813,
      "loss": 5.4083,
      "mean_token_accuracy": 0.21428395360708236,
      "num_tokens": 104960278.0,
      "step": 41405
    },
    {
      "entropy": 6.025417709350586,
      "epoch": 4.778995960761685,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000294912848116099,
      "loss": 5.4173,
      "mean_token_accuracy": 0.22592977732419967,
      "num_tokens": 104972766.0,
      "step": 41410
    },
    {
      "entropy": 6.184357357025147,
      "epoch": 4.779572994806694,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029487174534930104,
      "loss": 5.6082,
      "mean_token_accuracy": 0.2092226043343544,
      "num_tokens": 104986361.0,
      "step": 41415
    },
    {
      "entropy": 5.976289367675781,
      "epoch": 4.780150028851702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002948306419141269,
      "loss": 5.3971,
      "mean_token_accuracy": 0.22283859699964523,
      "num_tokens": 104998449.0,
      "step": 41420
    },
    {
      "entropy": 6.0617103099823,
      "epoch": 4.780727062896711,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002947895378119589,
      "loss": 5.5593,
      "mean_token_accuracy": 0.20727822929620743,
      "num_tokens": 105010661.0,
      "step": 41425
    },
    {
      "entropy": 6.135581827163696,
      "epoch": 4.781304096941719,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029474843304417957,
      "loss": 5.5758,
      "mean_token_accuracy": 0.2071564868092537,
      "num_tokens": 105022557.0,
      "step": 41430
    },
    {
      "entropy": 6.120849609375,
      "epoch": 4.781881130986728,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029470732761217167,
      "loss": 5.5602,
      "mean_token_accuracy": 0.21098630726337433,
      "num_tokens": 105033873.0,
      "step": 41435
    },
    {
      "entropy": 5.958856248855591,
      "epoch": 4.782458165031737,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00029466622151731747,
      "loss": 5.3475,
      "mean_token_accuracy": 0.22708093076944352,
      "num_tokens": 105047006.0,
      "step": 41440
    },
    {
      "entropy": 6.025007057189941,
      "epoch": 4.783035199076745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002946251147609997,
      "loss": 5.4787,
      "mean_token_accuracy": 0.21694047451019288,
      "num_tokens": 105058269.0,
      "step": 41445
    },
    {
      "entropy": 6.091573572158813,
      "epoch": 4.7836122331217545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029458400734460097,
      "loss": 5.4496,
      "mean_token_accuracy": 0.2201585814356804,
      "num_tokens": 105071489.0,
      "step": 41450
    },
    {
      "entropy": 6.081538105010987,
      "epoch": 4.784189267166763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002945428992695039,
      "loss": 5.4953,
      "mean_token_accuracy": 0.21984046697616577,
      "num_tokens": 105083729.0,
      "step": 41455
    },
    {
      "entropy": 6.01912636756897,
      "epoch": 4.784766301211771,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0002945017905370911,
      "loss": 5.5144,
      "mean_token_accuracy": 0.2168918326497078,
      "num_tokens": 105097091.0,
      "step": 41460
    },
    {
      "entropy": 5.99334602355957,
      "epoch": 4.78534333525678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029446068114874523,
      "loss": 5.4633,
      "mean_token_accuracy": 0.21733350604772567,
      "num_tokens": 105108298.0,
      "step": 41465
    },
    {
      "entropy": 6.097963333129883,
      "epoch": 4.785920369301789,
      "grad_norm": 0.984375,
      "learning_rate": 0.000294419571105849,
      "loss": 5.5477,
      "mean_token_accuracy": 0.21434574127197265,
      "num_tokens": 105121276.0,
      "step": 41470
    },
    {
      "entropy": 6.149650573730469,
      "epoch": 4.7864974033467975,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002943784604097852,
      "loss": 5.6099,
      "mean_token_accuracy": 0.20802673995494841,
      "num_tokens": 105132970.0,
      "step": 41475
    },
    {
      "entropy": 6.031307792663574,
      "epoch": 4.787074437391806,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002943373490619365,
      "loss": 5.5006,
      "mean_token_accuracy": 0.21768438518047334,
      "num_tokens": 105146247.0,
      "step": 41480
    },
    {
      "entropy": 6.0566246032714846,
      "epoch": 4.787651471436815,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002942962370636856,
      "loss": 5.5398,
      "mean_token_accuracy": 0.21586243510246278,
      "num_tokens": 105160490.0,
      "step": 41485
    },
    {
      "entropy": 6.0782383441925045,
      "epoch": 4.788228505481824,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002942551244164154,
      "loss": 5.5135,
      "mean_token_accuracy": 0.2113226607441902,
      "num_tokens": 105172828.0,
      "step": 41490
    },
    {
      "entropy": 6.04362998008728,
      "epoch": 4.788805539526832,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002942140111215087,
      "loss": 5.4744,
      "mean_token_accuracy": 0.21949948668479918,
      "num_tokens": 105185855.0,
      "step": 41495
    },
    {
      "entropy": 6.111118793487549,
      "epoch": 4.7893825735718405,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002941728971803483,
      "loss": 5.5255,
      "mean_token_accuracy": 0.2162104919552803,
      "num_tokens": 105197598.0,
      "step": 41500
    },
    {
      "entropy": 6.049832344055176,
      "epoch": 4.789959607616849,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002941317825943169,
      "loss": 5.4874,
      "mean_token_accuracy": 0.21283673495054245,
      "num_tokens": 105209179.0,
      "step": 41505
    },
    {
      "entropy": 5.980673933029175,
      "epoch": 4.790536641661858,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002940906673647976,
      "loss": 5.3883,
      "mean_token_accuracy": 0.2258274033665657,
      "num_tokens": 105222565.0,
      "step": 41510
    },
    {
      "entropy": 5.91896653175354,
      "epoch": 4.791113675706867,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002940495514931731,
      "loss": 5.3778,
      "mean_token_accuracy": 0.22174662202596665,
      "num_tokens": 105234708.0,
      "step": 41515
    },
    {
      "entropy": 6.088687229156494,
      "epoch": 4.791690709751875,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029400843498082654,
      "loss": 5.512,
      "mean_token_accuracy": 0.2184618815779686,
      "num_tokens": 105246265.0,
      "step": 41520
    },
    {
      "entropy": 6.053173589706421,
      "epoch": 4.792267743796884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002939673178291406,
      "loss": 5.4935,
      "mean_token_accuracy": 0.21846381723880767,
      "num_tokens": 105258512.0,
      "step": 41525
    },
    {
      "entropy": 6.0618218898773195,
      "epoch": 4.792844777841893,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002939262000394984,
      "loss": 5.4616,
      "mean_token_accuracy": 0.21793500781059266,
      "num_tokens": 105270020.0,
      "step": 41530
    },
    {
      "entropy": 6.00270676612854,
      "epoch": 4.793421811886901,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00029388508161328283,
      "loss": 5.5347,
      "mean_token_accuracy": 0.21377929896116257,
      "num_tokens": 105283388.0,
      "step": 41535
    },
    {
      "entropy": 6.119080591201782,
      "epoch": 4.79399884593191,
      "grad_norm": 1.0,
      "learning_rate": 0.0002938439625518769,
      "loss": 5.5872,
      "mean_token_accuracy": 0.20289211869239807,
      "num_tokens": 105296644.0,
      "step": 41540
    },
    {
      "entropy": 6.114306449890137,
      "epoch": 4.794575879976919,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029380284285666366,
      "loss": 5.5289,
      "mean_token_accuracy": 0.21145639717578887,
      "num_tokens": 105310268.0,
      "step": 41545
    },
    {
      "entropy": 6.04633092880249,
      "epoch": 4.795152914021927,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002937617225290261,
      "loss": 5.4764,
      "mean_token_accuracy": 0.21371156573295594,
      "num_tokens": 105323180.0,
      "step": 41550
    },
    {
      "entropy": 6.04849100112915,
      "epoch": 4.795729948066936,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029372060157034744,
      "loss": 5.4599,
      "mean_token_accuracy": 0.21785809993743896,
      "num_tokens": 105336206.0,
      "step": 41555
    },
    {
      "entropy": 6.0739411354064945,
      "epoch": 4.796306982111945,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002936794799820105,
      "loss": 5.6084,
      "mean_token_accuracy": 0.21283904016017913,
      "num_tokens": 105348759.0,
      "step": 41560
    },
    {
      "entropy": 5.9736074924469,
      "epoch": 4.7968840161569535,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029363835776539867,
      "loss": 5.3722,
      "mean_token_accuracy": 0.23410444408655168,
      "num_tokens": 105361318.0,
      "step": 41565
    },
    {
      "entropy": 6.09433913230896,
      "epoch": 4.797461050201962,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029359723492189486,
      "loss": 5.5378,
      "mean_token_accuracy": 0.2156742572784424,
      "num_tokens": 105374536.0,
      "step": 41570
    },
    {
      "entropy": 6.020211172103882,
      "epoch": 4.79803808424697,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002935561114528823,
      "loss": 5.4753,
      "mean_token_accuracy": 0.22142890840768814,
      "num_tokens": 105386972.0,
      "step": 41575
    },
    {
      "entropy": 6.014218759536743,
      "epoch": 4.798615118291979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002935149873597442,
      "loss": 5.4352,
      "mean_token_accuracy": 0.21854821145534514,
      "num_tokens": 105399898.0,
      "step": 41580
    },
    {
      "entropy": 6.062388706207275,
      "epoch": 4.799192152336988,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002934738626438636,
      "loss": 5.5497,
      "mean_token_accuracy": 0.20619071424007415,
      "num_tokens": 105412039.0,
      "step": 41585
    },
    {
      "entropy": 6.112994432449341,
      "epoch": 4.7997691863819965,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029343273730662384,
      "loss": 5.5497,
      "mean_token_accuracy": 0.20765966325998306,
      "num_tokens": 105423735.0,
      "step": 41590
    },
    {
      "entropy": 6.000195646286011,
      "epoch": 4.800346220427005,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002933916113494081,
      "loss": 5.4678,
      "mean_token_accuracy": 0.21727635562419892,
      "num_tokens": 105435753.0,
      "step": 41595
    },
    {
      "entropy": 6.107688093185425,
      "epoch": 4.800923254472014,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029335048477359974,
      "loss": 5.5165,
      "mean_token_accuracy": 0.20927537977695465,
      "num_tokens": 105446925.0,
      "step": 41600
    },
    {
      "entropy": 6.10490665435791,
      "epoch": 4.801500288517023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002933093575805819,
      "loss": 5.5291,
      "mean_token_accuracy": 0.21377342492341994,
      "num_tokens": 105459764.0,
      "step": 41605
    },
    {
      "entropy": 6.066596460342407,
      "epoch": 4.802077322562031,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002932682297717379,
      "loss": 5.4906,
      "mean_token_accuracy": 0.21600044667720794,
      "num_tokens": 105474508.0,
      "step": 41610
    },
    {
      "entropy": 6.020422315597534,
      "epoch": 4.802654356607039,
      "grad_norm": 1.0,
      "learning_rate": 0.000293227101348451,
      "loss": 5.4502,
      "mean_token_accuracy": 0.22096077054739,
      "num_tokens": 105488679.0,
      "step": 41615
    },
    {
      "entropy": 6.084559202194214,
      "epoch": 4.803231390652049,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002931859723121047,
      "loss": 5.4825,
      "mean_token_accuracy": 0.21959920674562455,
      "num_tokens": 105501859.0,
      "step": 41620
    },
    {
      "entropy": 6.030162286758423,
      "epoch": 4.803808424697057,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002931448426640822,
      "loss": 5.4456,
      "mean_token_accuracy": 0.21540820002555847,
      "num_tokens": 105514757.0,
      "step": 41625
    },
    {
      "entropy": 5.99525408744812,
      "epoch": 4.804385458742066,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000293103712405767,
      "loss": 5.4925,
      "mean_token_accuracy": 0.21668670028448106,
      "num_tokens": 105526747.0,
      "step": 41630
    },
    {
      "entropy": 6.059666109085083,
      "epoch": 4.804962492787075,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002930625815385424,
      "loss": 5.4937,
      "mean_token_accuracy": 0.21525411754846574,
      "num_tokens": 105539892.0,
      "step": 41635
    },
    {
      "entropy": 6.075288391113281,
      "epoch": 4.805539526832083,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002930214500637917,
      "loss": 5.4666,
      "mean_token_accuracy": 0.2126723572611809,
      "num_tokens": 105552958.0,
      "step": 41640
    },
    {
      "entropy": 5.961076736450195,
      "epoch": 4.806116560877092,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002929803179828986,
      "loss": 5.4337,
      "mean_token_accuracy": 0.219391368329525,
      "num_tokens": 105565617.0,
      "step": 41645
    },
    {
      "entropy": 6.092890930175781,
      "epoch": 4.8066935949221,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002929391852972465,
      "loss": 5.5209,
      "mean_token_accuracy": 0.21151095479726792,
      "num_tokens": 105578644.0,
      "step": 41650
    },
    {
      "entropy": 6.0649425983428955,
      "epoch": 4.807270628967109,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029289805200821874,
      "loss": 5.5233,
      "mean_token_accuracy": 0.21722752898931502,
      "num_tokens": 105591833.0,
      "step": 41655
    },
    {
      "entropy": 6.060564374923706,
      "epoch": 4.807847663012118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029285691811719883,
      "loss": 5.4705,
      "mean_token_accuracy": 0.21344552636146547,
      "num_tokens": 105602644.0,
      "step": 41660
    },
    {
      "entropy": 6.047170495986938,
      "epoch": 4.808424697057126,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029281578362557045,
      "loss": 5.4983,
      "mean_token_accuracy": 0.21394772678613663,
      "num_tokens": 105615392.0,
      "step": 41665
    },
    {
      "entropy": 6.023516368865967,
      "epoch": 4.809001731102135,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029277464853471696,
      "loss": 5.4865,
      "mean_token_accuracy": 0.21132300347089766,
      "num_tokens": 105628227.0,
      "step": 41670
    },
    {
      "entropy": 6.008621788024902,
      "epoch": 4.809578765147144,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029273351284602204,
      "loss": 5.4758,
      "mean_token_accuracy": 0.21179116517305374,
      "num_tokens": 105640587.0,
      "step": 41675
    },
    {
      "entropy": 6.045520448684693,
      "epoch": 4.810155799192152,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029269237656086923,
      "loss": 5.4284,
      "mean_token_accuracy": 0.2278468430042267,
      "num_tokens": 105653131.0,
      "step": 41680
    },
    {
      "entropy": 6.005474328994751,
      "epoch": 4.810732833237161,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002926512396806421,
      "loss": 5.4152,
      "mean_token_accuracy": 0.22410652041435242,
      "num_tokens": 105666303.0,
      "step": 41685
    },
    {
      "entropy": 6.016945743560791,
      "epoch": 4.811309867282169,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00029261010220672437,
      "loss": 5.5054,
      "mean_token_accuracy": 0.21547868549823762,
      "num_tokens": 105678776.0,
      "step": 41690
    },
    {
      "entropy": 6.053600406646728,
      "epoch": 4.811886901327179,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002925689641404995,
      "loss": 5.5078,
      "mean_token_accuracy": 0.21156597137451172,
      "num_tokens": 105692403.0,
      "step": 41695
    },
    {
      "entropy": 6.0752159595489506,
      "epoch": 4.812463935372187,
      "grad_norm": 1.0,
      "learning_rate": 0.00029252782548335127,
      "loss": 5.5587,
      "mean_token_accuracy": 0.207550747692585,
      "num_tokens": 105704505.0,
      "step": 41700
    },
    {
      "entropy": 6.037029600143432,
      "epoch": 4.813040969417195,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029248668623666336,
      "loss": 5.409,
      "mean_token_accuracy": 0.2211407944560051,
      "num_tokens": 105717250.0,
      "step": 41705
    },
    {
      "entropy": 6.090919733047485,
      "epoch": 4.813618003462205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002924455464018194,
      "loss": 5.5354,
      "mean_token_accuracy": 0.21743592321872712,
      "num_tokens": 105729208.0,
      "step": 41710
    },
    {
      "entropy": 6.0492791652679445,
      "epoch": 4.814195037507213,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00029240440598020325,
      "loss": 5.4572,
      "mean_token_accuracy": 0.22621232122182847,
      "num_tokens": 105741727.0,
      "step": 41715
    },
    {
      "entropy": 6.051457023620605,
      "epoch": 4.8147720715522215,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002923632649731984,
      "loss": 5.5007,
      "mean_token_accuracy": 0.21788737028837205,
      "num_tokens": 105753349.0,
      "step": 41720
    },
    {
      "entropy": 6.037710857391358,
      "epoch": 4.81534910559723,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002923221233821888,
      "loss": 5.4983,
      "mean_token_accuracy": 0.21543957144021988,
      "num_tokens": 105764915.0,
      "step": 41725
    },
    {
      "entropy": 5.956029844284058,
      "epoch": 4.815926139642239,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002922809812085583,
      "loss": 5.4408,
      "mean_token_accuracy": 0.22386123090982438,
      "num_tokens": 105779055.0,
      "step": 41730
    },
    {
      "entropy": 6.117808675765991,
      "epoch": 4.816503173687248,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002922398384536905,
      "loss": 5.5681,
      "mean_token_accuracy": 0.2072658881545067,
      "num_tokens": 105792307.0,
      "step": 41735
    },
    {
      "entropy": 6.140282201766968,
      "epoch": 4.817080207732256,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002921986951189693,
      "loss": 5.5109,
      "mean_token_accuracy": 0.21260603070259093,
      "num_tokens": 105805995.0,
      "step": 41740
    },
    {
      "entropy": 6.092316770553589,
      "epoch": 4.8176572417772645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029215755120577853,
      "loss": 5.5303,
      "mean_token_accuracy": 0.21037748456001282,
      "num_tokens": 105817961.0,
      "step": 41745
    },
    {
      "entropy": 6.042257165908813,
      "epoch": 4.818234275822274,
      "grad_norm": 0.984375,
      "learning_rate": 0.000292116406715502,
      "loss": 5.5017,
      "mean_token_accuracy": 0.21728186160326005,
      "num_tokens": 105829051.0,
      "step": 41750
    },
    {
      "entropy": 6.023566961288452,
      "epoch": 4.818811309867282,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029207526164952375,
      "loss": 5.476,
      "mean_token_accuracy": 0.21348415911197663,
      "num_tokens": 105841370.0,
      "step": 41755
    },
    {
      "entropy": 6.0289342403411865,
      "epoch": 4.819388343912291,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00029203411600922744,
      "loss": 5.4397,
      "mean_token_accuracy": 0.22370845675468445,
      "num_tokens": 105854895.0,
      "step": 41760
    },
    {
      "entropy": 6.034477281570434,
      "epoch": 4.819965377957299,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029199296979599723,
      "loss": 5.5269,
      "mean_token_accuracy": 0.20602592825889587,
      "num_tokens": 105866650.0,
      "step": 41765
    },
    {
      "entropy": 6.10467209815979,
      "epoch": 4.820542412002308,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029195182301121683,
      "loss": 5.5568,
      "mean_token_accuracy": 0.21620430946350097,
      "num_tokens": 105878682.0,
      "step": 41770
    },
    {
      "entropy": 6.048243761062622,
      "epoch": 4.821119446047317,
      "grad_norm": 0.84765625,
      "learning_rate": 0.00029191067565627034,
      "loss": 5.4833,
      "mean_token_accuracy": 0.22307748198509217,
      "num_tokens": 105892127.0,
      "step": 41775
    },
    {
      "entropy": 6.030547142028809,
      "epoch": 4.821696480092325,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029186952773254163,
      "loss": 5.45,
      "mean_token_accuracy": 0.21983271688222886,
      "num_tokens": 105904634.0,
      "step": 41780
    },
    {
      "entropy": 6.097190904617309,
      "epoch": 4.8222735141373345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029182837924141464,
      "loss": 5.5553,
      "mean_token_accuracy": 0.20591573119163514,
      "num_tokens": 105917267.0,
      "step": 41785
    },
    {
      "entropy": 6.059438848495484,
      "epoch": 4.822850548182343,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002917872301842736,
      "loss": 5.5112,
      "mean_token_accuracy": 0.21491246223449706,
      "num_tokens": 105930612.0,
      "step": 41790
    },
    {
      "entropy": 6.121038198471069,
      "epoch": 4.823427582227351,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002917460805625024,
      "loss": 5.5893,
      "mean_token_accuracy": 0.20605832785367967,
      "num_tokens": 105943162.0,
      "step": 41795
    },
    {
      "entropy": 6.0964601039886475,
      "epoch": 4.82400461627236,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002917049303774851,
      "loss": 5.5121,
      "mean_token_accuracy": 0.21484218537807465,
      "num_tokens": 105955907.0,
      "step": 41800
    },
    {
      "entropy": 6.092317819595337,
      "epoch": 4.824581650317369,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029166377963060574,
      "loss": 5.5166,
      "mean_token_accuracy": 0.2166803926229477,
      "num_tokens": 105970078.0,
      "step": 41805
    },
    {
      "entropy": 6.027702856063843,
      "epoch": 4.8251586843623775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002916226283232484,
      "loss": 5.4195,
      "mean_token_accuracy": 0.21365920156240464,
      "num_tokens": 105981817.0,
      "step": 41810
    },
    {
      "entropy": 5.950418281555176,
      "epoch": 4.825735718407386,
      "grad_norm": 0.921875,
      "learning_rate": 0.00029158147645679726,
      "loss": 5.3022,
      "mean_token_accuracy": 0.23412099331617356,
      "num_tokens": 105994146.0,
      "step": 41815
    },
    {
      "entropy": 6.082836294174195,
      "epoch": 4.826312752452394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002915403240326364,
      "loss": 5.5549,
      "mean_token_accuracy": 0.21540736258029938,
      "num_tokens": 106007120.0,
      "step": 41820
    },
    {
      "entropy": 6.000719499588013,
      "epoch": 4.826889786497404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029149917105214995,
      "loss": 5.4785,
      "mean_token_accuracy": 0.22026052474975585,
      "num_tokens": 106021807.0,
      "step": 41825
    },
    {
      "entropy": 6.106930923461914,
      "epoch": 4.827466820542412,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029145801751672204,
      "loss": 5.5743,
      "mean_token_accuracy": 0.20442667454481125,
      "num_tokens": 106034250.0,
      "step": 41830
    },
    {
      "entropy": 6.142507934570313,
      "epoch": 4.8280438545874205,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00029141686342773693,
      "loss": 5.5439,
      "mean_token_accuracy": 0.21021172404289246,
      "num_tokens": 106048507.0,
      "step": 41835
    },
    {
      "entropy": 5.977290153503418,
      "epoch": 4.828620888632429,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029137570878657877,
      "loss": 5.3779,
      "mean_token_accuracy": 0.23080142587423325,
      "num_tokens": 106060923.0,
      "step": 41840
    },
    {
      "entropy": 6.045567178726197,
      "epoch": 4.829197922677438,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029133455359463175,
      "loss": 5.4597,
      "mean_token_accuracy": 0.21228379756212234,
      "num_tokens": 106073518.0,
      "step": 41845
    },
    {
      "entropy": 5.915381193161011,
      "epoch": 4.829774956722447,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002912933978532802,
      "loss": 5.3213,
      "mean_token_accuracy": 0.22838466316461564,
      "num_tokens": 106086453.0,
      "step": 41850
    },
    {
      "entropy": 5.9941082954406735,
      "epoch": 4.830351990767455,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002912522415639082,
      "loss": 5.4301,
      "mean_token_accuracy": 0.2179124668240547,
      "num_tokens": 106099086.0,
      "step": 41855
    },
    {
      "entropy": 6.013785266876221,
      "epoch": 4.830929024812464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002912110847279002,
      "loss": 5.5111,
      "mean_token_accuracy": 0.21364900171756745,
      "num_tokens": 106112964.0,
      "step": 41860
    },
    {
      "entropy": 6.029106664657593,
      "epoch": 4.831506058857473,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002911699273466403,
      "loss": 5.4382,
      "mean_token_accuracy": 0.2227054789662361,
      "num_tokens": 106126915.0,
      "step": 41865
    },
    {
      "entropy": 6.107553482055664,
      "epoch": 4.832083092902481,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029112876942151314,
      "loss": 5.5537,
      "mean_token_accuracy": 0.2108403339982033,
      "num_tokens": 106139704.0,
      "step": 41870
    },
    {
      "entropy": 6.074220895767212,
      "epoch": 4.83266012694749,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002910876109539027,
      "loss": 5.4674,
      "mean_token_accuracy": 0.2125176638364792,
      "num_tokens": 106152352.0,
      "step": 41875
    },
    {
      "entropy": 6.032215929031372,
      "epoch": 4.833237160992499,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029104645194519345,
      "loss": 5.4944,
      "mean_token_accuracy": 0.22198151797056198,
      "num_tokens": 106165333.0,
      "step": 41880
    },
    {
      "entropy": 6.142093706130981,
      "epoch": 4.833814195037507,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002910052923967698,
      "loss": 5.5704,
      "mean_token_accuracy": 0.21293933987617492,
      "num_tokens": 106177522.0,
      "step": 41885
    },
    {
      "entropy": 6.072480821609497,
      "epoch": 4.834391229082516,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002909641323100161,
      "loss": 5.5168,
      "mean_token_accuracy": 0.20957802683115007,
      "num_tokens": 106189969.0,
      "step": 41890
    },
    {
      "entropy": 6.072681617736817,
      "epoch": 4.834968263127524,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002909229716863167,
      "loss": 5.516,
      "mean_token_accuracy": 0.21778736561536788,
      "num_tokens": 106202016.0,
      "step": 41895
    },
    {
      "entropy": 6.030926895141602,
      "epoch": 4.8355452971725335,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002908818105270561,
      "loss": 5.4172,
      "mean_token_accuracy": 0.22090569734573365,
      "num_tokens": 106214558.0,
      "step": 41900
    },
    {
      "entropy": 6.065970706939697,
      "epoch": 4.836122331217542,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0002908406488336186,
      "loss": 5.507,
      "mean_token_accuracy": 0.2176358923316002,
      "num_tokens": 106227845.0,
      "step": 41905
    },
    {
      "entropy": 6.102239036560059,
      "epoch": 4.83669936526255,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029079948660738883,
      "loss": 5.5083,
      "mean_token_accuracy": 0.2149672254920006,
      "num_tokens": 106240592.0,
      "step": 41910
    },
    {
      "entropy": 6.102728366851807,
      "epoch": 4.837276399307559,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002907583238497512,
      "loss": 5.5413,
      "mean_token_accuracy": 0.20700691789388656,
      "num_tokens": 106254902.0,
      "step": 41915
    },
    {
      "entropy": 6.133925247192383,
      "epoch": 4.837853433352568,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002907171605620901,
      "loss": 5.5952,
      "mean_token_accuracy": 0.20930389165878296,
      "num_tokens": 106268142.0,
      "step": 41920
    },
    {
      "entropy": 6.102343606948852,
      "epoch": 4.838430467397576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029067599674579013,
      "loss": 5.4961,
      "mean_token_accuracy": 0.21084205359220504,
      "num_tokens": 106279177.0,
      "step": 41925
    },
    {
      "entropy": 6.131048059463501,
      "epoch": 4.839007501442585,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002906348324022358,
      "loss": 5.6683,
      "mean_token_accuracy": 0.20389731377363204,
      "num_tokens": 106291526.0,
      "step": 41930
    },
    {
      "entropy": 6.0369102478027346,
      "epoch": 4.839584535487594,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002905936675328117,
      "loss": 5.4636,
      "mean_token_accuracy": 0.21381203830242157,
      "num_tokens": 106305520.0,
      "step": 41935
    },
    {
      "entropy": 6.152507972717285,
      "epoch": 4.840161569532603,
      "grad_norm": 1.0,
      "learning_rate": 0.00029055250213890226,
      "loss": 5.4809,
      "mean_token_accuracy": 0.21613842397928237,
      "num_tokens": 106316947.0,
      "step": 41940
    },
    {
      "entropy": 6.043904113769531,
      "epoch": 4.840738603577611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029051133622189215,
      "loss": 5.5282,
      "mean_token_accuracy": 0.21606215536594392,
      "num_tokens": 106329961.0,
      "step": 41945
    },
    {
      "entropy": 6.02126407623291,
      "epoch": 4.841315637622619,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00029047016978316596,
      "loss": 5.5217,
      "mean_token_accuracy": 0.21577370017766953,
      "num_tokens": 106342369.0,
      "step": 41950
    },
    {
      "entropy": 6.136128950119018,
      "epoch": 4.841892671667629,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029042900282410833,
      "loss": 5.5612,
      "mean_token_accuracy": 0.20706587582826613,
      "num_tokens": 106353810.0,
      "step": 41955
    },
    {
      "entropy": 6.052432537078857,
      "epoch": 4.842469705712637,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029038783534610384,
      "loss": 5.4642,
      "mean_token_accuracy": 0.21803044825792312,
      "num_tokens": 106365920.0,
      "step": 41960
    },
    {
      "entropy": 6.064763593673706,
      "epoch": 4.8430467397576455,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002903466673505371,
      "loss": 5.4617,
      "mean_token_accuracy": 0.22036785036325454,
      "num_tokens": 106377959.0,
      "step": 41965
    },
    {
      "entropy": 6.029982423782348,
      "epoch": 4.843623773802655,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002903054988387928,
      "loss": 5.5568,
      "mean_token_accuracy": 0.22074573785066604,
      "num_tokens": 106391106.0,
      "step": 41970
    },
    {
      "entropy": 6.103996276855469,
      "epoch": 4.844200807847663,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029026432981225583,
      "loss": 5.4919,
      "mean_token_accuracy": 0.21445810049772263,
      "num_tokens": 106402958.0,
      "step": 41975
    },
    {
      "entropy": 6.130579900741577,
      "epoch": 4.844777841892672,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029022316027231055,
      "loss": 5.5813,
      "mean_token_accuracy": 0.21052951812744142,
      "num_tokens": 106415710.0,
      "step": 41980
    },
    {
      "entropy": 6.053104591369629,
      "epoch": 4.84535487593768,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002901819902203419,
      "loss": 5.478,
      "mean_token_accuracy": 0.22126443982124328,
      "num_tokens": 106429004.0,
      "step": 41985
    },
    {
      "entropy": 6.061376333236694,
      "epoch": 4.8459319099826885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029014081965773457,
      "loss": 5.5617,
      "mean_token_accuracy": 0.21525579541921616,
      "num_tokens": 106441590.0,
      "step": 41990
    },
    {
      "entropy": 6.005907726287842,
      "epoch": 4.846508944027698,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029009964858587326,
      "loss": 5.4526,
      "mean_token_accuracy": 0.21998029947280884,
      "num_tokens": 106454567.0,
      "step": 41995
    },
    {
      "entropy": 6.065983867645263,
      "epoch": 4.847085978072706,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029005847700614275,
      "loss": 5.5299,
      "mean_token_accuracy": 0.2138586938381195,
      "num_tokens": 106468022.0,
      "step": 42000
    },
    {
      "epoch": 4.847085978072706,
      "eval_entropy": 5.868718833451519,
      "eval_loss": 5.651386260986328,
      "eval_mean_token_accuracy": 0.21457421298109403,
      "eval_num_tokens": 106468022.0,
      "eval_runtime": 23.0621,
      "eval_samples_per_second": 1220.01,
      "eval_steps_per_second": 152.501,
      "step": 42000
    },
    {
      "entropy": 6.01589789390564,
      "epoch": 4.847663012117715,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00029001730491992785,
      "loss": 5.4704,
      "mean_token_accuracy": 0.22318455576896667,
      "num_tokens": 106481191.0,
      "step": 42005
    },
    {
      "entropy": 6.032896566390991,
      "epoch": 4.848240046162724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028997613232861335,
      "loss": 5.4867,
      "mean_token_accuracy": 0.2166577696800232,
      "num_tokens": 106492887.0,
      "step": 42010
    },
    {
      "entropy": 6.0560060977935795,
      "epoch": 4.848817080207732,
      "grad_norm": 1.0,
      "learning_rate": 0.0002899349592335841,
      "loss": 5.4865,
      "mean_token_accuracy": 0.21756270974874498,
      "num_tokens": 106506121.0,
      "step": 42015
    },
    {
      "entropy": 6.040227794647217,
      "epoch": 4.849394114252741,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028989378563622486,
      "loss": 5.4693,
      "mean_token_accuracy": 0.2161123275756836,
      "num_tokens": 106518625.0,
      "step": 42020
    },
    {
      "entropy": 6.110466575622558,
      "epoch": 4.849971148297749,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002898526115379206,
      "loss": 5.5449,
      "mean_token_accuracy": 0.2067861005663872,
      "num_tokens": 106532438.0,
      "step": 42025
    },
    {
      "entropy": 6.008976411819458,
      "epoch": 4.8505481823427585,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002898114369400562,
      "loss": 5.4401,
      "mean_token_accuracy": 0.2157757431268692,
      "num_tokens": 106544725.0,
      "step": 42030
    },
    {
      "entropy": 5.973070573806763,
      "epoch": 4.851125216387767,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002897702618440163,
      "loss": 5.4348,
      "mean_token_accuracy": 0.21859724223613738,
      "num_tokens": 106557527.0,
      "step": 42035
    },
    {
      "entropy": 6.100250720977783,
      "epoch": 4.851702250432775,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002897290862511861,
      "loss": 5.5714,
      "mean_token_accuracy": 0.20295725166797637,
      "num_tokens": 106569254.0,
      "step": 42040
    },
    {
      "entropy": 6.081954669952393,
      "epoch": 4.852279284477785,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028968791016295044,
      "loss": 5.4889,
      "mean_token_accuracy": 0.21217907667160035,
      "num_tokens": 106581554.0,
      "step": 42045
    },
    {
      "entropy": 6.087125492095947,
      "epoch": 4.852856318522793,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028964673358069414,
      "loss": 5.531,
      "mean_token_accuracy": 0.20894382297992706,
      "num_tokens": 106594993.0,
      "step": 42050
    },
    {
      "entropy": 5.986484622955322,
      "epoch": 4.8534333525678015,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028960555650580223,
      "loss": 5.3768,
      "mean_token_accuracy": 0.23385561406612396,
      "num_tokens": 106607408.0,
      "step": 42055
    },
    {
      "entropy": 5.986668729782105,
      "epoch": 4.85401038661281,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002895643789396596,
      "loss": 5.5039,
      "mean_token_accuracy": 0.21539508700370788,
      "num_tokens": 106620338.0,
      "step": 42060
    },
    {
      "entropy": 6.07614951133728,
      "epoch": 4.854587420657818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002895232008836515,
      "loss": 5.4915,
      "mean_token_accuracy": 0.2159276932477951,
      "num_tokens": 106632523.0,
      "step": 42065
    },
    {
      "entropy": 6.074080848693848,
      "epoch": 4.855164454702828,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002894820223391627,
      "loss": 5.4465,
      "mean_token_accuracy": 0.22079320549964904,
      "num_tokens": 106646033.0,
      "step": 42070
    },
    {
      "entropy": 5.912979936599731,
      "epoch": 4.855741488747836,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028944084330757814,
      "loss": 5.3442,
      "mean_token_accuracy": 0.2312565639615059,
      "num_tokens": 106658754.0,
      "step": 42075
    },
    {
      "entropy": 5.999143934249878,
      "epoch": 4.8563185227928445,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002893996637902831,
      "loss": 5.5024,
      "mean_token_accuracy": 0.21659717857837676,
      "num_tokens": 106671653.0,
      "step": 42080
    },
    {
      "entropy": 6.176541519165039,
      "epoch": 4.856895556837854,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00028935848378866254,
      "loss": 5.5554,
      "mean_token_accuracy": 0.21208382844924928,
      "num_tokens": 106683241.0,
      "step": 42085
    },
    {
      "entropy": 5.99263710975647,
      "epoch": 4.857472590882862,
      "grad_norm": 0.859375,
      "learning_rate": 0.0002893173033041015,
      "loss": 5.4094,
      "mean_token_accuracy": 0.22358872443437577,
      "num_tokens": 106696702.0,
      "step": 42090
    },
    {
      "entropy": 5.993108606338501,
      "epoch": 4.858049624927871,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000289276122337985,
      "loss": 5.4576,
      "mean_token_accuracy": 0.2163296863436699,
      "num_tokens": 106708588.0,
      "step": 42095
    },
    {
      "entropy": 6.107447910308838,
      "epoch": 4.858626658972879,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002892349408916983,
      "loss": 5.5259,
      "mean_token_accuracy": 0.212727190554142,
      "num_tokens": 106721138.0,
      "step": 42100
    },
    {
      "entropy": 6.028909301757812,
      "epoch": 4.859203693017888,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002891937589666264,
      "loss": 5.4274,
      "mean_token_accuracy": 0.21823488026857377,
      "num_tokens": 106734430.0,
      "step": 42105
    },
    {
      "entropy": 6.0682297706604,
      "epoch": 4.859780727062897,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002891525765641545,
      "loss": 5.5248,
      "mean_token_accuracy": 0.20743386149406434,
      "num_tokens": 106746474.0,
      "step": 42110
    },
    {
      "entropy": 6.177525377273559,
      "epoch": 4.860357761107905,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028911139368566766,
      "loss": 5.5242,
      "mean_token_accuracy": 0.21512916535139084,
      "num_tokens": 106758486.0,
      "step": 42115
    },
    {
      "entropy": 6.060549402236939,
      "epoch": 4.8609347951529145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002890702103325512,
      "loss": 5.4737,
      "mean_token_accuracy": 0.22206600606441498,
      "num_tokens": 106770737.0,
      "step": 42120
    },
    {
      "entropy": 6.089299535751342,
      "epoch": 4.861511829197923,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028902902650619004,
      "loss": 5.6143,
      "mean_token_accuracy": 0.20724078714847566,
      "num_tokens": 106782983.0,
      "step": 42125
    },
    {
      "entropy": 6.085077571868896,
      "epoch": 4.862088863242931,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002889878422079697,
      "loss": 5.5509,
      "mean_token_accuracy": 0.21497012674808502,
      "num_tokens": 106795890.0,
      "step": 42130
    },
    {
      "entropy": 6.077599716186524,
      "epoch": 4.86266589728794,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002889466574392751,
      "loss": 5.4605,
      "mean_token_accuracy": 0.22520923614501953,
      "num_tokens": 106808603.0,
      "step": 42135
    },
    {
      "entropy": 6.052433681488037,
      "epoch": 4.863242931332948,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002889054722014917,
      "loss": 5.5067,
      "mean_token_accuracy": 0.21787955164909362,
      "num_tokens": 106821414.0,
      "step": 42140
    },
    {
      "entropy": 6.053203964233399,
      "epoch": 4.8638199653779575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028886428649600466,
      "loss": 5.493,
      "mean_token_accuracy": 0.21190803945064546,
      "num_tokens": 106833714.0,
      "step": 42145
    },
    {
      "entropy": 6.074897193908692,
      "epoch": 4.864396999422966,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028882310032419915,
      "loss": 5.5547,
      "mean_token_accuracy": 0.21016849875450133,
      "num_tokens": 106846975.0,
      "step": 42150
    },
    {
      "entropy": 6.05290036201477,
      "epoch": 4.864974033467974,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002887819136874606,
      "loss": 5.4482,
      "mean_token_accuracy": 0.21339240074157714,
      "num_tokens": 106859654.0,
      "step": 42155
    },
    {
      "entropy": 6.081863164901733,
      "epoch": 4.865551067512984,
      "grad_norm": 1.0,
      "learning_rate": 0.0002887407265871742,
      "loss": 5.5591,
      "mean_token_accuracy": 0.21253576129674911,
      "num_tokens": 106872532.0,
      "step": 42160
    },
    {
      "entropy": 6.083760690689087,
      "epoch": 4.866128101557992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028869953902472536,
      "loss": 5.5787,
      "mean_token_accuracy": 0.20476481318473816,
      "num_tokens": 106884396.0,
      "step": 42165
    },
    {
      "entropy": 6.082341480255127,
      "epoch": 4.866705135603,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028865835100149926,
      "loss": 5.5117,
      "mean_token_accuracy": 0.21547332853078843,
      "num_tokens": 106897069.0,
      "step": 42170
    },
    {
      "entropy": 6.089553546905518,
      "epoch": 4.867282169648009,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002886171625188813,
      "loss": 5.5414,
      "mean_token_accuracy": 0.21292500644922258,
      "num_tokens": 106910347.0,
      "step": 42175
    },
    {
      "entropy": 6.010037231445312,
      "epoch": 4.867859203693018,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002885759735782568,
      "loss": 5.4427,
      "mean_token_accuracy": 0.2145748034119606,
      "num_tokens": 106923552.0,
      "step": 42180
    },
    {
      "entropy": 5.980557918548584,
      "epoch": 4.868436237738027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002885347841810112,
      "loss": 5.3473,
      "mean_token_accuracy": 0.22816805094480513,
      "num_tokens": 106935452.0,
      "step": 42185
    },
    {
      "entropy": 6.046111965179444,
      "epoch": 4.869013271783035,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002884935943285299,
      "loss": 5.4902,
      "mean_token_accuracy": 0.21253160685300826,
      "num_tokens": 106947460.0,
      "step": 42190
    },
    {
      "entropy": 6.020666694641113,
      "epoch": 4.869590305828044,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002884524040221983,
      "loss": 5.4503,
      "mean_token_accuracy": 0.22385992258787155,
      "num_tokens": 106959219.0,
      "step": 42195
    },
    {
      "entropy": 6.055613994598389,
      "epoch": 4.870167339873053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002884112132634017,
      "loss": 5.5362,
      "mean_token_accuracy": 0.21146431863307952,
      "num_tokens": 106971208.0,
      "step": 42200
    },
    {
      "entropy": 6.0770158767700195,
      "epoch": 4.870744373918061,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002883700220535256,
      "loss": 5.5127,
      "mean_token_accuracy": 0.2192384496331215,
      "num_tokens": 106984729.0,
      "step": 42205
    },
    {
      "entropy": 6.076972627639771,
      "epoch": 4.87132140796307,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028832883039395544,
      "loss": 5.5113,
      "mean_token_accuracy": 0.21844467669725418,
      "num_tokens": 106997551.0,
      "step": 42210
    },
    {
      "entropy": 6.08613429069519,
      "epoch": 4.871898442008078,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002882876382860768,
      "loss": 5.539,
      "mean_token_accuracy": 0.2118827849626541,
      "num_tokens": 107010368.0,
      "step": 42215
    },
    {
      "entropy": 6.1925537109375,
      "epoch": 4.872475476053087,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002882464457312749,
      "loss": 5.5783,
      "mean_token_accuracy": 0.20799676179885865,
      "num_tokens": 107023217.0,
      "step": 42220
    },
    {
      "entropy": 6.08561520576477,
      "epoch": 4.873052510098096,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002882052527309354,
      "loss": 5.548,
      "mean_token_accuracy": 0.21356784254312516,
      "num_tokens": 107037056.0,
      "step": 42225
    },
    {
      "entropy": 5.9867613315582275,
      "epoch": 4.873629544143104,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002881640592864439,
      "loss": 5.4267,
      "mean_token_accuracy": 0.2184175208210945,
      "num_tokens": 107050454.0,
      "step": 42230
    },
    {
      "entropy": 6.033309507369995,
      "epoch": 4.874206578188113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002881228653991857,
      "loss": 5.4266,
      "mean_token_accuracy": 0.22961031049489974,
      "num_tokens": 107062365.0,
      "step": 42235
    },
    {
      "entropy": 5.905626058578491,
      "epoch": 4.874783612233122,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0002880816710705464,
      "loss": 5.3426,
      "mean_token_accuracy": 0.2287436991930008,
      "num_tokens": 107075173.0,
      "step": 42240
    },
    {
      "entropy": 6.067697954177857,
      "epoch": 4.87536064627813,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002880404763019117,
      "loss": 5.4985,
      "mean_token_accuracy": 0.21249936521053314,
      "num_tokens": 107086272.0,
      "step": 42245
    },
    {
      "entropy": 6.111725378036499,
      "epoch": 4.875937680323139,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028799928109466707,
      "loss": 5.5503,
      "mean_token_accuracy": 0.21670805662870407,
      "num_tokens": 107098790.0,
      "step": 42250
    },
    {
      "entropy": 6.00868501663208,
      "epoch": 4.876514714368148,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028795808545019797,
      "loss": 5.449,
      "mean_token_accuracy": 0.21645189821720123,
      "num_tokens": 107110296.0,
      "step": 42255
    },
    {
      "entropy": 6.091307353973389,
      "epoch": 4.877091748413156,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002879168893698901,
      "loss": 5.5438,
      "mean_token_accuracy": 0.21070765554904938,
      "num_tokens": 107123888.0,
      "step": 42260
    },
    {
      "entropy": 6.08305721282959,
      "epoch": 4.877668782458165,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002878756928551292,
      "loss": 5.5659,
      "mean_token_accuracy": 0.21068717688322067,
      "num_tokens": 107137204.0,
      "step": 42265
    },
    {
      "entropy": 6.0707193374633786,
      "epoch": 4.878245816503174,
      "grad_norm": 0.890625,
      "learning_rate": 0.00028783449590730073,
      "loss": 5.4271,
      "mean_token_accuracy": 0.2317424550652504,
      "num_tokens": 107150466.0,
      "step": 42270
    },
    {
      "entropy": 6.033425903320312,
      "epoch": 4.8788228505481825,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002877932985277903,
      "loss": 5.4922,
      "mean_token_accuracy": 0.2129114642739296,
      "num_tokens": 107162490.0,
      "step": 42275
    },
    {
      "entropy": 6.040828704833984,
      "epoch": 4.879399884593191,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028775210071798365,
      "loss": 5.5157,
      "mean_token_accuracy": 0.2181906819343567,
      "num_tokens": 107176562.0,
      "step": 42280
    },
    {
      "entropy": 6.044349002838135,
      "epoch": 4.879976918638199,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028771090247926644,
      "loss": 5.4775,
      "mean_token_accuracy": 0.2157879039645195,
      "num_tokens": 107189075.0,
      "step": 42285
    },
    {
      "entropy": 6.037503862380982,
      "epoch": 4.880553952683209,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002876697038130243,
      "loss": 5.459,
      "mean_token_accuracy": 0.22896388471126555,
      "num_tokens": 107201462.0,
      "step": 42290
    },
    {
      "entropy": 6.076305913925171,
      "epoch": 4.881130986728217,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002876285047206431,
      "loss": 5.5356,
      "mean_token_accuracy": 0.2162700816988945,
      "num_tokens": 107214250.0,
      "step": 42295
    },
    {
      "entropy": 6.01119589805603,
      "epoch": 4.8817080207732255,
      "grad_norm": 0.9375,
      "learning_rate": 0.00028758730520350826,
      "loss": 5.4665,
      "mean_token_accuracy": 0.21472443789243698,
      "num_tokens": 107226680.0,
      "step": 42300
    },
    {
      "entropy": 6.094110774993896,
      "epoch": 4.882285054818234,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028754610526300574,
      "loss": 5.5523,
      "mean_token_accuracy": 0.20786332637071608,
      "num_tokens": 107239245.0,
      "step": 42305
    },
    {
      "entropy": 6.06618185043335,
      "epoch": 4.882862088863243,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028750490490052116,
      "loss": 5.4779,
      "mean_token_accuracy": 0.21719125360250474,
      "num_tokens": 107252004.0,
      "step": 42310
    },
    {
      "entropy": 5.980462598800659,
      "epoch": 4.883439122908252,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028746370411744036,
      "loss": 5.498,
      "mean_token_accuracy": 0.21749580800533294,
      "num_tokens": 107264008.0,
      "step": 42315
    },
    {
      "entropy": 6.078024911880493,
      "epoch": 4.88401615695326,
      "grad_norm": 1.09375,
      "learning_rate": 0.000287422502915149,
      "loss": 5.5544,
      "mean_token_accuracy": 0.21611079424619675,
      "num_tokens": 107276119.0,
      "step": 42320
    },
    {
      "entropy": 6.115493535995483,
      "epoch": 4.8845931909982685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000287381301295033,
      "loss": 5.5364,
      "mean_token_accuracy": 0.21301726698875428,
      "num_tokens": 107288837.0,
      "step": 42325
    },
    {
      "entropy": 6.054299020767212,
      "epoch": 4.885170225043278,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028734009925847803,
      "loss": 5.4678,
      "mean_token_accuracy": 0.21994746029376983,
      "num_tokens": 107301831.0,
      "step": 42330
    },
    {
      "entropy": 6.094528770446777,
      "epoch": 4.885747259088286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028729889680687,
      "loss": 5.5109,
      "mean_token_accuracy": 0.21619769483804702,
      "num_tokens": 107313680.0,
      "step": 42335
    },
    {
      "entropy": 6.124050188064575,
      "epoch": 4.886324293133295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002872576939415947,
      "loss": 5.6108,
      "mean_token_accuracy": 0.20930839627981185,
      "num_tokens": 107326160.0,
      "step": 42340
    },
    {
      "entropy": 6.069322681427002,
      "epoch": 4.886901327178304,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002872164906640378,
      "loss": 5.5046,
      "mean_token_accuracy": 0.2140672102570534,
      "num_tokens": 107339479.0,
      "step": 42345
    },
    {
      "entropy": 6.110053443908692,
      "epoch": 4.887478361223312,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028717528697558536,
      "loss": 5.6014,
      "mean_token_accuracy": 0.20572398751974105,
      "num_tokens": 107352327.0,
      "step": 42350
    },
    {
      "entropy": 6.1258350849151615,
      "epoch": 4.888055395268321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028713408287762325,
      "loss": 5.5285,
      "mean_token_accuracy": 0.21264512538909913,
      "num_tokens": 107365232.0,
      "step": 42355
    },
    {
      "entropy": 5.916987323760987,
      "epoch": 4.888632429313329,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028709287837153737,
      "loss": 5.3993,
      "mean_token_accuracy": 0.22558772414922715,
      "num_tokens": 107378632.0,
      "step": 42360
    },
    {
      "entropy": 6.063654899597168,
      "epoch": 4.8892094633583385,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002870516734587134,
      "loss": 5.5048,
      "mean_token_accuracy": 0.2170914277434349,
      "num_tokens": 107390729.0,
      "step": 42365
    },
    {
      "entropy": 6.103580665588379,
      "epoch": 4.889786497403347,
      "grad_norm": 0.90625,
      "learning_rate": 0.0002870104681405374,
      "loss": 5.5364,
      "mean_token_accuracy": 0.21508759111166,
      "num_tokens": 107402497.0,
      "step": 42370
    },
    {
      "entropy": 6.030554294586182,
      "epoch": 4.890363531448355,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002869692624183953,
      "loss": 5.4201,
      "mean_token_accuracy": 0.22807493060827255,
      "num_tokens": 107416276.0,
      "step": 42375
    },
    {
      "entropy": 6.030253982543945,
      "epoch": 4.890940565493364,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00028692805629367294,
      "loss": 5.4502,
      "mean_token_accuracy": 0.21923165321350097,
      "num_tokens": 107429806.0,
      "step": 42380
    },
    {
      "entropy": 6.042533731460571,
      "epoch": 4.891517599538373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028688684976775646,
      "loss": 5.4597,
      "mean_token_accuracy": 0.21634673923254014,
      "num_tokens": 107441707.0,
      "step": 42385
    },
    {
      "entropy": 6.091143178939819,
      "epoch": 4.8920946335833815,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002868456428420316,
      "loss": 5.5421,
      "mean_token_accuracy": 0.21324512213468552,
      "num_tokens": 107455572.0,
      "step": 42390
    },
    {
      "entropy": 6.001121187210083,
      "epoch": 4.89267166762839,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002868044355178845,
      "loss": 5.4195,
      "mean_token_accuracy": 0.215337535738945,
      "num_tokens": 107468324.0,
      "step": 42395
    },
    {
      "entropy": 6.098383808135987,
      "epoch": 4.893248701673398,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000286763227796701,
      "loss": 5.5066,
      "mean_token_accuracy": 0.21163651645183562,
      "num_tokens": 107479984.0,
      "step": 42400
    },
    {
      "entropy": 6.159455966949463,
      "epoch": 4.893825735718408,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002867220196798672,
      "loss": 5.5945,
      "mean_token_accuracy": 0.2068368449807167,
      "num_tokens": 107492374.0,
      "step": 42405
    },
    {
      "entropy": 6.081700134277344,
      "epoch": 4.894402769763416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000286680811168769,
      "loss": 5.579,
      "mean_token_accuracy": 0.20862532258033753,
      "num_tokens": 107504176.0,
      "step": 42410
    },
    {
      "entropy": 6.02947735786438,
      "epoch": 4.8949798038084245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028663960226479266,
      "loss": 5.4765,
      "mean_token_accuracy": 0.2185306578874588,
      "num_tokens": 107515692.0,
      "step": 42415
    },
    {
      "entropy": 6.000588607788086,
      "epoch": 4.895556837853434,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028659839296932407,
      "loss": 5.4555,
      "mean_token_accuracy": 0.21624449789524078,
      "num_tokens": 107527802.0,
      "step": 42420
    },
    {
      "entropy": 6.088767051696777,
      "epoch": 4.896133871898442,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028655718328374924,
      "loss": 5.5646,
      "mean_token_accuracy": 0.2129533752799034,
      "num_tokens": 107539496.0,
      "step": 42425
    },
    {
      "entropy": 6.032118463516236,
      "epoch": 4.896710905943451,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028651597320945426,
      "loss": 5.4675,
      "mean_token_accuracy": 0.21630681902170182,
      "num_tokens": 107551265.0,
      "step": 42430
    },
    {
      "entropy": 6.050626230239868,
      "epoch": 4.897287939988459,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002864747627478253,
      "loss": 5.4763,
      "mean_token_accuracy": 0.21886650770902633,
      "num_tokens": 107562448.0,
      "step": 42435
    },
    {
      "entropy": 6.113722896575927,
      "epoch": 4.897864974033468,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028643355190024843,
      "loss": 5.5995,
      "mean_token_accuracy": 0.20752845257520675,
      "num_tokens": 107575386.0,
      "step": 42440
    },
    {
      "entropy": 6.135255479812622,
      "epoch": 4.898442008078477,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028639234066810967,
      "loss": 5.5253,
      "mean_token_accuracy": 0.21387504786252975,
      "num_tokens": 107588569.0,
      "step": 42445
    },
    {
      "entropy": 6.001686763763428,
      "epoch": 4.899019042123485,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028635112905279515,
      "loss": 5.4387,
      "mean_token_accuracy": 0.21784890443086624,
      "num_tokens": 107602053.0,
      "step": 42450
    },
    {
      "entropy": 5.9554901123046875,
      "epoch": 4.899596076168494,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002863099170556911,
      "loss": 5.4166,
      "mean_token_accuracy": 0.21788729429244996,
      "num_tokens": 107614445.0,
      "step": 42455
    },
    {
      "entropy": 6.041926288604737,
      "epoch": 4.900173110213503,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002862687046781836,
      "loss": 5.483,
      "mean_token_accuracy": 0.21106954514980317,
      "num_tokens": 107628981.0,
      "step": 42460
    },
    {
      "entropy": 6.1609711170196535,
      "epoch": 4.900750144258511,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002862274919216588,
      "loss": 5.5463,
      "mean_token_accuracy": 0.20651800483465194,
      "num_tokens": 107641952.0,
      "step": 42465
    },
    {
      "entropy": 6.140533542633056,
      "epoch": 4.90132717830352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028618627878750286,
      "loss": 5.566,
      "mean_token_accuracy": 0.20501160025596618,
      "num_tokens": 107652949.0,
      "step": 42470
    },
    {
      "entropy": 6.0371479988098145,
      "epoch": 4.901904212348528,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00028614506527710205,
      "loss": 5.4885,
      "mean_token_accuracy": 0.21945639252662658,
      "num_tokens": 107665878.0,
      "step": 42475
    },
    {
      "entropy": 6.07598910331726,
      "epoch": 4.902481246393537,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028610385139184245,
      "loss": 5.5248,
      "mean_token_accuracy": 0.20897335112094878,
      "num_tokens": 107678666.0,
      "step": 42480
    },
    {
      "entropy": 6.0083451747894285,
      "epoch": 4.903058280438546,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028606263713311044,
      "loss": 5.4359,
      "mean_token_accuracy": 0.22138419449329377,
      "num_tokens": 107691259.0,
      "step": 42485
    },
    {
      "entropy": 6.014791488647461,
      "epoch": 4.903635314483554,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002860214225022919,
      "loss": 5.512,
      "mean_token_accuracy": 0.21296166628599167,
      "num_tokens": 107703848.0,
      "step": 42490
    },
    {
      "entropy": 6.062704277038574,
      "epoch": 4.904212348528564,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002859802075007735,
      "loss": 5.5111,
      "mean_token_accuracy": 0.21279639154672622,
      "num_tokens": 107716770.0,
      "step": 42495
    },
    {
      "entropy": 6.024626636505127,
      "epoch": 4.904789382573572,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002859389921299412,
      "loss": 5.4977,
      "mean_token_accuracy": 0.21681968569755555,
      "num_tokens": 107730046.0,
      "step": 42500
    },
    {
      "entropy": 6.009398889541626,
      "epoch": 4.90536641661858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028589777639118134,
      "loss": 5.3802,
      "mean_token_accuracy": 0.22787190675735475,
      "num_tokens": 107742806.0,
      "step": 42505
    },
    {
      "entropy": 6.074196100234985,
      "epoch": 4.905943450663589,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028585656028588016,
      "loss": 5.5676,
      "mean_token_accuracy": 0.21569931656122207,
      "num_tokens": 107758001.0,
      "step": 42510
    },
    {
      "entropy": 6.077644395828247,
      "epoch": 4.906520484708598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000285815343815424,
      "loss": 5.4791,
      "mean_token_accuracy": 0.21419319212436677,
      "num_tokens": 107771008.0,
      "step": 42515
    },
    {
      "entropy": 6.026212453842163,
      "epoch": 4.907097518753607,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002857741269811991,
      "loss": 5.5135,
      "mean_token_accuracy": 0.2148035004734993,
      "num_tokens": 107784276.0,
      "step": 42520
    },
    {
      "entropy": 6.076160526275634,
      "epoch": 4.907674552798615,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002857329097845918,
      "loss": 5.5011,
      "mean_token_accuracy": 0.20898256748914718,
      "num_tokens": 107796555.0,
      "step": 42525
    },
    {
      "entropy": 6.092764568328858,
      "epoch": 4.908251586843623,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002856916922269884,
      "loss": 5.5552,
      "mean_token_accuracy": 0.2083060175180435,
      "num_tokens": 107811899.0,
      "step": 42530
    },
    {
      "entropy": 6.0735053539276125,
      "epoch": 4.908828620888633,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028565047430977524,
      "loss": 5.5252,
      "mean_token_accuracy": 0.21149604320526122,
      "num_tokens": 107824514.0,
      "step": 42535
    },
    {
      "entropy": 6.0582245826721195,
      "epoch": 4.909405654933641,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002856092560343387,
      "loss": 5.4537,
      "mean_token_accuracy": 0.2141177847981453,
      "num_tokens": 107837523.0,
      "step": 42540
    },
    {
      "entropy": 6.008212518692017,
      "epoch": 4.9099826889786495,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0002855680374020651,
      "loss": 5.4323,
      "mean_token_accuracy": 0.2156248614192009,
      "num_tokens": 107849033.0,
      "step": 42545
    },
    {
      "entropy": 6.048239517211914,
      "epoch": 4.910559723023658,
      "grad_norm": 1.0,
      "learning_rate": 0.00028552681841434083,
      "loss": 5.5069,
      "mean_token_accuracy": 0.21879953145980835,
      "num_tokens": 107860856.0,
      "step": 42550
    },
    {
      "entropy": 6.029404544830323,
      "epoch": 4.911136757068667,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002854855990725522,
      "loss": 5.4954,
      "mean_token_accuracy": 0.21359726041555405,
      "num_tokens": 107872802.0,
      "step": 42555
    },
    {
      "entropy": 6.089948749542236,
      "epoch": 4.911713791113676,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028544437937808563,
      "loss": 5.5862,
      "mean_token_accuracy": 0.20500381290912628,
      "num_tokens": 107885882.0,
      "step": 42560
    },
    {
      "entropy": 6.095817565917969,
      "epoch": 4.912290825158684,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002854031593323276,
      "loss": 5.4565,
      "mean_token_accuracy": 0.21755782663822174,
      "num_tokens": 107897363.0,
      "step": 42565
    },
    {
      "entropy": 6.0812242984771725,
      "epoch": 4.912867859203693,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00028536193893666446,
      "loss": 5.4826,
      "mean_token_accuracy": 0.2122122198343277,
      "num_tokens": 107910329.0,
      "step": 42570
    },
    {
      "entropy": 6.067883825302124,
      "epoch": 4.913444893248702,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002853207181924827,
      "loss": 5.5474,
      "mean_token_accuracy": 0.20889407247304917,
      "num_tokens": 107922316.0,
      "step": 42575
    },
    {
      "entropy": 6.051030445098877,
      "epoch": 4.91402192729371,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028527949710116865,
      "loss": 5.4007,
      "mean_token_accuracy": 0.22093999087810517,
      "num_tokens": 107934005.0,
      "step": 42580
    },
    {
      "entropy": 6.043297147750854,
      "epoch": 4.914598961338719,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028523827566410883,
      "loss": 5.4023,
      "mean_token_accuracy": 0.2306014209985733,
      "num_tokens": 107946981.0,
      "step": 42585
    },
    {
      "entropy": 5.920707130432129,
      "epoch": 4.915175995383728,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002851970538826897,
      "loss": 5.3459,
      "mean_token_accuracy": 0.2286318376660347,
      "num_tokens": 107959018.0,
      "step": 42590
    },
    {
      "entropy": 5.990842580795288,
      "epoch": 4.915753029428736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002851558317582977,
      "loss": 5.4864,
      "mean_token_accuracy": 0.21997285038232803,
      "num_tokens": 107971495.0,
      "step": 42595
    },
    {
      "entropy": 6.134239387512207,
      "epoch": 4.916330063473745,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00028511460929231934,
      "loss": 5.5897,
      "mean_token_accuracy": 0.20877509117126464,
      "num_tokens": 107984159.0,
      "step": 42600
    },
    {
      "entropy": 6.062919950485229,
      "epoch": 4.916907097518754,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028507338648614116,
      "loss": 5.4625,
      "mean_token_accuracy": 0.21399381309747695,
      "num_tokens": 107996661.0,
      "step": 42605
    },
    {
      "entropy": 6.043512678146362,
      "epoch": 4.9174841315637625,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028503216334114964,
      "loss": 5.4954,
      "mean_token_accuracy": 0.21791517734527588,
      "num_tokens": 108010054.0,
      "step": 42610
    },
    {
      "entropy": 6.085896015167236,
      "epoch": 4.918061165608771,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028499093985873125,
      "loss": 5.5665,
      "mean_token_accuracy": 0.20890969783067703,
      "num_tokens": 108022667.0,
      "step": 42615
    },
    {
      "entropy": 5.995922994613648,
      "epoch": 4.918638199653779,
      "grad_norm": 1.0,
      "learning_rate": 0.0002849497160402726,
      "loss": 5.4307,
      "mean_token_accuracy": 0.22446546256542205,
      "num_tokens": 108034498.0,
      "step": 42620
    },
    {
      "entropy": 6.070241260528564,
      "epoch": 4.919215233698788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002849084918871602,
      "loss": 5.5213,
      "mean_token_accuracy": 0.21241340041160583,
      "num_tokens": 108047678.0,
      "step": 42625
    },
    {
      "entropy": 6.091761493682862,
      "epoch": 4.919792267743797,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002848672674007805,
      "loss": 5.552,
      "mean_token_accuracy": 0.21082081496715546,
      "num_tokens": 108061072.0,
      "step": 42630
    },
    {
      "entropy": 6.126362943649292,
      "epoch": 4.9203693017888055,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028482604258252026,
      "loss": 5.5749,
      "mean_token_accuracy": 0.21069298386573793,
      "num_tokens": 108073710.0,
      "step": 42635
    },
    {
      "entropy": 6.041939353942871,
      "epoch": 4.920946335833814,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002847848174337659,
      "loss": 5.454,
      "mean_token_accuracy": 0.22192743569612502,
      "num_tokens": 108086377.0,
      "step": 42640
    },
    {
      "entropy": 6.077037334442139,
      "epoch": 4.921523369878823,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028474359195590417,
      "loss": 5.4961,
      "mean_token_accuracy": 0.21405644416809083,
      "num_tokens": 108098181.0,
      "step": 42645
    },
    {
      "entropy": 6.02266674041748,
      "epoch": 4.922100403923832,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0002847023661503214,
      "loss": 5.4658,
      "mean_token_accuracy": 0.2223532721400261,
      "num_tokens": 108111365.0,
      "step": 42650
    },
    {
      "entropy": 6.054823303222657,
      "epoch": 4.92267743796884,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002846611400184045,
      "loss": 5.4981,
      "mean_token_accuracy": 0.21125366240739823,
      "num_tokens": 108123718.0,
      "step": 42655
    },
    {
      "entropy": 5.946111488342285,
      "epoch": 4.9232544720138485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002846199135615399,
      "loss": 5.3657,
      "mean_token_accuracy": 0.2319044515490532,
      "num_tokens": 108136533.0,
      "step": 42660
    },
    {
      "entropy": 6.037826299667358,
      "epoch": 4.923831506058858,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002845786867811143,
      "loss": 5.4683,
      "mean_token_accuracy": 0.21365686357021332,
      "num_tokens": 108149340.0,
      "step": 42665
    },
    {
      "entropy": 5.891231966018677,
      "epoch": 4.924408540103866,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00028453745967851435,
      "loss": 5.3095,
      "mean_token_accuracy": 0.23251685202121736,
      "num_tokens": 108163229.0,
      "step": 42670
    },
    {
      "entropy": 6.097901916503906,
      "epoch": 4.924985574148875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002844962322551266,
      "loss": 5.549,
      "mean_token_accuracy": 0.2108125388622284,
      "num_tokens": 108175847.0,
      "step": 42675
    },
    {
      "entropy": 6.03236231803894,
      "epoch": 4.925562608193884,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028445500451233785,
      "loss": 5.5276,
      "mean_token_accuracy": 0.217990905046463,
      "num_tokens": 108189888.0,
      "step": 42680
    },
    {
      "entropy": 5.996472597122192,
      "epoch": 4.926139642238892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028441377645153476,
      "loss": 5.4023,
      "mean_token_accuracy": 0.22852868884801864,
      "num_tokens": 108201555.0,
      "step": 42685
    },
    {
      "entropy": 5.986435270309448,
      "epoch": 4.926716676283901,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002843725480741039,
      "loss": 5.3497,
      "mean_token_accuracy": 0.2356192350387573,
      "num_tokens": 108214386.0,
      "step": 42690
    },
    {
      "entropy": 6.0332691192626955,
      "epoch": 4.927293710328909,
      "grad_norm": 1.03125,
      "learning_rate": 0.000284331319381432,
      "loss": 5.4503,
      "mean_token_accuracy": 0.2227146580815315,
      "num_tokens": 108227359.0,
      "step": 42695
    },
    {
      "entropy": 6.008638000488281,
      "epoch": 4.927870744373918,
      "grad_norm": 1.0,
      "learning_rate": 0.000284290090374906,
      "loss": 5.4864,
      "mean_token_accuracy": 0.2140437811613083,
      "num_tokens": 108239407.0,
      "step": 42700
    },
    {
      "entropy": 6.025502490997314,
      "epoch": 4.928447778418927,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028424886105591225,
      "loss": 5.4031,
      "mean_token_accuracy": 0.22531924545764923,
      "num_tokens": 108252372.0,
      "step": 42705
    },
    {
      "entropy": 6.013317394256592,
      "epoch": 4.929024812463935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002842076314258377,
      "loss": 5.4331,
      "mean_token_accuracy": 0.22245500832796097,
      "num_tokens": 108263791.0,
      "step": 42710
    },
    {
      "entropy": 6.021922779083252,
      "epoch": 4.929601846508944,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028416640148606903,
      "loss": 5.4973,
      "mean_token_accuracy": 0.21262186020612717,
      "num_tokens": 108276897.0,
      "step": 42715
    },
    {
      "entropy": 5.952204942703247,
      "epoch": 4.930178880553953,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028412517123799306,
      "loss": 5.4224,
      "mean_token_accuracy": 0.21960787028074263,
      "num_tokens": 108290577.0,
      "step": 42720
    },
    {
      "entropy": 5.980778169631958,
      "epoch": 4.9307559145989615,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002840839406829965,
      "loss": 5.3573,
      "mean_token_accuracy": 0.22540011405944824,
      "num_tokens": 108303426.0,
      "step": 42725
    },
    {
      "entropy": 6.144718885421753,
      "epoch": 4.93133294864397,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000284042709822466,
      "loss": 5.5896,
      "mean_token_accuracy": 0.19863436967134476,
      "num_tokens": 108315736.0,
      "step": 42730
    },
    {
      "entropy": 6.060486268997193,
      "epoch": 4.931909982688978,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002840014786577885,
      "loss": 5.4504,
      "mean_token_accuracy": 0.2206668183207512,
      "num_tokens": 108329150.0,
      "step": 42735
    },
    {
      "entropy": 6.104358339309693,
      "epoch": 4.932487016733988,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028396024719035075,
      "loss": 5.5301,
      "mean_token_accuracy": 0.20877788215875626,
      "num_tokens": 108341334.0,
      "step": 42740
    },
    {
      "entropy": 6.003159046173096,
      "epoch": 4.933064050778996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002839190154215395,
      "loss": 5.4754,
      "mean_token_accuracy": 0.21796131879091263,
      "num_tokens": 108353039.0,
      "step": 42745
    },
    {
      "entropy": 6.089808797836303,
      "epoch": 4.933641084824004,
      "grad_norm": 0.890625,
      "learning_rate": 0.0002838777833527416,
      "loss": 5.5257,
      "mean_token_accuracy": 0.21428242921829224,
      "num_tokens": 108366293.0,
      "step": 42750
    },
    {
      "entropy": 6.1134514808654785,
      "epoch": 4.934218118869014,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002838365509853439,
      "loss": 5.5635,
      "mean_token_accuracy": 0.21114464700222016,
      "num_tokens": 108378857.0,
      "step": 42755
    },
    {
      "entropy": 5.956152105331421,
      "epoch": 4.934795152914022,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002837953183207332,
      "loss": 5.4261,
      "mean_token_accuracy": 0.222879259288311,
      "num_tokens": 108391113.0,
      "step": 42760
    },
    {
      "entropy": 6.114244365692139,
      "epoch": 4.935372186959031,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002837540853602963,
      "loss": 5.4697,
      "mean_token_accuracy": 0.2117037668824196,
      "num_tokens": 108404304.0,
      "step": 42765
    },
    {
      "entropy": 5.991312551498413,
      "epoch": 4.935949221004039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028371285210542006,
      "loss": 5.3663,
      "mean_token_accuracy": 0.22580697536468505,
      "num_tokens": 108415973.0,
      "step": 42770
    },
    {
      "entropy": 5.948039293289185,
      "epoch": 4.936526255049047,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002836716185574915,
      "loss": 5.4508,
      "mean_token_accuracy": 0.21183349639177323,
      "num_tokens": 108428668.0,
      "step": 42775
    },
    {
      "entropy": 6.064259481430054,
      "epoch": 4.937103289094057,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028363038471789724,
      "loss": 5.5303,
      "mean_token_accuracy": 0.2114143803715706,
      "num_tokens": 108441722.0,
      "step": 42780
    },
    {
      "entropy": 6.030072116851807,
      "epoch": 4.937680323139065,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002835891505880243,
      "loss": 5.459,
      "mean_token_accuracy": 0.22549148201942443,
      "num_tokens": 108454157.0,
      "step": 42785
    },
    {
      "entropy": 6.032482719421386,
      "epoch": 4.9382573571840735,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00028354791616925964,
      "loss": 5.436,
      "mean_token_accuracy": 0.22359249591827393,
      "num_tokens": 108467028.0,
      "step": 42790
    },
    {
      "entropy": 5.9496900081634525,
      "epoch": 4.938834391229083,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002835066814629899,
      "loss": 5.404,
      "mean_token_accuracy": 0.22584114223718643,
      "num_tokens": 108479460.0,
      "step": 42795
    },
    {
      "entropy": 5.9788023948669435,
      "epoch": 4.939411425274091,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002834654464706023,
      "loss": 5.437,
      "mean_token_accuracy": 0.22376679480075837,
      "num_tokens": 108491698.0,
      "step": 42800
    },
    {
      "entropy": 6.008862018585205,
      "epoch": 4.9399884593191,
      "grad_norm": 1.0,
      "learning_rate": 0.0002834242111934835,
      "loss": 5.388,
      "mean_token_accuracy": 0.22189870476722717,
      "num_tokens": 108503191.0,
      "step": 42805
    },
    {
      "entropy": 6.1356651306152346,
      "epoch": 4.940565493364108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028338297563302054,
      "loss": 5.5563,
      "mean_token_accuracy": 0.20657484531402587,
      "num_tokens": 108516269.0,
      "step": 42810
    },
    {
      "entropy": 6.085603809356689,
      "epoch": 4.941142527409117,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028334173979060047,
      "loss": 5.5043,
      "mean_token_accuracy": 0.222974793612957,
      "num_tokens": 108529880.0,
      "step": 42815
    },
    {
      "entropy": 6.02394061088562,
      "epoch": 4.941719561454126,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002833005036676101,
      "loss": 5.4745,
      "mean_token_accuracy": 0.2152271181344986,
      "num_tokens": 108542063.0,
      "step": 42820
    },
    {
      "entropy": 6.01603946685791,
      "epoch": 4.942296595499134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028325926726543635,
      "loss": 5.4658,
      "mean_token_accuracy": 0.223656365275383,
      "num_tokens": 108554654.0,
      "step": 42825
    },
    {
      "entropy": 5.984672260284424,
      "epoch": 4.9428736295441436,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002832180305854663,
      "loss": 5.4662,
      "mean_token_accuracy": 0.2179507076740265,
      "num_tokens": 108567375.0,
      "step": 42830
    },
    {
      "entropy": 6.03062219619751,
      "epoch": 4.943450663589152,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002831767936290869,
      "loss": 5.4641,
      "mean_token_accuracy": 0.2168751984834671,
      "num_tokens": 108580721.0,
      "step": 42835
    },
    {
      "entropy": 6.084998703002929,
      "epoch": 4.94402769763416,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028313555639768504,
      "loss": 5.5261,
      "mean_token_accuracy": 0.21340830624103546,
      "num_tokens": 108594106.0,
      "step": 42840
    },
    {
      "entropy": 5.9906532764434814,
      "epoch": 4.944604731679169,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002830943188926477,
      "loss": 5.4681,
      "mean_token_accuracy": 0.21862661093473434,
      "num_tokens": 108605373.0,
      "step": 42845
    },
    {
      "entropy": 5.989369010925293,
      "epoch": 4.945181765724178,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028305308111536216,
      "loss": 5.441,
      "mean_token_accuracy": 0.2241882011294365,
      "num_tokens": 108618208.0,
      "step": 42850
    },
    {
      "entropy": 6.142806434631348,
      "epoch": 4.9457587997691865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002830118430672151,
      "loss": 5.6231,
      "mean_token_accuracy": 0.21054959148168564,
      "num_tokens": 108631260.0,
      "step": 42855
    },
    {
      "entropy": 6.096552181243896,
      "epoch": 4.946335833814195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028297060474959373,
      "loss": 5.5498,
      "mean_token_accuracy": 0.20989176630973816,
      "num_tokens": 108644577.0,
      "step": 42860
    },
    {
      "entropy": 6.04664978981018,
      "epoch": 4.946912867859203,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028292936616388503,
      "loss": 5.5143,
      "mean_token_accuracy": 0.2106948733329773,
      "num_tokens": 108656239.0,
      "step": 42865
    },
    {
      "entropy": 6.0369397640228275,
      "epoch": 4.947489901904213,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000282888127311476,
      "loss": 5.4179,
      "mean_token_accuracy": 0.22064224183559417,
      "num_tokens": 108668562.0,
      "step": 42870
    },
    {
      "entropy": 5.892840051651001,
      "epoch": 4.948066935949221,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028284688819375374,
      "loss": 5.2901,
      "mean_token_accuracy": 0.23528130054473878,
      "num_tokens": 108680186.0,
      "step": 42875
    },
    {
      "entropy": 5.976221227645874,
      "epoch": 4.9486439699942295,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028280564881210535,
      "loss": 5.4734,
      "mean_token_accuracy": 0.2145787388086319,
      "num_tokens": 108692444.0,
      "step": 42880
    },
    {
      "entropy": 6.023977613449096,
      "epoch": 4.949221004039238,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002827644091679178,
      "loss": 5.4631,
      "mean_token_accuracy": 0.2159357711672783,
      "num_tokens": 108704972.0,
      "step": 42885
    },
    {
      "entropy": 5.972660732269287,
      "epoch": 4.949798038084247,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028272316926257815,
      "loss": 5.3349,
      "mean_token_accuracy": 0.23300850242376328,
      "num_tokens": 108718735.0,
      "step": 42890
    },
    {
      "entropy": 6.079352760314942,
      "epoch": 4.950375072129256,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00028268192909747367,
      "loss": 5.5774,
      "mean_token_accuracy": 0.21249482333660125,
      "num_tokens": 108732119.0,
      "step": 42895
    },
    {
      "entropy": 6.102401876449585,
      "epoch": 4.950952106174264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028264068867399117,
      "loss": 5.5164,
      "mean_token_accuracy": 0.21498086750507356,
      "num_tokens": 108743445.0,
      "step": 42900
    },
    {
      "entropy": 6.060680246353149,
      "epoch": 4.951529140219273,
      "grad_norm": 0.921875,
      "learning_rate": 0.00028259944799351796,
      "loss": 5.4436,
      "mean_token_accuracy": 0.22306081652641296,
      "num_tokens": 108756566.0,
      "step": 42905
    },
    {
      "entropy": 6.000123119354248,
      "epoch": 4.952106174264282,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002825582070574411,
      "loss": 5.4621,
      "mean_token_accuracy": 0.21929712891578673,
      "num_tokens": 108768382.0,
      "step": 42910
    },
    {
      "entropy": 5.923867130279541,
      "epoch": 4.95268320830929,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002825169658671477,
      "loss": 5.3787,
      "mean_token_accuracy": 0.229587422311306,
      "num_tokens": 108780905.0,
      "step": 42915
    },
    {
      "entropy": 6.134624481201172,
      "epoch": 4.953260242354299,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002824757244240249,
      "loss": 5.6432,
      "mean_token_accuracy": 0.2025127574801445,
      "num_tokens": 108794029.0,
      "step": 42920
    },
    {
      "entropy": 6.021833848953247,
      "epoch": 4.953837276399308,
      "grad_norm": 3.28125,
      "learning_rate": 0.0002824344827294599,
      "loss": 5.4937,
      "mean_token_accuracy": 0.22164767533540725,
      "num_tokens": 108805936.0,
      "step": 42925
    },
    {
      "entropy": 6.116102600097657,
      "epoch": 4.954414310444316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002823932407848396,
      "loss": 5.5467,
      "mean_token_accuracy": 0.20866008400917052,
      "num_tokens": 108818259.0,
      "step": 42930
    },
    {
      "entropy": 6.096166610717773,
      "epoch": 4.954991344489325,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028235199859155143,
      "loss": 5.5349,
      "mean_token_accuracy": 0.2173505961894989,
      "num_tokens": 108830449.0,
      "step": 42935
    },
    {
      "entropy": 6.04797329902649,
      "epoch": 4.955568378534333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002823107561509824,
      "loss": 5.4565,
      "mean_token_accuracy": 0.22156455367803574,
      "num_tokens": 108843586.0,
      "step": 42940
    },
    {
      "entropy": 6.0117346286773685,
      "epoch": 4.9561454125793425,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028226951346451974,
      "loss": 5.4601,
      "mean_token_accuracy": 0.22760563492774963,
      "num_tokens": 108856190.0,
      "step": 42945
    },
    {
      "entropy": 6.040597820281983,
      "epoch": 4.956722446624351,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002822282705335505,
      "loss": 5.6003,
      "mean_token_accuracy": 0.2147225633263588,
      "num_tokens": 108869677.0,
      "step": 42950
    },
    {
      "entropy": 6.011884307861328,
      "epoch": 4.957299480669359,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028218702735946207,
      "loss": 5.3966,
      "mean_token_accuracy": 0.22292168587446212,
      "num_tokens": 108882953.0,
      "step": 42955
    },
    {
      "entropy": 6.014325761795044,
      "epoch": 4.957876514714368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028214578394364153,
      "loss": 5.4345,
      "mean_token_accuracy": 0.22130564898252486,
      "num_tokens": 108896471.0,
      "step": 42960
    },
    {
      "entropy": 6.01017689704895,
      "epoch": 4.958453548759377,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002821045402874762,
      "loss": 5.3785,
      "mean_token_accuracy": 0.22776084244251252,
      "num_tokens": 108909922.0,
      "step": 42965
    },
    {
      "entropy": 6.095320177078247,
      "epoch": 4.9590305828043855,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028206329639235303,
      "loss": 5.5167,
      "mean_token_accuracy": 0.21930584758520127,
      "num_tokens": 108922972.0,
      "step": 42970
    },
    {
      "entropy": 6.008806943893433,
      "epoch": 4.959607616849394,
      "grad_norm": 1.0,
      "learning_rate": 0.00028202205225965943,
      "loss": 5.4834,
      "mean_token_accuracy": 0.22035084366798402,
      "num_tokens": 108934523.0,
      "step": 42975
    },
    {
      "entropy": 6.064395523071289,
      "epoch": 4.960184650894403,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002819808078907827,
      "loss": 5.4931,
      "mean_token_accuracy": 0.21269882917404176,
      "num_tokens": 108947148.0,
      "step": 42980
    },
    {
      "entropy": 5.988593244552613,
      "epoch": 4.960761684939412,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002819395632871098,
      "loss": 5.4012,
      "mean_token_accuracy": 0.2195812940597534,
      "num_tokens": 108960236.0,
      "step": 42985
    },
    {
      "entropy": 6.007198333740234,
      "epoch": 4.96133871898442,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028189831845002816,
      "loss": 5.4779,
      "mean_token_accuracy": 0.22291641533374787,
      "num_tokens": 108974464.0,
      "step": 42990
    },
    {
      "entropy": 6.01093864440918,
      "epoch": 4.9619157530294284,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002818570733809251,
      "loss": 5.4372,
      "mean_token_accuracy": 0.22423449605703355,
      "num_tokens": 108988986.0,
      "step": 42995
    },
    {
      "entropy": 5.990017652511597,
      "epoch": 4.962492787074438,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028181582808118775,
      "loss": 5.4336,
      "mean_token_accuracy": 0.22402408570051194,
      "num_tokens": 109002899.0,
      "step": 43000
    },
    {
      "entropy": 6.000071907043457,
      "epoch": 4.963069821119446,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002817745825522034,
      "loss": 5.3667,
      "mean_token_accuracy": 0.23070352673530578,
      "num_tokens": 109015312.0,
      "step": 43005
    },
    {
      "entropy": 6.1429602146148685,
      "epoch": 4.963646855164455,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002817333367953593,
      "loss": 5.5415,
      "mean_token_accuracy": 0.21176917403936385,
      "num_tokens": 109027569.0,
      "step": 43010
    },
    {
      "entropy": 6.108979845046997,
      "epoch": 4.964223889209463,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028169209081204283,
      "loss": 5.5457,
      "mean_token_accuracy": 0.20894451737403869,
      "num_tokens": 109039718.0,
      "step": 43015
    },
    {
      "entropy": 6.055894899368286,
      "epoch": 4.964800923254472,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002816508446036412,
      "loss": 5.4795,
      "mean_token_accuracy": 0.21691870242357253,
      "num_tokens": 109051525.0,
      "step": 43020
    },
    {
      "entropy": 6.05994143486023,
      "epoch": 4.965377957299481,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002816095981715417,
      "loss": 5.4981,
      "mean_token_accuracy": 0.22089499235153198,
      "num_tokens": 109063865.0,
      "step": 43025
    },
    {
      "entropy": 6.045294380187988,
      "epoch": 4.965954991344489,
      "grad_norm": 1.0,
      "learning_rate": 0.00028156835151713154,
      "loss": 5.511,
      "mean_token_accuracy": 0.21797371059656143,
      "num_tokens": 109075768.0,
      "step": 43030
    },
    {
      "entropy": 6.0039207458496096,
      "epoch": 4.966532025389498,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028152710464179827,
      "loss": 5.4396,
      "mean_token_accuracy": 0.22051458954811096,
      "num_tokens": 109088417.0,
      "step": 43035
    },
    {
      "entropy": 6.089677095413208,
      "epoch": 4.967109059434507,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002814858575469291,
      "loss": 5.5869,
      "mean_token_accuracy": 0.2123097538948059,
      "num_tokens": 109100170.0,
      "step": 43040
    },
    {
      "entropy": 6.0384680271148685,
      "epoch": 4.967686093479515,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028144461023391125,
      "loss": 5.434,
      "mean_token_accuracy": 0.22179457694292068,
      "num_tokens": 109112680.0,
      "step": 43045
    },
    {
      "entropy": 6.027723884582519,
      "epoch": 4.968263127524524,
      "grad_norm": 1.0,
      "learning_rate": 0.00028140336270413214,
      "loss": 5.4376,
      "mean_token_accuracy": 0.22216033339500427,
      "num_tokens": 109125169.0,
      "step": 43050
    },
    {
      "entropy": 6.083027029037476,
      "epoch": 4.968840161569533,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028136211495897914,
      "loss": 5.5017,
      "mean_token_accuracy": 0.21493292301893235,
      "num_tokens": 109137049.0,
      "step": 43055
    },
    {
      "entropy": 6.128494739532471,
      "epoch": 4.969417195614541,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028132086699983954,
      "loss": 5.5474,
      "mean_token_accuracy": 0.21425151973962783,
      "num_tokens": 109149005.0,
      "step": 43060
    },
    {
      "entropy": 6.004573059082031,
      "epoch": 4.96999422965955,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028127961882810074,
      "loss": 5.4998,
      "mean_token_accuracy": 0.21845198571681976,
      "num_tokens": 109162628.0,
      "step": 43065
    },
    {
      "entropy": 6.028212213516236,
      "epoch": 4.970571263704558,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028123837044515,
      "loss": 5.5116,
      "mean_token_accuracy": 0.21566485464572907,
      "num_tokens": 109175105.0,
      "step": 43070
    },
    {
      "entropy": 6.056178283691406,
      "epoch": 4.971148297749568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002811971218523749,
      "loss": 5.4763,
      "mean_token_accuracy": 0.22011447995901107,
      "num_tokens": 109187387.0,
      "step": 43075
    },
    {
      "entropy": 6.075297784805298,
      "epoch": 4.971725331794576,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002811558730511626,
      "loss": 5.5008,
      "mean_token_accuracy": 0.2134289026260376,
      "num_tokens": 109198705.0,
      "step": 43080
    },
    {
      "entropy": 6.108932161331177,
      "epoch": 4.972302365839584,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00028111462404290064,
      "loss": 5.5391,
      "mean_token_accuracy": 0.21483826339244844,
      "num_tokens": 109212001.0,
      "step": 43085
    },
    {
      "entropy": 6.006114387512207,
      "epoch": 4.972879399884593,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002810733748289762,
      "loss": 5.4496,
      "mean_token_accuracy": 0.22773292660713196,
      "num_tokens": 109223733.0,
      "step": 43090
    },
    {
      "entropy": 6.04915771484375,
      "epoch": 4.973456433929602,
      "grad_norm": 0.953125,
      "learning_rate": 0.000281032125410777,
      "loss": 5.4692,
      "mean_token_accuracy": 0.21967923939228057,
      "num_tokens": 109236055.0,
      "step": 43095
    },
    {
      "entropy": 6.019994020462036,
      "epoch": 4.9740334679746105,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002809908757896901,
      "loss": 5.4873,
      "mean_token_accuracy": 0.2136274680495262,
      "num_tokens": 109247662.0,
      "step": 43100
    },
    {
      "entropy": 6.084213733673096,
      "epoch": 4.974610502019619,
      "grad_norm": 1.0,
      "learning_rate": 0.00028094962596710314,
      "loss": 5.5586,
      "mean_token_accuracy": 0.21334403455257417,
      "num_tokens": 109259932.0,
      "step": 43105
    },
    {
      "entropy": 6.086196756362915,
      "epoch": 4.975187536064627,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002809083759444035,
      "loss": 5.5243,
      "mean_token_accuracy": 0.21859664022922515,
      "num_tokens": 109273122.0,
      "step": 43110
    },
    {
      "entropy": 6.07145586013794,
      "epoch": 4.975764570109637,
      "grad_norm": 1.0,
      "learning_rate": 0.0002808671257229785,
      "loss": 5.5414,
      "mean_token_accuracy": 0.2049272373318672,
      "num_tokens": 109285600.0,
      "step": 43115
    },
    {
      "entropy": 6.0434510707855225,
      "epoch": 4.976341604154645,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002808258753042157,
      "loss": 5.4547,
      "mean_token_accuracy": 0.2106258124113083,
      "num_tokens": 109297820.0,
      "step": 43120
    },
    {
      "entropy": 6.080720090866089,
      "epoch": 4.9769186381996535,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002807846246895024,
      "loss": 5.5158,
      "mean_token_accuracy": 0.21140980273485183,
      "num_tokens": 109309313.0,
      "step": 43125
    },
    {
      "entropy": 6.0035765171051025,
      "epoch": 4.977495672244663,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028074337388022617,
      "loss": 5.4669,
      "mean_token_accuracy": 0.22074911445379258,
      "num_tokens": 109321987.0,
      "step": 43130
    },
    {
      "entropy": 6.025129985809326,
      "epoch": 4.978072706289671,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002807021228777744,
      "loss": 5.4636,
      "mean_token_accuracy": 0.21835483461618424,
      "num_tokens": 109333911.0,
      "step": 43135
    },
    {
      "entropy": 6.079827880859375,
      "epoch": 4.97864974033468,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002806608716835346,
      "loss": 5.5737,
      "mean_token_accuracy": 0.2068240001797676,
      "num_tokens": 109346755.0,
      "step": 43140
    },
    {
      "entropy": 6.012156867980957,
      "epoch": 4.979226774379688,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028061962029889414,
      "loss": 5.4835,
      "mean_token_accuracy": 0.21426208466291427,
      "num_tokens": 109360584.0,
      "step": 43145
    },
    {
      "entropy": 6.022485733032227,
      "epoch": 4.979803808424697,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002805783687252405,
      "loss": 5.4609,
      "mean_token_accuracy": 0.21524961292743683,
      "num_tokens": 109372165.0,
      "step": 43150
    },
    {
      "entropy": 6.079935550689697,
      "epoch": 4.980380842469706,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028053711696396133,
      "loss": 5.478,
      "mean_token_accuracy": 0.2228081315755844,
      "num_tokens": 109384653.0,
      "step": 43155
    },
    {
      "entropy": 5.981617975234985,
      "epoch": 4.980957876514714,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028049586501644385,
      "loss": 5.4568,
      "mean_token_accuracy": 0.22147443294525146,
      "num_tokens": 109397203.0,
      "step": 43160
    },
    {
      "entropy": 6.013154315948486,
      "epoch": 4.981534910559723,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028045461288407576,
      "loss": 5.4935,
      "mean_token_accuracy": 0.21586854606866837,
      "num_tokens": 109410059.0,
      "step": 43165
    },
    {
      "entropy": 6.0732649803161625,
      "epoch": 4.982111944604732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028041336056824445,
      "loss": 5.5211,
      "mean_token_accuracy": 0.21231584399938583,
      "num_tokens": 109423509.0,
      "step": 43170
    },
    {
      "entropy": 6.084903860092163,
      "epoch": 4.98268897864974,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028037210807033754,
      "loss": 5.5499,
      "mean_token_accuracy": 0.21949520707130432,
      "num_tokens": 109436885.0,
      "step": 43175
    },
    {
      "entropy": 6.051956510543823,
      "epoch": 4.983266012694749,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00028033085539174227,
      "loss": 5.4899,
      "mean_token_accuracy": 0.21759622544050217,
      "num_tokens": 109450583.0,
      "step": 43180
    },
    {
      "entropy": 6.083387994766236,
      "epoch": 4.983843046739757,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002802896025338464,
      "loss": 5.5652,
      "mean_token_accuracy": 0.21513134241104126,
      "num_tokens": 109462769.0,
      "step": 43185
    },
    {
      "entropy": 6.070533704757691,
      "epoch": 4.9844200807847665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002802483494980375,
      "loss": 5.5389,
      "mean_token_accuracy": 0.21402060985565186,
      "num_tokens": 109475565.0,
      "step": 43190
    },
    {
      "entropy": 6.102189350128174,
      "epoch": 4.984997114829775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002802070962857028,
      "loss": 5.505,
      "mean_token_accuracy": 0.21299435645341874,
      "num_tokens": 109488102.0,
      "step": 43195
    },
    {
      "entropy": 6.13311505317688,
      "epoch": 4.985574148874783,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00028016584289823,
      "loss": 5.5641,
      "mean_token_accuracy": 0.21746873259544372,
      "num_tokens": 109500656.0,
      "step": 43200
    },
    {
      "entropy": 6.082438278198242,
      "epoch": 4.986151182919793,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028012458933700674,
      "loss": 5.4965,
      "mean_token_accuracy": 0.2175798863172531,
      "num_tokens": 109514305.0,
      "step": 43205
    },
    {
      "entropy": 6.057926034927368,
      "epoch": 4.986728216964801,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002800833356034204,
      "loss": 5.5214,
      "mean_token_accuracy": 0.21364598125219345,
      "num_tokens": 109526506.0,
      "step": 43210
    },
    {
      "entropy": 6.049296712875366,
      "epoch": 4.9873052510098095,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002800420816988587,
      "loss": 5.4578,
      "mean_token_accuracy": 0.2186858057975769,
      "num_tokens": 109539256.0,
      "step": 43215
    },
    {
      "entropy": 6.113685178756714,
      "epoch": 4.987882285054818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000280000827624709,
      "loss": 5.5207,
      "mean_token_accuracy": 0.21198918521404267,
      "num_tokens": 109550667.0,
      "step": 43220
    },
    {
      "entropy": 6.017138242721558,
      "epoch": 4.988459319099827,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00027995957338235893,
      "loss": 5.4269,
      "mean_token_accuracy": 0.2152409479022026,
      "num_tokens": 109563739.0,
      "step": 43225
    },
    {
      "entropy": 6.085301399230957,
      "epoch": 4.989036353144836,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027991831897319615,
      "loss": 5.5452,
      "mean_token_accuracy": 0.21240639090538024,
      "num_tokens": 109575417.0,
      "step": 43230
    },
    {
      "entropy": 6.048946905136108,
      "epoch": 4.989613387189844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002798770643986081,
      "loss": 5.4895,
      "mean_token_accuracy": 0.2093876227736473,
      "num_tokens": 109588770.0,
      "step": 43235
    },
    {
      "entropy": 6.095553636550903,
      "epoch": 4.990190421234853,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002798358096599824,
      "loss": 5.5944,
      "mean_token_accuracy": 0.21540394723415374,
      "num_tokens": 109600760.0,
      "step": 43240
    },
    {
      "entropy": 6.009504175186157,
      "epoch": 4.990767455279862,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002797945547587066,
      "loss": 5.4494,
      "mean_token_accuracy": 0.21644032150506973,
      "num_tokens": 109613731.0,
      "step": 43245
    },
    {
      "entropy": 6.057441568374633,
      "epoch": 4.99134448932487,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002797532996961684,
      "loss": 5.4726,
      "mean_token_accuracy": 0.2163127601146698,
      "num_tokens": 109626051.0,
      "step": 43250
    },
    {
      "entropy": 6.102918529510498,
      "epoch": 4.991921523369879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002797120444737553,
      "loss": 5.5338,
      "mean_token_accuracy": 0.21369542926549911,
      "num_tokens": 109638095.0,
      "step": 43255
    },
    {
      "entropy": 6.021718120574951,
      "epoch": 4.992498557414887,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027967078909285495,
      "loss": 5.53,
      "mean_token_accuracy": 0.21121876537799836,
      "num_tokens": 109650330.0,
      "step": 43260
    },
    {
      "entropy": 6.084610033035278,
      "epoch": 4.993075591459896,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0002796295335548549,
      "loss": 5.4748,
      "mean_token_accuracy": 0.2149845376610756,
      "num_tokens": 109663537.0,
      "step": 43265
    },
    {
      "entropy": 6.093011045455933,
      "epoch": 4.993652625504905,
      "grad_norm": 0.9375,
      "learning_rate": 0.00027958827786114283,
      "loss": 5.4804,
      "mean_token_accuracy": 0.2195744812488556,
      "num_tokens": 109675382.0,
      "step": 43270
    },
    {
      "entropy": 6.0562254905700685,
      "epoch": 4.994229659549913,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00027954702201310634,
      "loss": 5.5745,
      "mean_token_accuracy": 0.21120424717664718,
      "num_tokens": 109688649.0,
      "step": 43275
    },
    {
      "entropy": 6.116166400909424,
      "epoch": 4.9948066935949225,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002795057660121329,
      "loss": 5.5472,
      "mean_token_accuracy": 0.2126206561923027,
      "num_tokens": 109701483.0,
      "step": 43280
    },
    {
      "entropy": 6.038506937026978,
      "epoch": 4.995383727639931,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00027946450985961036,
      "loss": 5.4858,
      "mean_token_accuracy": 0.2234092429280281,
      "num_tokens": 109715764.0,
      "step": 43285
    },
    {
      "entropy": 6.095676422119141,
      "epoch": 4.995960761684939,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00027942325355692627,
      "loss": 5.5783,
      "mean_token_accuracy": 0.2065901055932045,
      "num_tokens": 109728320.0,
      "step": 43290
    },
    {
      "entropy": 6.054702234268189,
      "epoch": 4.996537795729948,
      "grad_norm": 1.0,
      "learning_rate": 0.00027938199710546817,
      "loss": 5.4618,
      "mean_token_accuracy": 0.2177264451980591,
      "num_tokens": 109740110.0,
      "step": 43295
    },
    {
      "entropy": 6.076013517379761,
      "epoch": 4.997114829774957,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027934074050662384,
      "loss": 5.4899,
      "mean_token_accuracy": 0.2115508198738098,
      "num_tokens": 109753186.0,
      "step": 43300
    },
    {
      "entropy": 6.026408863067627,
      "epoch": 4.997691863819965,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002792994837617808,
      "loss": 5.3889,
      "mean_token_accuracy": 0.22212812453508377,
      "num_tokens": 109766125.0,
      "step": 43305
    },
    {
      "entropy": 6.00216121673584,
      "epoch": 4.998268897864974,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00027925822687232675,
      "loss": 5.4285,
      "mean_token_accuracy": 0.2226571962237358,
      "num_tokens": 109779108.0,
      "step": 43310
    },
    {
      "entropy": 5.972737264633179,
      "epoch": 4.998845931909983,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002792169698396494,
      "loss": 5.46,
      "mean_token_accuracy": 0.21331502348184586,
      "num_tokens": 109791171.0,
      "step": 43315
    },
    {
      "entropy": 6.031192111968994,
      "epoch": 4.999422965954992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002791757126651364,
      "loss": 5.5314,
      "mean_token_accuracy": 0.2099149838089943,
      "num_tokens": 109803198.0,
      "step": 43320
    },
    {
      "entropy": 5.979723358154297,
      "epoch": 5.0,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002791344553501752,
      "loss": 5.4097,
      "mean_token_accuracy": 0.222554911673069,
      "num_tokens": 109815585.0,
      "step": 43325
    },
    {
      "entropy": 6.057002878189087,
      "epoch": 5.000577034045008,
      "grad_norm": 0.875,
      "learning_rate": 0.00027909319789615384,
      "loss": 5.411,
      "mean_token_accuracy": 0.21840555667877198,
      "num_tokens": 109829087.0,
      "step": 43330
    },
    {
      "entropy": 6.0135149478912355,
      "epoch": 5.001154068090018,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002790519403044597,
      "loss": 5.3842,
      "mean_token_accuracy": 0.2211799815297127,
      "num_tokens": 109843102.0,
      "step": 43335
    },
    {
      "entropy": 6.011963129043579,
      "epoch": 5.001731102135026,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027901068257648063,
      "loss": 5.3962,
      "mean_token_accuracy": 0.2208527460694313,
      "num_tokens": 109855373.0,
      "step": 43340
    },
    {
      "entropy": 6.0466187477111815,
      "epoch": 5.002308136180035,
      "grad_norm": 0.9375,
      "learning_rate": 0.00027896942471360413,
      "loss": 5.4358,
      "mean_token_accuracy": 0.21855471283197403,
      "num_tokens": 109868291.0,
      "step": 43345
    },
    {
      "entropy": 6.09973292350769,
      "epoch": 5.002885170225043,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00027892816671721817,
      "loss": 5.5186,
      "mean_token_accuracy": 0.21402830481529236,
      "num_tokens": 109881775.0,
      "step": 43350
    },
    {
      "entropy": 6.074065446853638,
      "epoch": 5.003462204270052,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002788869085887102,
      "loss": 5.4648,
      "mean_token_accuracy": 0.21933338642120362,
      "num_tokens": 109893968.0,
      "step": 43355
    },
    {
      "entropy": 6.00483021736145,
      "epoch": 5.004039238315061,
      "grad_norm": 1.0,
      "learning_rate": 0.00027884565032946796,
      "loss": 5.386,
      "mean_token_accuracy": 0.22136774957180022,
      "num_tokens": 109906610.0,
      "step": 43360
    },
    {
      "entropy": 6.038157987594604,
      "epoch": 5.004616272360069,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002788043919408791,
      "loss": 5.4556,
      "mean_token_accuracy": 0.22546171098947526,
      "num_tokens": 109919641.0,
      "step": 43365
    },
    {
      "entropy": 6.093091297149658,
      "epoch": 5.0051933064050775,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027876313342433155,
      "loss": 5.4462,
      "mean_token_accuracy": 0.21399179100990295,
      "num_tokens": 109930948.0,
      "step": 43370
    },
    {
      "entropy": 5.999747323989868,
      "epoch": 5.005770340450087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027872187478121286,
      "loss": 5.3871,
      "mean_token_accuracy": 0.22151034474372863,
      "num_tokens": 109943885.0,
      "step": 43375
    },
    {
      "entropy": 6.021550607681275,
      "epoch": 5.006347374495095,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002786806160129108,
      "loss": 5.4329,
      "mean_token_accuracy": 0.2237292245030403,
      "num_tokens": 109955418.0,
      "step": 43380
    },
    {
      "entropy": 6.058311700820923,
      "epoch": 5.006924408540104,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002786393571208129,
      "loss": 5.4124,
      "mean_token_accuracy": 0.22067750245332718,
      "num_tokens": 109967392.0,
      "step": 43385
    },
    {
      "entropy": 6.070983219146728,
      "epoch": 5.007501442585112,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027859809810630717,
      "loss": 5.4457,
      "mean_token_accuracy": 0.21794684380292892,
      "num_tokens": 109979129.0,
      "step": 43390
    },
    {
      "entropy": 6.035621166229248,
      "epoch": 5.008078476630121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002785568389707811,
      "loss": 5.4222,
      "mean_token_accuracy": 0.21813954561948776,
      "num_tokens": 109990672.0,
      "step": 43395
    },
    {
      "entropy": 5.960146522521972,
      "epoch": 5.00865551067513,
      "grad_norm": 1.0,
      "learning_rate": 0.0002785155797156226,
      "loss": 5.3678,
      "mean_token_accuracy": 0.22357719093561174,
      "num_tokens": 110005009.0,
      "step": 43400
    },
    {
      "entropy": 6.027435159683227,
      "epoch": 5.009232544720138,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002784743203422192,
      "loss": 5.3714,
      "mean_token_accuracy": 0.22740354835987092,
      "num_tokens": 110017239.0,
      "step": 43405
    },
    {
      "entropy": 5.953833150863647,
      "epoch": 5.0098095787651475,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00027843306085195886,
      "loss": 5.2906,
      "mean_token_accuracy": 0.22527462989091873,
      "num_tokens": 110029789.0,
      "step": 43410
    },
    {
      "entropy": 6.080698251724243,
      "epoch": 5.010386612810156,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002783918012462292,
      "loss": 5.5204,
      "mean_token_accuracy": 0.2130571871995926,
      "num_tokens": 110042094.0,
      "step": 43415
    },
    {
      "entropy": 6.062127113342285,
      "epoch": 5.010963646855164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000278350541526418,
      "loss": 5.4487,
      "mean_token_accuracy": 0.22809096574783325,
      "num_tokens": 110055464.0,
      "step": 43420
    },
    {
      "entropy": 6.0731995582580565,
      "epoch": 5.011540680900173,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027830928169391295,
      "loss": 5.462,
      "mean_token_accuracy": 0.21829243004322052,
      "num_tokens": 110067771.0,
      "step": 43425
    },
    {
      "entropy": 6.026186466217041,
      "epoch": 5.012117714945182,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002782680217501019,
      "loss": 5.3742,
      "mean_token_accuracy": 0.22001130729913712,
      "num_tokens": 110079746.0,
      "step": 43430
    },
    {
      "entropy": 6.065733003616333,
      "epoch": 5.0126947489901905,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00027822676169637255,
      "loss": 5.4586,
      "mean_token_accuracy": 0.21646685749292374,
      "num_tokens": 110091970.0,
      "step": 43435
    },
    {
      "entropy": 6.041999244689942,
      "epoch": 5.013271783035199,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027818550153411276,
      "loss": 5.3826,
      "mean_token_accuracy": 0.22398083806037902,
      "num_tokens": 110104545.0,
      "step": 43440
    },
    {
      "entropy": 6.009548568725586,
      "epoch": 5.013848817080207,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0002781442412647101,
      "loss": 5.3611,
      "mean_token_accuracy": 0.23100339472293854,
      "num_tokens": 110117280.0,
      "step": 43445
    },
    {
      "entropy": 6.00257682800293,
      "epoch": 5.014425851125217,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002781029808895525,
      "loss": 5.4074,
      "mean_token_accuracy": 0.2265898421406746,
      "num_tokens": 110130526.0,
      "step": 43450
    },
    {
      "entropy": 6.010365962982178,
      "epoch": 5.015002885170225,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027806172041002766,
      "loss": 5.427,
      "mean_token_accuracy": 0.2144506588578224,
      "num_tokens": 110143336.0,
      "step": 43455
    },
    {
      "entropy": 6.060679864883423,
      "epoch": 5.0155799192152335,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027802045982752336,
      "loss": 5.4366,
      "mean_token_accuracy": 0.21671856790781022,
      "num_tokens": 110155220.0,
      "step": 43460
    },
    {
      "entropy": 5.9916623592376705,
      "epoch": 5.016156953260242,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00027797919914342736,
      "loss": 5.3761,
      "mean_token_accuracy": 0.22382241487503052,
      "num_tokens": 110169163.0,
      "step": 43465
    },
    {
      "entropy": 5.984676265716553,
      "epoch": 5.016733987305251,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002779379383591275,
      "loss": 5.3955,
      "mean_token_accuracy": 0.22768711745738984,
      "num_tokens": 110181887.0,
      "step": 43470
    },
    {
      "entropy": 6.030897855758667,
      "epoch": 5.01731102135026,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002778966774760116,
      "loss": 5.3853,
      "mean_token_accuracy": 0.228704833984375,
      "num_tokens": 110194402.0,
      "step": 43475
    },
    {
      "entropy": 6.107543134689331,
      "epoch": 5.017888055395268,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027785541649546743,
      "loss": 5.5118,
      "mean_token_accuracy": 0.21228811740875245,
      "num_tokens": 110207521.0,
      "step": 43480
    },
    {
      "entropy": 6.047583723068238,
      "epoch": 5.018465089440277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002778141554188826,
      "loss": 5.4229,
      "mean_token_accuracy": 0.22110394090414048,
      "num_tokens": 110219240.0,
      "step": 43485
    },
    {
      "entropy": 6.052866363525391,
      "epoch": 5.019042123485286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027777289424764514,
      "loss": 5.4499,
      "mean_token_accuracy": 0.21959130465984344,
      "num_tokens": 110231186.0,
      "step": 43490
    },
    {
      "entropy": 6.035445356369019,
      "epoch": 5.019619157530294,
      "grad_norm": 1.0,
      "learning_rate": 0.00027773163298314276,
      "loss": 5.452,
      "mean_token_accuracy": 0.22017614245414735,
      "num_tokens": 110243185.0,
      "step": 43495
    },
    {
      "entropy": 6.103040552139282,
      "epoch": 5.020196191575303,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002776903716267632,
      "loss": 5.4572,
      "mean_token_accuracy": 0.21248776316642762,
      "num_tokens": 110254840.0,
      "step": 43500
    },
    {
      "entropy": 6.074078559875488,
      "epoch": 5.020773225620312,
      "grad_norm": 1.0,
      "learning_rate": 0.00027764911017989443,
      "loss": 5.4705,
      "mean_token_accuracy": 0.2233603611588478,
      "num_tokens": 110267263.0,
      "step": 43505
    },
    {
      "entropy": 6.103983449935913,
      "epoch": 5.02135025966532,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027760784864392404,
      "loss": 5.5296,
      "mean_token_accuracy": 0.21100655496120452,
      "num_tokens": 110280057.0,
      "step": 43510
    },
    {
      "entropy": 6.048934555053711,
      "epoch": 5.021927293710329,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027756658702024,
      "loss": 5.4041,
      "mean_token_accuracy": 0.21984410434961318,
      "num_tokens": 110292827.0,
      "step": 43515
    },
    {
      "entropy": 5.979860544204712,
      "epoch": 5.022504327755337,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002775253253102301,
      "loss": 5.306,
      "mean_token_accuracy": 0.22724293917417526,
      "num_tokens": 110305449.0,
      "step": 43520
    },
    {
      "entropy": 6.053369379043579,
      "epoch": 5.0230813618003465,
      "grad_norm": 1.0,
      "learning_rate": 0.00027748406351528216,
      "loss": 5.4624,
      "mean_token_accuracy": 0.21502453833818436,
      "num_tokens": 110317922.0,
      "step": 43525
    },
    {
      "entropy": 6.0136621475219725,
      "epoch": 5.023658395845355,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002774428016367839,
      "loss": 5.4129,
      "mean_token_accuracy": 0.22424203306436538,
      "num_tokens": 110330377.0,
      "step": 43530
    },
    {
      "entropy": 6.015083312988281,
      "epoch": 5.024235429890363,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027740153967612327,
      "loss": 5.4141,
      "mean_token_accuracy": 0.223326875269413,
      "num_tokens": 110343353.0,
      "step": 43535
    },
    {
      "entropy": 6.013021326065063,
      "epoch": 5.024812463935373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00027736027763468807,
      "loss": 5.4195,
      "mean_token_accuracy": 0.22434018552303314,
      "num_tokens": 110355953.0,
      "step": 43540
    },
    {
      "entropy": 6.14027214050293,
      "epoch": 5.025389497980381,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027731901551386606,
      "loss": 5.549,
      "mean_token_accuracy": 0.21244252920150758,
      "num_tokens": 110367905.0,
      "step": 43545
    },
    {
      "entropy": 6.060546398162842,
      "epoch": 5.0259665320253895,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002772777533150451,
      "loss": 5.4001,
      "mean_token_accuracy": 0.22276355773210527,
      "num_tokens": 110381255.0,
      "step": 43550
    },
    {
      "entropy": 6.036908960342407,
      "epoch": 5.026543566070398,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002772364910396131,
      "loss": 5.4117,
      "mean_token_accuracy": 0.22621753811836243,
      "num_tokens": 110393287.0,
      "step": 43555
    },
    {
      "entropy": 6.002397584915161,
      "epoch": 5.027120600115407,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027719522868895774,
      "loss": 5.4344,
      "mean_token_accuracy": 0.221749347448349,
      "num_tokens": 110406378.0,
      "step": 43560
    },
    {
      "entropy": 6.049661874771118,
      "epoch": 5.027697634160416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000277153966264467,
      "loss": 5.3826,
      "mean_token_accuracy": 0.22492537200450896,
      "num_tokens": 110419153.0,
      "step": 43565
    },
    {
      "entropy": 6.08701982498169,
      "epoch": 5.028274668205424,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002771127037675288,
      "loss": 5.4385,
      "mean_token_accuracy": 0.22303762584924697,
      "num_tokens": 110431418.0,
      "step": 43570
    },
    {
      "entropy": 5.97638430595398,
      "epoch": 5.028851702250432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002770714411995307,
      "loss": 5.428,
      "mean_token_accuracy": 0.22077428847551345,
      "num_tokens": 110444999.0,
      "step": 43575
    },
    {
      "entropy": 6.061025905609131,
      "epoch": 5.029428736295442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027703017856186077,
      "loss": 5.4722,
      "mean_token_accuracy": 0.2129136011004448,
      "num_tokens": 110456263.0,
      "step": 43580
    },
    {
      "entropy": 6.047892427444458,
      "epoch": 5.03000577034045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002769889158559067,
      "loss": 5.4443,
      "mean_token_accuracy": 0.2185959190130234,
      "num_tokens": 110469019.0,
      "step": 43585
    },
    {
      "entropy": 6.002608871459961,
      "epoch": 5.030582804385459,
      "grad_norm": 1.0,
      "learning_rate": 0.0002769476530830566,
      "loss": 5.4108,
      "mean_token_accuracy": 0.2232239380478859,
      "num_tokens": 110481046.0,
      "step": 43590
    },
    {
      "entropy": 6.076695919036865,
      "epoch": 5.031159838430467,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000276906390244698,
      "loss": 5.4771,
      "mean_token_accuracy": 0.21442841589450837,
      "num_tokens": 110495091.0,
      "step": 43595
    },
    {
      "entropy": 6.061804246902466,
      "epoch": 5.031736872475476,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027686512734221896,
      "loss": 5.4872,
      "mean_token_accuracy": 0.219915272295475,
      "num_tokens": 110507465.0,
      "step": 43600
    },
    {
      "entropy": 6.036294889450073,
      "epoch": 5.032313906520485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027682386437700735,
      "loss": 5.4281,
      "mean_token_accuracy": 0.22256449609994888,
      "num_tokens": 110519564.0,
      "step": 43605
    },
    {
      "entropy": 6.070567226409912,
      "epoch": 5.032890940565493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027678260135045093,
      "loss": 5.3928,
      "mean_token_accuracy": 0.22905297130346297,
      "num_tokens": 110532961.0,
      "step": 43610
    },
    {
      "entropy": 6.1446428298950195,
      "epoch": 5.033467974610502,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027674133826393753,
      "loss": 5.4993,
      "mean_token_accuracy": 0.21308950036764146,
      "num_tokens": 110544445.0,
      "step": 43615
    },
    {
      "entropy": 6.089685678482056,
      "epoch": 5.034045008655511,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002767000751188552,
      "loss": 5.498,
      "mean_token_accuracy": 0.2153544694185257,
      "num_tokens": 110558462.0,
      "step": 43620
    },
    {
      "entropy": 5.9756715297698975,
      "epoch": 5.034622042700519,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002766588119165916,
      "loss": 5.3546,
      "mean_token_accuracy": 0.2213977411389351,
      "num_tokens": 110571071.0,
      "step": 43625
    },
    {
      "entropy": 6.030717897415161,
      "epoch": 5.035199076745528,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002766175486585348,
      "loss": 5.4131,
      "mean_token_accuracy": 0.22668559104204178,
      "num_tokens": 110583430.0,
      "step": 43630
    },
    {
      "entropy": 6.100792455673218,
      "epoch": 5.035776110790537,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002765762853460724,
      "loss": 5.4906,
      "mean_token_accuracy": 0.21438062191009521,
      "num_tokens": 110596730.0,
      "step": 43635
    },
    {
      "entropy": 6.107109117507934,
      "epoch": 5.036353144835545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002765350219805925,
      "loss": 5.4758,
      "mean_token_accuracy": 0.21680520325899125,
      "num_tokens": 110608793.0,
      "step": 43640
    },
    {
      "entropy": 6.078278303146362,
      "epoch": 5.036930178880554,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00027649375856348286,
      "loss": 5.469,
      "mean_token_accuracy": 0.21492241621017455,
      "num_tokens": 110621432.0,
      "step": 43645
    },
    {
      "entropy": 5.941734361648559,
      "epoch": 5.037507212925562,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027645249509613147,
      "loss": 5.4082,
      "mean_token_accuracy": 0.22579283863306046,
      "num_tokens": 110634059.0,
      "step": 43650
    },
    {
      "entropy": 6.008061933517456,
      "epoch": 5.0380842469705716,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027641123157992603,
      "loss": 5.4025,
      "mean_token_accuracy": 0.22261736840009688,
      "num_tokens": 110647070.0,
      "step": 43655
    },
    {
      "entropy": 6.00016417503357,
      "epoch": 5.03866128101558,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00027636996801625466,
      "loss": 5.4018,
      "mean_token_accuracy": 0.21993223428726197,
      "num_tokens": 110660062.0,
      "step": 43660
    },
    {
      "entropy": 6.1118992328643795,
      "epoch": 5.039238315060588,
      "grad_norm": 0.9375,
      "learning_rate": 0.00027632870440650497,
      "loss": 5.4389,
      "mean_token_accuracy": 0.22095718383789062,
      "num_tokens": 110673009.0,
      "step": 43665
    },
    {
      "entropy": 5.991368150711059,
      "epoch": 5.039815349105597,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000276287440752065,
      "loss": 5.4128,
      "mean_token_accuracy": 0.22291670441627504,
      "num_tokens": 110685180.0,
      "step": 43670
    },
    {
      "entropy": 6.073290252685547,
      "epoch": 5.040392383150606,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002762461770543226,
      "loss": 5.439,
      "mean_token_accuracy": 0.224125699698925,
      "num_tokens": 110697883.0,
      "step": 43675
    },
    {
      "entropy": 5.9949424266815186,
      "epoch": 5.0409694171956145,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027620491331466565,
      "loss": 5.3806,
      "mean_token_accuracy": 0.22720801532268525,
      "num_tokens": 110710916.0,
      "step": 43680
    },
    {
      "entropy": 6.005471849441529,
      "epoch": 5.041546451240623,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002761636495344821,
      "loss": 5.3962,
      "mean_token_accuracy": 0.22628005594015121,
      "num_tokens": 110723291.0,
      "step": 43685
    },
    {
      "entropy": 5.998024559020996,
      "epoch": 5.042123485285632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027612238571515975,
      "loss": 5.3743,
      "mean_token_accuracy": 0.22809399515390397,
      "num_tokens": 110736008.0,
      "step": 43690
    },
    {
      "entropy": 6.045644950866699,
      "epoch": 5.042700519330641,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027608112185808646,
      "loss": 5.4274,
      "mean_token_accuracy": 0.22308100312948226,
      "num_tokens": 110749464.0,
      "step": 43695
    },
    {
      "entropy": 6.015965366363526,
      "epoch": 5.043277553375649,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002760398579646502,
      "loss": 5.4474,
      "mean_token_accuracy": 0.2218312442302704,
      "num_tokens": 110762437.0,
      "step": 43700
    },
    {
      "entropy": 6.03645691871643,
      "epoch": 5.0438545874206575,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00027599859403623883,
      "loss": 5.4247,
      "mean_token_accuracy": 0.2206249788403511,
      "num_tokens": 110775364.0,
      "step": 43705
    },
    {
      "entropy": 6.012313747406006,
      "epoch": 5.044431621465667,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002759573300742403,
      "loss": 5.3688,
      "mean_token_accuracy": 0.2273314744234085,
      "num_tokens": 110787668.0,
      "step": 43710
    },
    {
      "entropy": 6.045507001876831,
      "epoch": 5.045008655510675,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002759160660800424,
      "loss": 5.4285,
      "mean_token_accuracy": 0.21761375963687896,
      "num_tokens": 110799522.0,
      "step": 43715
    },
    {
      "entropy": 6.0433127880096436,
      "epoch": 5.045585689555684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000275874802055033,
      "loss": 5.3969,
      "mean_token_accuracy": 0.22407684326171876,
      "num_tokens": 110811576.0,
      "step": 43720
    },
    {
      "entropy": 5.968475580215454,
      "epoch": 5.046162723600692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002758335380006002,
      "loss": 5.3672,
      "mean_token_accuracy": 0.22569257915019988,
      "num_tokens": 110823824.0,
      "step": 43725
    },
    {
      "entropy": 6.141472578048706,
      "epoch": 5.046739757645701,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002757922739181317,
      "loss": 5.5027,
      "mean_token_accuracy": 0.2123544842004776,
      "num_tokens": 110837656.0,
      "step": 43730
    },
    {
      "entropy": 6.100065803527832,
      "epoch": 5.04731679169071,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00027575100980901544,
      "loss": 5.4927,
      "mean_token_accuracy": 0.2117149218916893,
      "num_tokens": 110852148.0,
      "step": 43735
    },
    {
      "entropy": 6.093953561782837,
      "epoch": 5.047893825735718,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027570974567463934,
      "loss": 5.5321,
      "mean_token_accuracy": 0.2087482437491417,
      "num_tokens": 110865884.0,
      "step": 43740
    },
    {
      "entropy": 6.048941373825073,
      "epoch": 5.048470859780727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027566848151639134,
      "loss": 5.4201,
      "mean_token_accuracy": 0.22149043828248977,
      "num_tokens": 110877512.0,
      "step": 43745
    },
    {
      "entropy": 6.0919266700744625,
      "epoch": 5.049047893825736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002756272173356593,
      "loss": 5.4356,
      "mean_token_accuracy": 0.2235841080546379,
      "num_tokens": 110890499.0,
      "step": 43750
    },
    {
      "entropy": 6.026390504837036,
      "epoch": 5.049624927870744,
      "grad_norm": 1.0,
      "learning_rate": 0.0002755859531338311,
      "loss": 5.4783,
      "mean_token_accuracy": 0.2191991999745369,
      "num_tokens": 110903376.0,
      "step": 43755
    },
    {
      "entropy": 6.04155707359314,
      "epoch": 5.050201961915753,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00027554468891229463,
      "loss": 5.446,
      "mean_token_accuracy": 0.21936483234167098,
      "num_tokens": 110916376.0,
      "step": 43760
    },
    {
      "entropy": 6.027285242080689,
      "epoch": 5.050778995960762,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002755034246724378,
      "loss": 5.47,
      "mean_token_accuracy": 0.2123220220208168,
      "num_tokens": 110928719.0,
      "step": 43765
    },
    {
      "entropy": 6.05591197013855,
      "epoch": 5.0513560300057705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00027546216041564856,
      "loss": 5.4302,
      "mean_token_accuracy": 0.2223159059882164,
      "num_tokens": 110941361.0,
      "step": 43770
    },
    {
      "entropy": 6.076092004776001,
      "epoch": 5.051933064050779,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002754208961433149,
      "loss": 5.4691,
      "mean_token_accuracy": 0.22302059978246688,
      "num_tokens": 110954068.0,
      "step": 43775
    },
    {
      "entropy": 5.895540618896485,
      "epoch": 5.052510098095787,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002753796318568244,
      "loss": 5.2876,
      "mean_token_accuracy": 0.23152270168066025,
      "num_tokens": 110967819.0,
      "step": 43780
    },
    {
      "entropy": 6.023781394958496,
      "epoch": 5.053087132140797,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002753383675575653,
      "loss": 5.4126,
      "mean_token_accuracy": 0.22144342511892318,
      "num_tokens": 110981153.0,
      "step": 43785
    },
    {
      "entropy": 6.085490942001343,
      "epoch": 5.053664166185805,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027529710324692536,
      "loss": 5.5039,
      "mean_token_accuracy": 0.21074877828359603,
      "num_tokens": 110993242.0,
      "step": 43790
    },
    {
      "entropy": 6.038920068740845,
      "epoch": 5.0542412002308135,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027525583892629255,
      "loss": 5.3804,
      "mean_token_accuracy": 0.2290875256061554,
      "num_tokens": 111006937.0,
      "step": 43795
    },
    {
      "entropy": 5.986760091781616,
      "epoch": 5.054818234275822,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027521457459705464,
      "loss": 5.4085,
      "mean_token_accuracy": 0.22843277156352998,
      "num_tokens": 111019942.0,
      "step": 43800
    },
    {
      "entropy": 6.02552604675293,
      "epoch": 5.055395268320831,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027517331026059963,
      "loss": 5.3876,
      "mean_token_accuracy": 0.21961864829063416,
      "num_tokens": 111031199.0,
      "step": 43805
    },
    {
      "entropy": 6.061905431747436,
      "epoch": 5.05597230236584,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002751320459183154,
      "loss": 5.4812,
      "mean_token_accuracy": 0.21339097172021865,
      "num_tokens": 111043252.0,
      "step": 43810
    },
    {
      "entropy": 6.060674571990967,
      "epoch": 5.056549336410848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002750907815715899,
      "loss": 5.4729,
      "mean_token_accuracy": 0.22339835464954377,
      "num_tokens": 111054969.0,
      "step": 43815
    },
    {
      "entropy": 6.088775539398194,
      "epoch": 5.057126370455857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00027504951722181104,
      "loss": 5.4099,
      "mean_token_accuracy": 0.22435894310474397,
      "num_tokens": 111067028.0,
      "step": 43820
    },
    {
      "entropy": 6.014452743530273,
      "epoch": 5.057703404500866,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002750082528703666,
      "loss": 5.4367,
      "mean_token_accuracy": 0.22171470671892166,
      "num_tokens": 111079317.0,
      "step": 43825
    },
    {
      "entropy": 5.963097953796387,
      "epoch": 5.058280438545874,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002749669885186447,
      "loss": 5.3484,
      "mean_token_accuracy": 0.22546017915010452,
      "num_tokens": 111092052.0,
      "step": 43830
    },
    {
      "entropy": 6.066402864456177,
      "epoch": 5.058857472590883,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002749257241680331,
      "loss": 5.4791,
      "mean_token_accuracy": 0.2186070203781128,
      "num_tokens": 111104616.0,
      "step": 43835
    },
    {
      "entropy": 5.969967842102051,
      "epoch": 5.059434506635892,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002748844598199196,
      "loss": 5.3698,
      "mean_token_accuracy": 0.22845522463321685,
      "num_tokens": 111118394.0,
      "step": 43840
    },
    {
      "entropy": 6.05379638671875,
      "epoch": 5.0600115406809,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00027484319547569237,
      "loss": 5.4199,
      "mean_token_accuracy": 0.2178450882434845,
      "num_tokens": 111130812.0,
      "step": 43845
    },
    {
      "entropy": 6.103090095520019,
      "epoch": 5.060588574725909,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002748019311367391,
      "loss": 5.4998,
      "mean_token_accuracy": 0.21108583956956864,
      "num_tokens": 111143332.0,
      "step": 43850
    },
    {
      "entropy": 6.078620862960816,
      "epoch": 5.061165608770917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002747606668044479,
      "loss": 5.4217,
      "mean_token_accuracy": 0.2254311040043831,
      "num_tokens": 111156273.0,
      "step": 43855
    },
    {
      "entropy": 6.064867258071899,
      "epoch": 5.0617426428159265,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027471940248020637,
      "loss": 5.4615,
      "mean_token_accuracy": 0.21778440475463867,
      "num_tokens": 111169408.0,
      "step": 43860
    },
    {
      "entropy": 5.954033660888672,
      "epoch": 5.062319676860935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027467813816540277,
      "loss": 5.348,
      "mean_token_accuracy": 0.22130090445280076,
      "num_tokens": 111181346.0,
      "step": 43865
    },
    {
      "entropy": 6.066328668594361,
      "epoch": 5.062896710905943,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002746368738614247,
      "loss": 5.4252,
      "mean_token_accuracy": 0.218902787566185,
      "num_tokens": 111193013.0,
      "step": 43870
    },
    {
      "entropy": 6.0883392810821535,
      "epoch": 5.063473744950952,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027459560956966036,
      "loss": 5.5008,
      "mean_token_accuracy": 0.21469320207834244,
      "num_tokens": 111205255.0,
      "step": 43875
    },
    {
      "entropy": 6.073800945281983,
      "epoch": 5.064050778995961,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002745543452914973,
      "loss": 5.4458,
      "mean_token_accuracy": 0.22309898883104323,
      "num_tokens": 111218825.0,
      "step": 43880
    },
    {
      "entropy": 6.072566986083984,
      "epoch": 5.064627813040969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027451308102832375,
      "loss": 5.4736,
      "mean_token_accuracy": 0.22132294476032258,
      "num_tokens": 111231655.0,
      "step": 43885
    },
    {
      "entropy": 6.0426851272583,
      "epoch": 5.065204847085978,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002744718167815274,
      "loss": 5.4215,
      "mean_token_accuracy": 0.2223377600312233,
      "num_tokens": 111245039.0,
      "step": 43890
    },
    {
      "entropy": 6.04015703201294,
      "epoch": 5.065781881130987,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002744305525524963,
      "loss": 5.4903,
      "mean_token_accuracy": 0.21566438525915146,
      "num_tokens": 111257790.0,
      "step": 43895
    },
    {
      "entropy": 6.1207441806793215,
      "epoch": 5.066358915175996,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002743892883426182,
      "loss": 5.5735,
      "mean_token_accuracy": 0.21036375612020491,
      "num_tokens": 111270435.0,
      "step": 43900
    },
    {
      "entropy": 6.15555214881897,
      "epoch": 5.066935949221004,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027434802415328115,
      "loss": 5.5271,
      "mean_token_accuracy": 0.21542391180992126,
      "num_tokens": 111283418.0,
      "step": 43905
    },
    {
      "entropy": 6.092133522033691,
      "epoch": 5.067512983266012,
      "grad_norm": 1.0,
      "learning_rate": 0.00027430675998587294,
      "loss": 5.4792,
      "mean_token_accuracy": 0.21991305351257323,
      "num_tokens": 111296717.0,
      "step": 43910
    },
    {
      "entropy": 5.90102252960205,
      "epoch": 5.068090017311022,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027426549584178163,
      "loss": 5.2205,
      "mean_token_accuracy": 0.2448872983455658,
      "num_tokens": 111310222.0,
      "step": 43915
    },
    {
      "entropy": 6.018193483352661,
      "epoch": 5.06866705135603,
      "grad_norm": 1.0,
      "learning_rate": 0.00027422423172239485,
      "loss": 5.4635,
      "mean_token_accuracy": 0.21214070320129394,
      "num_tokens": 111321838.0,
      "step": 43920
    },
    {
      "entropy": 5.953082609176636,
      "epoch": 5.0692440854010385,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027418296762910075,
      "loss": 5.3204,
      "mean_token_accuracy": 0.2320536643266678,
      "num_tokens": 111335297.0,
      "step": 43925
    },
    {
      "entropy": 6.072371149063111,
      "epoch": 5.069821119446047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000274141703563287,
      "loss": 5.4374,
      "mean_token_accuracy": 0.21538564413785935,
      "num_tokens": 111346571.0,
      "step": 43930
    },
    {
      "entropy": 5.9878864765167235,
      "epoch": 5.070398153491056,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002741004395263418,
      "loss": 5.3723,
      "mean_token_accuracy": 0.22695438116788863,
      "num_tokens": 111359080.0,
      "step": 43935
    },
    {
      "entropy": 6.03265528678894,
      "epoch": 5.070975187536065,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002740591755196527,
      "loss": 5.3929,
      "mean_token_accuracy": 0.2247341215610504,
      "num_tokens": 111372013.0,
      "step": 43940
    },
    {
      "entropy": 6.024111032485962,
      "epoch": 5.071552221581073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002740179115446079,
      "loss": 5.408,
      "mean_token_accuracy": 0.22659876495599746,
      "num_tokens": 111384097.0,
      "step": 43945
    },
    {
      "entropy": 6.062497806549072,
      "epoch": 5.0721292556260815,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002739766476025951,
      "loss": 5.4394,
      "mean_token_accuracy": 0.21750799417495728,
      "num_tokens": 111397025.0,
      "step": 43950
    },
    {
      "entropy": 6.1084648132324215,
      "epoch": 5.072706289671091,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027393538369500226,
      "loss": 5.4946,
      "mean_token_accuracy": 0.21657062768936158,
      "num_tokens": 111409256.0,
      "step": 43955
    },
    {
      "entropy": 5.967855548858642,
      "epoch": 5.073283323716099,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002738941198232172,
      "loss": 5.3623,
      "mean_token_accuracy": 0.23110650926828386,
      "num_tokens": 111422401.0,
      "step": 43960
    },
    {
      "entropy": 6.00794825553894,
      "epoch": 5.073860357761108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027385285598862795,
      "loss": 5.4588,
      "mean_token_accuracy": 0.22129245698451996,
      "num_tokens": 111434855.0,
      "step": 43965
    },
    {
      "entropy": 6.047057104110718,
      "epoch": 5.074437391806117,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002738115921926223,
      "loss": 5.4187,
      "mean_token_accuracy": 0.21995000541210175,
      "num_tokens": 111448089.0,
      "step": 43970
    },
    {
      "entropy": 6.019506359100342,
      "epoch": 5.075014425851125,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002737703284365882,
      "loss": 5.3858,
      "mean_token_accuracy": 0.2197961390018463,
      "num_tokens": 111460290.0,
      "step": 43975
    },
    {
      "entropy": 5.972984981536865,
      "epoch": 5.075591459896134,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002737290647219134,
      "loss": 5.3676,
      "mean_token_accuracy": 0.22668742537498474,
      "num_tokens": 111473901.0,
      "step": 43980
    },
    {
      "entropy": 6.074132013320923,
      "epoch": 5.076168493941142,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00027368780104998594,
      "loss": 5.4614,
      "mean_token_accuracy": 0.22270818054676056,
      "num_tokens": 111486309.0,
      "step": 43985
    },
    {
      "entropy": 6.084434890747071,
      "epoch": 5.0767455279861515,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002736465374221936,
      "loss": 5.5097,
      "mean_token_accuracy": 0.20677373558282852,
      "num_tokens": 111498220.0,
      "step": 43990
    },
    {
      "entropy": 6.0118903636932375,
      "epoch": 5.07732256203116,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0002736052738399243,
      "loss": 5.4077,
      "mean_token_accuracy": 0.2176373153924942,
      "num_tokens": 111511789.0,
      "step": 43995
    },
    {
      "entropy": 6.067039680480957,
      "epoch": 5.077899596076168,
      "grad_norm": 1.1875,
      "learning_rate": 0.00027356401030456597,
      "loss": 5.439,
      "mean_token_accuracy": 0.22006991356611252,
      "num_tokens": 111524304.0,
      "step": 44000
    },
    {
      "entropy": 6.071492624282837,
      "epoch": 5.078476630121177,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002735227468175064,
      "loss": 5.4405,
      "mean_token_accuracy": 0.21808840632438659,
      "num_tokens": 111537149.0,
      "step": 44005
    },
    {
      "entropy": 6.1197710037231445,
      "epoch": 5.079053664166186,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002734814833801335,
      "loss": 5.543,
      "mean_token_accuracy": 0.21046993136405945,
      "num_tokens": 111551166.0,
      "step": 44010
    },
    {
      "entropy": 6.014575672149658,
      "epoch": 5.0796306982111945,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002734402199938352,
      "loss": 5.3542,
      "mean_token_accuracy": 0.23357006460428237,
      "num_tokens": 111563982.0,
      "step": 44015
    },
    {
      "entropy": 6.074667358398438,
      "epoch": 5.080207732256203,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002733989566599993,
      "loss": 5.3914,
      "mean_token_accuracy": 0.22708164006471634,
      "num_tokens": 111577298.0,
      "step": 44020
    },
    {
      "entropy": 6.018869066238404,
      "epoch": 5.080784766301211,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00027335769338001373,
      "loss": 5.4182,
      "mean_token_accuracy": 0.22038844972848892,
      "num_tokens": 111590044.0,
      "step": 44025
    },
    {
      "entropy": 5.94426326751709,
      "epoch": 5.081361800346221,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00027331643015526623,
      "loss": 5.2582,
      "mean_token_accuracy": 0.24838035106658934,
      "num_tokens": 111603811.0,
      "step": 44030
    },
    {
      "entropy": 5.89699273109436,
      "epoch": 5.081938834391229,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002732751669871449,
      "loss": 5.3123,
      "mean_token_accuracy": 0.23269784897565843,
      "num_tokens": 111615198.0,
      "step": 44035
    },
    {
      "entropy": 5.9873229503631595,
      "epoch": 5.0825158684362375,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00027323390387703735,
      "loss": 5.341,
      "mean_token_accuracy": 0.2299254059791565,
      "num_tokens": 111627873.0,
      "step": 44040
    },
    {
      "entropy": 6.038514089584351,
      "epoch": 5.083092902481247,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002731926408263317,
      "loss": 5.4665,
      "mean_token_accuracy": 0.21882090121507644,
      "num_tokens": 111640493.0,
      "step": 44045
    },
    {
      "entropy": 5.97344479560852,
      "epoch": 5.083669936526255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002731513778364155,
      "loss": 5.3725,
      "mean_token_accuracy": 0.23172066509723663,
      "num_tokens": 111652701.0,
      "step": 44050
    },
    {
      "entropy": 6.013360834121704,
      "epoch": 5.084246970571264,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027311011490867694,
      "loss": 5.4463,
      "mean_token_accuracy": 0.21674323529005052,
      "num_tokens": 111664467.0,
      "step": 44055
    },
    {
      "entropy": 6.002835416793824,
      "epoch": 5.084824004616272,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002730688520445036,
      "loss": 5.3616,
      "mean_token_accuracy": 0.22356657832860946,
      "num_tokens": 111677129.0,
      "step": 44060
    },
    {
      "entropy": 6.014926195144653,
      "epoch": 5.085401038661281,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002730275892452836,
      "loss": 5.3801,
      "mean_token_accuracy": 0.22070738226175307,
      "num_tokens": 111689546.0,
      "step": 44065
    },
    {
      "entropy": 6.076750898361206,
      "epoch": 5.08597807270629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002729863265124046,
      "loss": 5.4684,
      "mean_token_accuracy": 0.22009487748146056,
      "num_tokens": 111702015.0,
      "step": 44070
    },
    {
      "entropy": 6.1154650211334225,
      "epoch": 5.086555106751298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002729450638472545,
      "loss": 5.5007,
      "mean_token_accuracy": 0.21336942613124849,
      "num_tokens": 111713559.0,
      "step": 44075
    },
    {
      "entropy": 6.013238716125488,
      "epoch": 5.087132140796307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002729038012512211,
      "loss": 5.4505,
      "mean_token_accuracy": 0.21332897394895553,
      "num_tokens": 111725249.0,
      "step": 44080
    },
    {
      "entropy": 6.0834942817687985,
      "epoch": 5.087709174841316,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002728625387256924,
      "loss": 5.4971,
      "mean_token_accuracy": 0.21392269134521485,
      "num_tokens": 111736978.0,
      "step": 44085
    },
    {
      "entropy": 6.066761207580567,
      "epoch": 5.088286208886324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027282127627205604,
      "loss": 5.4875,
      "mean_token_accuracy": 0.21301310658454894,
      "num_tokens": 111749129.0,
      "step": 44090
    },
    {
      "entropy": 6.019028425216675,
      "epoch": 5.088863242931333,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002727800138917001,
      "loss": 5.4452,
      "mean_token_accuracy": 0.22329931259155272,
      "num_tokens": 111762348.0,
      "step": 44095
    },
    {
      "entropy": 6.001850032806397,
      "epoch": 5.089440276976342,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027273875158601214,
      "loss": 5.4377,
      "mean_token_accuracy": 0.21423040330410004,
      "num_tokens": 111775847.0,
      "step": 44100
    },
    {
      "entropy": 6.078058576583862,
      "epoch": 5.0900173110213505,
      "grad_norm": 0.953125,
      "learning_rate": 0.00027269748935638026,
      "loss": 5.4679,
      "mean_token_accuracy": 0.22060146927833557,
      "num_tokens": 111789175.0,
      "step": 44105
    },
    {
      "entropy": 6.084034967422485,
      "epoch": 5.090594345066359,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002726562272041921,
      "loss": 5.466,
      "mean_token_accuracy": 0.21976256221532822,
      "num_tokens": 111801294.0,
      "step": 44110
    },
    {
      "entropy": 6.018796586990357,
      "epoch": 5.091171379111367,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027261496513083567,
      "loss": 5.3825,
      "mean_token_accuracy": 0.22294479012489318,
      "num_tokens": 111814006.0,
      "step": 44115
    },
    {
      "entropy": 5.953041124343872,
      "epoch": 5.091748413156377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027257370313769864,
      "loss": 5.2922,
      "mean_token_accuracy": 0.23157981038093567,
      "num_tokens": 111825714.0,
      "step": 44120
    },
    {
      "entropy": 6.021838808059693,
      "epoch": 5.092325447201385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000272532441226169,
      "loss": 5.5264,
      "mean_token_accuracy": 0.21223487854003906,
      "num_tokens": 111838264.0,
      "step": 44125
    },
    {
      "entropy": 6.104084587097168,
      "epoch": 5.092902481246393,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027249117939763437,
      "loss": 5.4681,
      "mean_token_accuracy": 0.2156369060277939,
      "num_tokens": 111850281.0,
      "step": 44130
    },
    {
      "entropy": 6.043744516372681,
      "epoch": 5.093479515291402,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027244991765348263,
      "loss": 5.4612,
      "mean_token_accuracy": 0.2171190470457077,
      "num_tokens": 111863512.0,
      "step": 44135
    },
    {
      "entropy": 5.979039287567138,
      "epoch": 5.094056549336411,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027240865599510174,
      "loss": 5.3363,
      "mean_token_accuracy": 0.2315061643719673,
      "num_tokens": 111875673.0,
      "step": 44140
    },
    {
      "entropy": 6.036716365814209,
      "epoch": 5.09463358338142,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00027236739442387944,
      "loss": 5.4812,
      "mean_token_accuracy": 0.21775710284709932,
      "num_tokens": 111889280.0,
      "step": 44145
    },
    {
      "entropy": 6.065521478652954,
      "epoch": 5.095210617426428,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027232613294120344,
      "loss": 5.4455,
      "mean_token_accuracy": 0.22125951945781708,
      "num_tokens": 111900895.0,
      "step": 44150
    },
    {
      "entropy": 6.0401082038879395,
      "epoch": 5.095787651471436,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002722848715484616,
      "loss": 5.4652,
      "mean_token_accuracy": 0.2165570318698883,
      "num_tokens": 111912874.0,
      "step": 44155
    },
    {
      "entropy": 5.986469411849976,
      "epoch": 5.096364685516446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027224361024704184,
      "loss": 5.3364,
      "mean_token_accuracy": 0.23090906739234923,
      "num_tokens": 111925787.0,
      "step": 44160
    },
    {
      "entropy": 6.135260963439942,
      "epoch": 5.096941719561454,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002722023490383318,
      "loss": 5.5271,
      "mean_token_accuracy": 0.21275327801704408,
      "num_tokens": 111937150.0,
      "step": 44165
    },
    {
      "entropy": 6.005226278305054,
      "epoch": 5.097518753606463,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027216108792371945,
      "loss": 5.4413,
      "mean_token_accuracy": 0.2194020479917526,
      "num_tokens": 111949570.0,
      "step": 44170
    },
    {
      "entropy": 6.039155673980713,
      "epoch": 5.098095787651472,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00027211982690459246,
      "loss": 5.5213,
      "mean_token_accuracy": 0.2094275176525116,
      "num_tokens": 111962220.0,
      "step": 44175
    },
    {
      "entropy": 6.079327917098999,
      "epoch": 5.09867282169648,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027207856598233867,
      "loss": 5.4439,
      "mean_token_accuracy": 0.21620655357837676,
      "num_tokens": 111975098.0,
      "step": 44180
    },
    {
      "entropy": 6.100356912612915,
      "epoch": 5.099249855741489,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027203730515834584,
      "loss": 5.4751,
      "mean_token_accuracy": 0.22408611178398133,
      "num_tokens": 111987629.0,
      "step": 44185
    },
    {
      "entropy": 6.056990146636963,
      "epoch": 5.099826889786497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002719960444340018,
      "loss": 5.4472,
      "mean_token_accuracy": 0.21904900521039963,
      "num_tokens": 112000416.0,
      "step": 44190
    },
    {
      "entropy": 6.029079914093018,
      "epoch": 5.100403923831506,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00027195478381069433,
      "loss": 5.4277,
      "mean_token_accuracy": 0.2184651955962181,
      "num_tokens": 112012531.0,
      "step": 44195
    },
    {
      "entropy": 5.997275733947754,
      "epoch": 5.100980957876515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027191352328981115,
      "loss": 5.3848,
      "mean_token_accuracy": 0.22130489200353623,
      "num_tokens": 112023662.0,
      "step": 44200
    },
    {
      "entropy": 6.000210189819336,
      "epoch": 5.101557991921523,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027187226287274006,
      "loss": 5.4165,
      "mean_token_accuracy": 0.21802253127098084,
      "num_tokens": 112036135.0,
      "step": 44205
    },
    {
      "entropy": 6.09615569114685,
      "epoch": 5.102135025966532,
      "grad_norm": 1.0,
      "learning_rate": 0.000271831002560869,
      "loss": 5.5001,
      "mean_token_accuracy": 0.21643033772706985,
      "num_tokens": 112050234.0,
      "step": 44210
    },
    {
      "entropy": 6.010221672058106,
      "epoch": 5.102712060011541,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027178974235558546,
      "loss": 5.341,
      "mean_token_accuracy": 0.22517186552286148,
      "num_tokens": 112063860.0,
      "step": 44215
    },
    {
      "entropy": 5.995897388458252,
      "epoch": 5.103289094056549,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027174848225827735,
      "loss": 5.4175,
      "mean_token_accuracy": 0.21974959522485732,
      "num_tokens": 112076381.0,
      "step": 44220
    },
    {
      "entropy": 5.970018625259399,
      "epoch": 5.103866128101558,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027170722227033245,
      "loss": 5.3887,
      "mean_token_accuracy": 0.22539203613996506,
      "num_tokens": 112088828.0,
      "step": 44225
    },
    {
      "entropy": 6.031030511856079,
      "epoch": 5.104443162146566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002716659623931386,
      "loss": 5.4167,
      "mean_token_accuracy": 0.22228151559829712,
      "num_tokens": 112102488.0,
      "step": 44230
    },
    {
      "entropy": 6.0417201042175295,
      "epoch": 5.1050201961915755,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027162470262808325,
      "loss": 5.3842,
      "mean_token_accuracy": 0.22782442569732667,
      "num_tokens": 112115372.0,
      "step": 44235
    },
    {
      "entropy": 6.065902519226074,
      "epoch": 5.105597230236584,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027158344297655446,
      "loss": 5.4537,
      "mean_token_accuracy": 0.21784008890390397,
      "num_tokens": 112128660.0,
      "step": 44240
    },
    {
      "entropy": 6.010843944549561,
      "epoch": 5.106174264281592,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027154218343993983,
      "loss": 5.4105,
      "mean_token_accuracy": 0.2173339068889618,
      "num_tokens": 112141122.0,
      "step": 44245
    },
    {
      "entropy": 6.104519653320312,
      "epoch": 5.106751298326602,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027150092401962724,
      "loss": 5.509,
      "mean_token_accuracy": 0.21041673868894578,
      "num_tokens": 112152992.0,
      "step": 44250
    },
    {
      "entropy": 5.994868278503418,
      "epoch": 5.10732833237161,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002714596647170042,
      "loss": 5.3521,
      "mean_token_accuracy": 0.23054351806640624,
      "num_tokens": 112166817.0,
      "step": 44255
    },
    {
      "entropy": 6.034130716323853,
      "epoch": 5.1079053664166185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027141840553345864,
      "loss": 5.4247,
      "mean_token_accuracy": 0.22292573004961014,
      "num_tokens": 112179934.0,
      "step": 44260
    },
    {
      "entropy": 6.0555338859558105,
      "epoch": 5.108482400461627,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002713771464703782,
      "loss": 5.4361,
      "mean_token_accuracy": 0.2198113650083542,
      "num_tokens": 112192916.0,
      "step": 44265
    },
    {
      "entropy": 6.013733434677124,
      "epoch": 5.109059434506636,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027133588752915075,
      "loss": 5.3808,
      "mean_token_accuracy": 0.22532901167869568,
      "num_tokens": 112205618.0,
      "step": 44270
    },
    {
      "entropy": 6.11332950592041,
      "epoch": 5.109636468551645,
      "grad_norm": 1.0,
      "learning_rate": 0.00027129462871116375,
      "loss": 5.5319,
      "mean_token_accuracy": 0.207612980902195,
      "num_tokens": 112217337.0,
      "step": 44275
    },
    {
      "entropy": 6.0811244487762455,
      "epoch": 5.110213502596653,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027125337001780513,
      "loss": 5.4945,
      "mean_token_accuracy": 0.21728063374757767,
      "num_tokens": 112229281.0,
      "step": 44280
    },
    {
      "entropy": 6.079291915893554,
      "epoch": 5.1107905366416615,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002712121114504625,
      "loss": 5.4784,
      "mean_token_accuracy": 0.21370008140802382,
      "num_tokens": 112241762.0,
      "step": 44285
    },
    {
      "entropy": 5.9831983089447025,
      "epoch": 5.111367570686671,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00027117085301052375,
      "loss": 5.3679,
      "mean_token_accuracy": 0.22413185387849807,
      "num_tokens": 112254457.0,
      "step": 44290
    },
    {
      "entropy": 5.995687818527221,
      "epoch": 5.111944604731679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002711295946993764,
      "loss": 5.3478,
      "mean_token_accuracy": 0.22914140671491623,
      "num_tokens": 112267961.0,
      "step": 44295
    },
    {
      "entropy": 6.033307790756226,
      "epoch": 5.112521638776688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002710883365184082,
      "loss": 5.4283,
      "mean_token_accuracy": 0.22290459126234055,
      "num_tokens": 112281107.0,
      "step": 44300
    },
    {
      "entropy": 6.054268741607666,
      "epoch": 5.113098672821696,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002710470784690068,
      "loss": 5.4281,
      "mean_token_accuracy": 0.2260906368494034,
      "num_tokens": 112294141.0,
      "step": 44305
    },
    {
      "entropy": 5.956476736068725,
      "epoch": 5.113675706866705,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002710058205525601,
      "loss": 5.4063,
      "mean_token_accuracy": 0.22675709128379823,
      "num_tokens": 112308105.0,
      "step": 44310
    },
    {
      "entropy": 5.948278856277466,
      "epoch": 5.114252740911714,
      "grad_norm": 1.0,
      "learning_rate": 0.00027096456277045546,
      "loss": 5.3837,
      "mean_token_accuracy": 0.23046501725912094,
      "num_tokens": 112321997.0,
      "step": 44315
    },
    {
      "entropy": 6.042380952835083,
      "epoch": 5.114829774956722,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00027092330512408085,
      "loss": 5.4116,
      "mean_token_accuracy": 0.22785115242004395,
      "num_tokens": 112335495.0,
      "step": 44320
    },
    {
      "entropy": 5.898266696929932,
      "epoch": 5.1154068090017315,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002708820476148238,
      "loss": 5.2687,
      "mean_token_accuracy": 0.23632698953151704,
      "num_tokens": 112350053.0,
      "step": 44325
    },
    {
      "entropy": 6.028501081466675,
      "epoch": 5.11598384304674,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002708407902440721,
      "loss": 5.4573,
      "mean_token_accuracy": 0.22067195177078247,
      "num_tokens": 112362197.0,
      "step": 44330
    },
    {
      "entropy": 5.971458530426025,
      "epoch": 5.116560877091748,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002707995330132133,
      "loss": 5.3196,
      "mean_token_accuracy": 0.23021223694086074,
      "num_tokens": 112374416.0,
      "step": 44335
    },
    {
      "entropy": 5.887787961959839,
      "epoch": 5.117137911136757,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002707582759236352,
      "loss": 5.275,
      "mean_token_accuracy": 0.23801162838935852,
      "num_tokens": 112387748.0,
      "step": 44340
    },
    {
      "entropy": 6.059664678573609,
      "epoch": 5.117714945181766,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027071701897672526,
      "loss": 5.3827,
      "mean_token_accuracy": 0.22549789100885392,
      "num_tokens": 112398878.0,
      "step": 44345
    },
    {
      "entropy": 6.133424520492554,
      "epoch": 5.1182919792267745,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00027067576217387133,
      "loss": 5.5131,
      "mean_token_accuracy": 0.21598127037286757,
      "num_tokens": 112410877.0,
      "step": 44350
    },
    {
      "entropy": 6.046455717086792,
      "epoch": 5.118869013271783,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027063450551646095,
      "loss": 5.4378,
      "mean_token_accuracy": 0.221391561627388,
      "num_tokens": 112422464.0,
      "step": 44355
    },
    {
      "entropy": 6.0207763671875,
      "epoch": 5.119446047316791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027059324900588185,
      "loss": 5.3938,
      "mean_token_accuracy": 0.22808483093976975,
      "num_tokens": 112434517.0,
      "step": 44360
    },
    {
      "entropy": 5.987870836257935,
      "epoch": 5.120023081361801,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002705519926435216,
      "loss": 5.4184,
      "mean_token_accuracy": 0.2249480426311493,
      "num_tokens": 112446568.0,
      "step": 44365
    },
    {
      "entropy": 6.004012250900269,
      "epoch": 5.120600115406809,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00027051073643076794,
      "loss": 5.3387,
      "mean_token_accuracy": 0.2288269206881523,
      "num_tokens": 112459472.0,
      "step": 44370
    },
    {
      "entropy": 6.024369525909424,
      "epoch": 5.1211771494518175,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00027046948036900837,
      "loss": 5.4111,
      "mean_token_accuracy": 0.21618709564208985,
      "num_tokens": 112472612.0,
      "step": 44375
    },
    {
      "entropy": 6.026199865341186,
      "epoch": 5.121754183496826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002704282244596306,
      "loss": 5.4618,
      "mean_token_accuracy": 0.22158183157444,
      "num_tokens": 112485299.0,
      "step": 44380
    },
    {
      "entropy": 6.083711814880371,
      "epoch": 5.122331217541835,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027038696870402217,
      "loss": 5.489,
      "mean_token_accuracy": 0.21162751764059068,
      "num_tokens": 112497870.0,
      "step": 44385
    },
    {
      "entropy": 5.962233543395996,
      "epoch": 5.122908251586844,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002703457131035709,
      "loss": 5.2949,
      "mean_token_accuracy": 0.22972021102905274,
      "num_tokens": 112510596.0,
      "step": 44390
    },
    {
      "entropy": 5.895632362365722,
      "epoch": 5.123485285631852,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027030445765966404,
      "loss": 5.2866,
      "mean_token_accuracy": 0.22984685003757477,
      "num_tokens": 112524421.0,
      "step": 44395
    },
    {
      "entropy": 6.001127767562866,
      "epoch": 5.124062319676861,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027026320237368955,
      "loss": 5.3675,
      "mean_token_accuracy": 0.2265291541814804,
      "num_tokens": 112536033.0,
      "step": 44400
    },
    {
      "entropy": 6.075658941268921,
      "epoch": 5.12463935372187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027022194724703484,
      "loss": 5.4754,
      "mean_token_accuracy": 0.2165545180439949,
      "num_tokens": 112549095.0,
      "step": 44405
    },
    {
      "entropy": 6.017171859741211,
      "epoch": 5.125216387766878,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002701806922810877,
      "loss": 5.4197,
      "mean_token_accuracy": 0.2216867044568062,
      "num_tokens": 112560939.0,
      "step": 44410
    },
    {
      "entropy": 6.074033641815186,
      "epoch": 5.125793421811887,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002701394374772354,
      "loss": 5.5165,
      "mean_token_accuracy": 0.218694768846035,
      "num_tokens": 112574602.0,
      "step": 44415
    },
    {
      "entropy": 6.1506813049316404,
      "epoch": 5.126370455856896,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002700981828368659,
      "loss": 5.4736,
      "mean_token_accuracy": 0.21896029114723206,
      "num_tokens": 112587227.0,
      "step": 44420
    },
    {
      "entropy": 6.021144580841065,
      "epoch": 5.126947489901904,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027005692836136637,
      "loss": 5.372,
      "mean_token_accuracy": 0.2248900443315506,
      "num_tokens": 112600624.0,
      "step": 44425
    },
    {
      "entropy": 5.963524866104126,
      "epoch": 5.127524523946913,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002700156740521248,
      "loss": 5.3836,
      "mean_token_accuracy": 0.22393372803926467,
      "num_tokens": 112612643.0,
      "step": 44430
    },
    {
      "entropy": 5.993826484680175,
      "epoch": 5.128101557991921,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002699744199105284,
      "loss": 5.448,
      "mean_token_accuracy": 0.2290252134203911,
      "num_tokens": 112625282.0,
      "step": 44435
    },
    {
      "entropy": 5.933115530014038,
      "epoch": 5.12867859203693,
      "grad_norm": 1.046875,
      "learning_rate": 0.000269933165937965,
      "loss": 5.3302,
      "mean_token_accuracy": 0.219595205783844,
      "num_tokens": 112636880.0,
      "step": 44440
    },
    {
      "entropy": 6.063958692550659,
      "epoch": 5.129255626081939,
      "grad_norm": 0.984375,
      "learning_rate": 0.000269891912135822,
      "loss": 5.4424,
      "mean_token_accuracy": 0.21860643923282624,
      "num_tokens": 112649381.0,
      "step": 44445
    },
    {
      "entropy": 6.005603313446045,
      "epoch": 5.129832660126947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000269850658505487,
      "loss": 5.4176,
      "mean_token_accuracy": 0.21815809458494187,
      "num_tokens": 112660787.0,
      "step": 44450
    },
    {
      "entropy": 6.045212125778198,
      "epoch": 5.130409694171956,
      "grad_norm": 0.921875,
      "learning_rate": 0.00026980940504834743,
      "loss": 5.4979,
      "mean_token_accuracy": 0.21226125508546828,
      "num_tokens": 112672777.0,
      "step": 44455
    },
    {
      "entropy": 5.912465381622314,
      "epoch": 5.130986728216965,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026976815176579116,
      "loss": 5.278,
      "mean_token_accuracy": 0.23653078079223633,
      "num_tokens": 112685894.0,
      "step": 44460
    },
    {
      "entropy": 6.073809814453125,
      "epoch": 5.131563762261973,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026972689865920534,
      "loss": 5.4485,
      "mean_token_accuracy": 0.21514449417591094,
      "num_tokens": 112698627.0,
      "step": 44465
    },
    {
      "entropy": 6.055767965316773,
      "epoch": 5.132140796306982,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002696856457299777,
      "loss": 5.4398,
      "mean_token_accuracy": 0.2137020170688629,
      "num_tokens": 112709642.0,
      "step": 44470
    },
    {
      "entropy": 5.933177280426025,
      "epoch": 5.132717830351991,
      "grad_norm": 1.0,
      "learning_rate": 0.0002696443929794957,
      "loss": 5.3004,
      "mean_token_accuracy": 0.23580056428909302,
      "num_tokens": 112723488.0,
      "step": 44475
    },
    {
      "entropy": 6.0547089099884035,
      "epoch": 5.133294864397,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000269603140409147,
      "loss": 5.5068,
      "mean_token_accuracy": 0.2118276223540306,
      "num_tokens": 112735320.0,
      "step": 44480
    },
    {
      "entropy": 6.050627851486206,
      "epoch": 5.133871898442008,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026956188802031893,
      "loss": 5.3951,
      "mean_token_accuracy": 0.22158414870500565,
      "num_tokens": 112747921.0,
      "step": 44485
    },
    {
      "entropy": 6.054494428634643,
      "epoch": 5.134448932487016,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00026952063581439905,
      "loss": 5.4695,
      "mean_token_accuracy": 0.22435725182294847,
      "num_tokens": 112760123.0,
      "step": 44490
    },
    {
      "entropy": 6.090556240081787,
      "epoch": 5.135025966532026,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026947938379277484,
      "loss": 5.499,
      "mean_token_accuracy": 0.2141335979104042,
      "num_tokens": 112773407.0,
      "step": 44495
    },
    {
      "entropy": 5.854138326644898,
      "epoch": 5.135603000577034,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026943813195683385,
      "loss": 5.2329,
      "mean_token_accuracy": 0.24058865010738373,
      "num_tokens": 112787594.0,
      "step": 44500
    },
    {
      "entropy": 5.976094913482666,
      "epoch": 5.1361800346220425,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002693968803079635,
      "loss": 5.3712,
      "mean_token_accuracy": 0.2286993995308876,
      "num_tokens": 112800766.0,
      "step": 44505
    },
    {
      "entropy": 5.990269804000855,
      "epoch": 5.136757068667051,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026935562884755133,
      "loss": 5.3783,
      "mean_token_accuracy": 0.22347110360860825,
      "num_tokens": 112814072.0,
      "step": 44510
    },
    {
      "entropy": 6.128317356109619,
      "epoch": 5.13733410271206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026931437757698475,
      "loss": 5.5226,
      "mean_token_accuracy": 0.21437612622976304,
      "num_tokens": 112825821.0,
      "step": 44515
    },
    {
      "entropy": 6.021499729156494,
      "epoch": 5.137911136757069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026927312649765127,
      "loss": 5.3849,
      "mean_token_accuracy": 0.22333071380853653,
      "num_tokens": 112837681.0,
      "step": 44520
    },
    {
      "entropy": 6.010371255874634,
      "epoch": 5.138488170802077,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026923187561093834,
      "loss": 5.4116,
      "mean_token_accuracy": 0.22103540748357772,
      "num_tokens": 112851476.0,
      "step": 44525
    },
    {
      "entropy": 6.081694412231445,
      "epoch": 5.139065204847086,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00026919062491823336,
      "loss": 5.5225,
      "mean_token_accuracy": 0.21470822393894196,
      "num_tokens": 112863722.0,
      "step": 44530
    },
    {
      "entropy": 6.045423269271851,
      "epoch": 5.139642238892095,
      "grad_norm": 0.9375,
      "learning_rate": 0.00026914937442092384,
      "loss": 5.3447,
      "mean_token_accuracy": 0.22357507795095444,
      "num_tokens": 112876399.0,
      "step": 44535
    },
    {
      "entropy": 6.038230133056641,
      "epoch": 5.140219272937103,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026910812412039734,
      "loss": 5.4592,
      "mean_token_accuracy": 0.22503597140312195,
      "num_tokens": 112889765.0,
      "step": 44540
    },
    {
      "entropy": 5.99134407043457,
      "epoch": 5.140796306982112,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000269066874018041,
      "loss": 5.3788,
      "mean_token_accuracy": 0.2356344297528267,
      "num_tokens": 112902136.0,
      "step": 44545
    },
    {
      "entropy": 5.959253358840942,
      "epoch": 5.141373341027121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002690256241152424,
      "loss": 5.3359,
      "mean_token_accuracy": 0.2273300901055336,
      "num_tokens": 112914324.0,
      "step": 44550
    },
    {
      "entropy": 5.926967668533325,
      "epoch": 5.141950375072129,
      "grad_norm": 1.21875,
      "learning_rate": 0.000268984374413389,
      "loss": 5.291,
      "mean_token_accuracy": 0.23139042407274246,
      "num_tokens": 112927732.0,
      "step": 44555
    },
    {
      "entropy": 6.006013488769531,
      "epoch": 5.142527409117138,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002689431249138682,
      "loss": 5.4005,
      "mean_token_accuracy": 0.22160426378250123,
      "num_tokens": 112940765.0,
      "step": 44560
    },
    {
      "entropy": 6.005873775482177,
      "epoch": 5.143104443162146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026890187561806747,
      "loss": 5.3421,
      "mean_token_accuracy": 0.22981645464897155,
      "num_tokens": 112954162.0,
      "step": 44565
    },
    {
      "entropy": 6.0560156345367435,
      "epoch": 5.1436814772071555,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000268860626527374,
      "loss": 5.4152,
      "mean_token_accuracy": 0.22137005925178527,
      "num_tokens": 112968166.0,
      "step": 44570
    },
    {
      "entropy": 5.985534906387329,
      "epoch": 5.144258511252164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002688193776431753,
      "loss": 5.3472,
      "mean_token_accuracy": 0.22974993139505387,
      "num_tokens": 112981246.0,
      "step": 44575
    },
    {
      "entropy": 6.000988245010376,
      "epoch": 5.144835545297172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002687781289668588,
      "loss": 5.3616,
      "mean_token_accuracy": 0.22493812292814255,
      "num_tokens": 112992828.0,
      "step": 44580
    },
    {
      "entropy": 6.131247282028198,
      "epoch": 5.145412579342181,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002687368804998119,
      "loss": 5.5341,
      "mean_token_accuracy": 0.21079110503196716,
      "num_tokens": 113005542.0,
      "step": 44585
    },
    {
      "entropy": 5.924923276901245,
      "epoch": 5.14598961338719,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002686956322434218,
      "loss": 5.2804,
      "mean_token_accuracy": 0.2421203076839447,
      "num_tokens": 113019656.0,
      "step": 44590
    },
    {
      "entropy": 5.983460855484009,
      "epoch": 5.1465666474321985,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026865438419907604,
      "loss": 5.3875,
      "mean_token_accuracy": 0.2214603915810585,
      "num_tokens": 113032007.0,
      "step": 44595
    },
    {
      "entropy": 6.03508620262146,
      "epoch": 5.147143681477207,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026861313636816195,
      "loss": 5.4286,
      "mean_token_accuracy": 0.22106152772903442,
      "num_tokens": 113044305.0,
      "step": 44600
    },
    {
      "entropy": 6.095164775848389,
      "epoch": 5.147720715522216,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026857188875206684,
      "loss": 5.5463,
      "mean_token_accuracy": 0.21622940748929978,
      "num_tokens": 113056412.0,
      "step": 44605
    },
    {
      "entropy": 6.004116106033325,
      "epoch": 5.148297749567225,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026853064135217797,
      "loss": 5.4397,
      "mean_token_accuracy": 0.21877771764993667,
      "num_tokens": 113068793.0,
      "step": 44610
    },
    {
      "entropy": 6.031514358520508,
      "epoch": 5.148874783612233,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026848939416988283,
      "loss": 5.4002,
      "mean_token_accuracy": 0.22823102176189422,
      "num_tokens": 113080869.0,
      "step": 44615
    },
    {
      "entropy": 5.997077178955078,
      "epoch": 5.1494518176572415,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002684481472065686,
      "loss": 5.3956,
      "mean_token_accuracy": 0.22227750718593597,
      "num_tokens": 113092764.0,
      "step": 44620
    },
    {
      "entropy": 5.971799850463867,
      "epoch": 5.150028851702251,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002684069004636229,
      "loss": 5.3942,
      "mean_token_accuracy": 0.22672227323055266,
      "num_tokens": 113105157.0,
      "step": 44625
    },
    {
      "entropy": 6.116018390655517,
      "epoch": 5.150605885747259,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002683656539424326,
      "loss": 5.4268,
      "mean_token_accuracy": 0.22036228477954864,
      "num_tokens": 113118600.0,
      "step": 44630
    },
    {
      "entropy": 6.110688591003418,
      "epoch": 5.151182919792268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002683244076443854,
      "loss": 5.503,
      "mean_token_accuracy": 0.2101044088602066,
      "num_tokens": 113131451.0,
      "step": 44635
    },
    {
      "entropy": 6.063616371154785,
      "epoch": 5.151759953837276,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002682831615708683,
      "loss": 5.4768,
      "mean_token_accuracy": 0.21707184612751007,
      "num_tokens": 113143676.0,
      "step": 44640
    },
    {
      "entropy": 6.077631187438965,
      "epoch": 5.152336987882285,
      "grad_norm": 1.0,
      "learning_rate": 0.0002682419157232688,
      "loss": 5.4585,
      "mean_token_accuracy": 0.2191653937101364,
      "num_tokens": 113156355.0,
      "step": 44645
    },
    {
      "entropy": 6.035669660568237,
      "epoch": 5.152914021927294,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026820067010297405,
      "loss": 5.435,
      "mean_token_accuracy": 0.22112427949905394,
      "num_tokens": 113170094.0,
      "step": 44650
    },
    {
      "entropy": 6.019174003601075,
      "epoch": 5.153491055972302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026815942471137145,
      "loss": 5.3489,
      "mean_token_accuracy": 0.22874856144189834,
      "num_tokens": 113182033.0,
      "step": 44655
    },
    {
      "entropy": 6.073612022399902,
      "epoch": 5.1540680900173115,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002681181795498481,
      "loss": 5.4862,
      "mean_token_accuracy": 0.21422306448221207,
      "num_tokens": 113194387.0,
      "step": 44660
    },
    {
      "entropy": 6.139192390441894,
      "epoch": 5.15464512406232,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026807693461979157,
      "loss": 5.5143,
      "mean_token_accuracy": 0.20934156328439713,
      "num_tokens": 113206754.0,
      "step": 44665
    },
    {
      "entropy": 6.045892953872681,
      "epoch": 5.155222158107328,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002680356899225887,
      "loss": 5.3808,
      "mean_token_accuracy": 0.2265696719288826,
      "num_tokens": 113218705.0,
      "step": 44670
    },
    {
      "entropy": 5.982802391052246,
      "epoch": 5.155799192152337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026799444545962705,
      "loss": 5.4217,
      "mean_token_accuracy": 0.22309057861566545,
      "num_tokens": 113231089.0,
      "step": 44675
    },
    {
      "entropy": 6.0090666770935055,
      "epoch": 5.156376226197346,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026795320123229366,
      "loss": 5.3326,
      "mean_token_accuracy": 0.22865242809057235,
      "num_tokens": 113243080.0,
      "step": 44680
    },
    {
      "entropy": 6.021140623092651,
      "epoch": 5.1569532602423545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002679119572419759,
      "loss": 5.3553,
      "mean_token_accuracy": 0.22254408597946168,
      "num_tokens": 113256273.0,
      "step": 44685
    },
    {
      "entropy": 5.988370132446289,
      "epoch": 5.157530294287363,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026787071349006095,
      "loss": 5.4323,
      "mean_token_accuracy": 0.2243028461933136,
      "num_tokens": 113268860.0,
      "step": 44690
    },
    {
      "entropy": 5.938693046569824,
      "epoch": 5.158107328332371,
      "grad_norm": 0.984375,
      "learning_rate": 0.000267829469977936,
      "loss": 5.3825,
      "mean_token_accuracy": 0.22711580097675324,
      "num_tokens": 113281216.0,
      "step": 44695
    },
    {
      "entropy": 6.030620813369751,
      "epoch": 5.158684362377381,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026778822670698823,
      "loss": 5.4024,
      "mean_token_accuracy": 0.21856730580329894,
      "num_tokens": 113294158.0,
      "step": 44700
    },
    {
      "entropy": 6.011349058151245,
      "epoch": 5.159261396422389,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026774698367860496,
      "loss": 5.4233,
      "mean_token_accuracy": 0.21693855077028273,
      "num_tokens": 113307424.0,
      "step": 44705
    },
    {
      "entropy": 5.949773693084717,
      "epoch": 5.159838430467397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002677057408941732,
      "loss": 5.3639,
      "mean_token_accuracy": 0.23049382269382476,
      "num_tokens": 113319331.0,
      "step": 44710
    },
    {
      "entropy": 6.027141046524048,
      "epoch": 5.160415464512406,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002676644983550802,
      "loss": 5.3948,
      "mean_token_accuracy": 0.2188984528183937,
      "num_tokens": 113331803.0,
      "step": 44715
    },
    {
      "entropy": 6.158699655532837,
      "epoch": 5.160992498557415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002676232560627132,
      "loss": 5.4916,
      "mean_token_accuracy": 0.21258330792188646,
      "num_tokens": 113343211.0,
      "step": 44720
    },
    {
      "entropy": 5.944718551635742,
      "epoch": 5.161569532602424,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026758201401845945,
      "loss": 5.3102,
      "mean_token_accuracy": 0.22187620997428895,
      "num_tokens": 113356503.0,
      "step": 44725
    },
    {
      "entropy": 6.034282588958741,
      "epoch": 5.162146566647432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026754077222370575,
      "loss": 5.4526,
      "mean_token_accuracy": 0.21784512400627137,
      "num_tokens": 113368274.0,
      "step": 44730
    },
    {
      "entropy": 5.991609239578247,
      "epoch": 5.162723600692441,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002674995306798396,
      "loss": 5.3517,
      "mean_token_accuracy": 0.2291390687227249,
      "num_tokens": 113382432.0,
      "step": 44735
    },
    {
      "entropy": 6.10571928024292,
      "epoch": 5.16330063473745,
      "grad_norm": 1.125,
      "learning_rate": 0.00026745828938824794,
      "loss": 5.4862,
      "mean_token_accuracy": 0.20711445659399033,
      "num_tokens": 113394358.0,
      "step": 44740
    },
    {
      "entropy": 6.102712631225586,
      "epoch": 5.163877668782458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026741704835031804,
      "loss": 5.5124,
      "mean_token_accuracy": 0.20864539444446564,
      "num_tokens": 113406785.0,
      "step": 44745
    },
    {
      "entropy": 6.035008192062378,
      "epoch": 5.1644547028274665,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002673758075674369,
      "loss": 5.3439,
      "mean_token_accuracy": 0.22151231169700622,
      "num_tokens": 113419815.0,
      "step": 44750
    },
    {
      "entropy": 5.990100765228272,
      "epoch": 5.165031736872476,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026733456704099173,
      "loss": 5.4019,
      "mean_token_accuracy": 0.2248249277472496,
      "num_tokens": 113431387.0,
      "step": 44755
    },
    {
      "entropy": 6.028573083877563,
      "epoch": 5.165608770917484,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026729332677236957,
      "loss": 5.4527,
      "mean_token_accuracy": 0.21799647212028503,
      "num_tokens": 113443690.0,
      "step": 44760
    },
    {
      "entropy": 6.070745992660522,
      "epoch": 5.166185804962493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002672520867629576,
      "loss": 5.4423,
      "mean_token_accuracy": 0.21718568354845047,
      "num_tokens": 113456695.0,
      "step": 44765
    },
    {
      "entropy": 5.85064058303833,
      "epoch": 5.166762839007501,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026721084701414273,
      "loss": 5.183,
      "mean_token_accuracy": 0.2460350960493088,
      "num_tokens": 113470521.0,
      "step": 44770
    },
    {
      "entropy": 5.933103847503662,
      "epoch": 5.16733987305251,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002671696075273122,
      "loss": 5.4274,
      "mean_token_accuracy": 0.22143907099962234,
      "num_tokens": 113484542.0,
      "step": 44775
    },
    {
      "entropy": 6.104251337051392,
      "epoch": 5.167916907097519,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000267128368303853,
      "loss": 5.5326,
      "mean_token_accuracy": 0.21054575890302657,
      "num_tokens": 113497962.0,
      "step": 44780
    },
    {
      "entropy": 6.085057878494263,
      "epoch": 5.168493941142527,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026708712934515226,
      "loss": 5.4555,
      "mean_token_accuracy": 0.22229008078575135,
      "num_tokens": 113511684.0,
      "step": 44785
    },
    {
      "entropy": 6.15748085975647,
      "epoch": 5.169070975187536,
      "grad_norm": 0.9375,
      "learning_rate": 0.00026704589065259696,
      "loss": 5.586,
      "mean_token_accuracy": 0.20541545897722244,
      "num_tokens": 113525055.0,
      "step": 44790
    },
    {
      "entropy": 6.105576229095459,
      "epoch": 5.169648009232545,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002670046522275743,
      "loss": 5.521,
      "mean_token_accuracy": 0.2187458097934723,
      "num_tokens": 113537940.0,
      "step": 44795
    },
    {
      "entropy": 6.056947994232178,
      "epoch": 5.170225043277553,
      "grad_norm": 1.046875,
      "learning_rate": 0.000266963414071471,
      "loss": 5.4461,
      "mean_token_accuracy": 0.21187059432268143,
      "num_tokens": 113549955.0,
      "step": 44800
    },
    {
      "entropy": 5.942605924606323,
      "epoch": 5.170802077322562,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00026692217618567445,
      "loss": 5.392,
      "mean_token_accuracy": 0.21558916121721267,
      "num_tokens": 113563947.0,
      "step": 44805
    },
    {
      "entropy": 6.0644816875457765,
      "epoch": 5.171379111367571,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00026688093857157137,
      "loss": 5.4009,
      "mean_token_accuracy": 0.22395075410604476,
      "num_tokens": 113577240.0,
      "step": 44810
    },
    {
      "entropy": 6.079711771011352,
      "epoch": 5.1719561454125795,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026683970123054894,
      "loss": 5.4251,
      "mean_token_accuracy": 0.22018396258354186,
      "num_tokens": 113590575.0,
      "step": 44815
    },
    {
      "entropy": 6.048147583007813,
      "epoch": 5.172533179457588,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002667984641639941,
      "loss": 5.5083,
      "mean_token_accuracy": 0.20902777314186097,
      "num_tokens": 113603039.0,
      "step": 44820
    },
    {
      "entropy": 6.053910923004151,
      "epoch": 5.173110213502596,
      "grad_norm": 1.0,
      "learning_rate": 0.0002667572273732938,
      "loss": 5.4217,
      "mean_token_accuracy": 0.22254975885152817,
      "num_tokens": 113616736.0,
      "step": 44825
    },
    {
      "entropy": 6.10822491645813,
      "epoch": 5.173687247547606,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002667159908598351,
      "loss": 5.4066,
      "mean_token_accuracy": 0.2141740381717682,
      "num_tokens": 113628177.0,
      "step": 44830
    },
    {
      "entropy": 5.989495897293091,
      "epoch": 5.174264281592614,
      "grad_norm": 1.015625,
      "learning_rate": 0.000266674754625005,
      "loss": 5.3938,
      "mean_token_accuracy": 0.22373245358467103,
      "num_tokens": 113639792.0,
      "step": 44835
    },
    {
      "entropy": 5.931725120544433,
      "epoch": 5.1748413156376225,
      "grad_norm": 1.0,
      "learning_rate": 0.0002666335186701903,
      "loss": 5.3315,
      "mean_token_accuracy": 0.2324563205242157,
      "num_tokens": 113652390.0,
      "step": 44840
    },
    {
      "entropy": 5.983621740341187,
      "epoch": 5.175418349682631,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026659228299677804,
      "loss": 5.341,
      "mean_token_accuracy": 0.22976835668087006,
      "num_tokens": 113664608.0,
      "step": 44845
    },
    {
      "entropy": 5.988905382156372,
      "epoch": 5.17599538372764,
      "grad_norm": 1.328125,
      "learning_rate": 0.00026655104760615515,
      "loss": 5.3004,
      "mean_token_accuracy": 0.23790231049060823,
      "num_tokens": 113676787.0,
      "step": 44850
    },
    {
      "entropy": 6.0241286277771,
      "epoch": 5.176572417772649,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026650981249970874,
      "loss": 5.3972,
      "mean_token_accuracy": 0.22648463100194932,
      "num_tokens": 113689022.0,
      "step": 44855
    },
    {
      "entropy": 5.948133754730224,
      "epoch": 5.177149451817657,
      "grad_norm": 0.953125,
      "learning_rate": 0.00026646857767882536,
      "loss": 5.4182,
      "mean_token_accuracy": 0.21624010056257248,
      "num_tokens": 113702744.0,
      "step": 44860
    },
    {
      "entropy": 5.973331546783447,
      "epoch": 5.1777264858626655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026642734314489224,
      "loss": 5.3599,
      "mean_token_accuracy": 0.23160626888275146,
      "num_tokens": 113715318.0,
      "step": 44865
    },
    {
      "entropy": 5.9750227451324465,
      "epoch": 5.178303519907675,
      "grad_norm": 1.0,
      "learning_rate": 0.0002663861088992962,
      "loss": 5.3173,
      "mean_token_accuracy": 0.2294831618666649,
      "num_tokens": 113728547.0,
      "step": 44870
    },
    {
      "entropy": 5.96131067276001,
      "epoch": 5.178880553952683,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002663448749434241,
      "loss": 5.3801,
      "mean_token_accuracy": 0.23026080429553986,
      "num_tokens": 113741188.0,
      "step": 44875
    },
    {
      "entropy": 5.997134399414063,
      "epoch": 5.179457587997692,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026630364127866273,
      "loss": 5.3637,
      "mean_token_accuracy": 0.22123554944992066,
      "num_tokens": 113753789.0,
      "step": 44880
    },
    {
      "entropy": 6.0182726860046385,
      "epoch": 5.180034622042701,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026626240790639916,
      "loss": 5.4342,
      "mean_token_accuracy": 0.22021446377038956,
      "num_tokens": 113767437.0,
      "step": 44885
    },
    {
      "entropy": 6.044240856170655,
      "epoch": 5.180611656087709,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002662211748280201,
      "loss": 5.4451,
      "mean_token_accuracy": 0.22081341296434404,
      "num_tokens": 113779768.0,
      "step": 44890
    },
    {
      "entropy": 6.061866664886475,
      "epoch": 5.181188690132718,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002661799420449126,
      "loss": 5.3919,
      "mean_token_accuracy": 0.22339577227830887,
      "num_tokens": 113792050.0,
      "step": 44895
    },
    {
      "entropy": 6.027600717544556,
      "epoch": 5.181765724177726,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002661387095584632,
      "loss": 5.4084,
      "mean_token_accuracy": 0.22313836514949797,
      "num_tokens": 113803828.0,
      "step": 44900
    },
    {
      "entropy": 6.0872053623199465,
      "epoch": 5.1823427582227355,
      "grad_norm": 1.046875,
      "learning_rate": 0.000266097477370059,
      "loss": 5.477,
      "mean_token_accuracy": 0.22224216163158417,
      "num_tokens": 113815807.0,
      "step": 44905
    },
    {
      "entropy": 6.047957468032837,
      "epoch": 5.182919792267744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026605624548108666,
      "loss": 5.4785,
      "mean_token_accuracy": 0.21966487765312195,
      "num_tokens": 113828608.0,
      "step": 44910
    },
    {
      "entropy": 6.124171400070191,
      "epoch": 5.183496826312752,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026601501389293317,
      "loss": 5.5078,
      "mean_token_accuracy": 0.20982252210378646,
      "num_tokens": 113839838.0,
      "step": 44915
    },
    {
      "entropy": 6.061716747283936,
      "epoch": 5.184073860357761,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026597378260698505,
      "loss": 5.4486,
      "mean_token_accuracy": 0.22021180093288423,
      "num_tokens": 113852233.0,
      "step": 44920
    },
    {
      "entropy": 6.138302183151245,
      "epoch": 5.18465089440277,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002659325516246294,
      "loss": 5.5179,
      "mean_token_accuracy": 0.2113403469324112,
      "num_tokens": 113864263.0,
      "step": 44925
    },
    {
      "entropy": 6.086659097671509,
      "epoch": 5.1852279284477785,
      "grad_norm": 1.0,
      "learning_rate": 0.0002658913209472528,
      "loss": 5.4711,
      "mean_token_accuracy": 0.21501243710517884,
      "num_tokens": 113877400.0,
      "step": 44930
    },
    {
      "entropy": 5.988292026519775,
      "epoch": 5.185804962492787,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002658500905762422,
      "loss": 5.3618,
      "mean_token_accuracy": 0.22192232459783554,
      "num_tokens": 113888797.0,
      "step": 44935
    },
    {
      "entropy": 5.968055295944214,
      "epoch": 5.186381996537795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026580886051298414,
      "loss": 5.3519,
      "mean_token_accuracy": 0.23253681212663652,
      "num_tokens": 113899931.0,
      "step": 44940
    },
    {
      "entropy": 6.133341598510742,
      "epoch": 5.186959030582805,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002657676307588656,
      "loss": 5.47,
      "mean_token_accuracy": 0.21322982162237167,
      "num_tokens": 113912482.0,
      "step": 44945
    },
    {
      "entropy": 6.11820912361145,
      "epoch": 5.187536064627813,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026572640131527315,
      "loss": 5.5278,
      "mean_token_accuracy": 0.21246123164892197,
      "num_tokens": 113925345.0,
      "step": 44950
    },
    {
      "entropy": 6.044606351852417,
      "epoch": 5.188113098672821,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002656851721835935,
      "loss": 5.4435,
      "mean_token_accuracy": 0.22434695512056352,
      "num_tokens": 113938800.0,
      "step": 44955
    },
    {
      "entropy": 6.106613445281982,
      "epoch": 5.188690132717831,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002656439433652136,
      "loss": 5.4949,
      "mean_token_accuracy": 0.21674506962299347,
      "num_tokens": 113951408.0,
      "step": 44960
    },
    {
      "entropy": 6.141059398651123,
      "epoch": 5.189267166762839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026560271486152,
      "loss": 5.5041,
      "mean_token_accuracy": 0.21088642328977586,
      "num_tokens": 113963593.0,
      "step": 44965
    },
    {
      "entropy": 6.05765290260315,
      "epoch": 5.189844200807848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00026556148667389936,
      "loss": 5.4583,
      "mean_token_accuracy": 0.21746560037136078,
      "num_tokens": 113975777.0,
      "step": 44970
    },
    {
      "entropy": 6.006952571868896,
      "epoch": 5.190421234852856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002655202588037384,
      "loss": 5.4302,
      "mean_token_accuracy": 0.21962960809469223,
      "num_tokens": 113988840.0,
      "step": 44975
    },
    {
      "entropy": 5.984322500228882,
      "epoch": 5.190998268897865,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00026547903125242393,
      "loss": 5.3413,
      "mean_token_accuracy": 0.22475039064884186,
      "num_tokens": 114001905.0,
      "step": 44980
    },
    {
      "entropy": 5.982054615020752,
      "epoch": 5.191575302942874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002654378040213425,
      "loss": 5.2853,
      "mean_token_accuracy": 0.23152008652687073,
      "num_tokens": 114014290.0,
      "step": 44985
    },
    {
      "entropy": 6.027858543395996,
      "epoch": 5.192152336987882,
      "grad_norm": 1.0,
      "learning_rate": 0.00026539657711188074,
      "loss": 5.4463,
      "mean_token_accuracy": 0.21847382932901382,
      "num_tokens": 114026105.0,
      "step": 44990
    },
    {
      "entropy": 5.993068695068359,
      "epoch": 5.192729371032891,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002653553505254253,
      "loss": 5.47,
      "mean_token_accuracy": 0.21498792469501496,
      "num_tokens": 114038643.0,
      "step": 44995
    },
    {
      "entropy": 5.895764303207398,
      "epoch": 5.1933064050779,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00026531412426336295,
      "loss": 5.3549,
      "mean_token_accuracy": 0.22395807206630708,
      "num_tokens": 114050253.0,
      "step": 45000
    },
    {
      "epoch": 5.1933064050779,
      "eval_entropy": 5.893099537823123,
      "eval_loss": 5.628840923309326,
      "eval_mean_token_accuracy": 0.21704788898367552,
      "eval_num_tokens": 114050253.0,
      "eval_runtime": 23.0197,
      "eval_samples_per_second": 1222.259,
      "eval_steps_per_second": 152.782,
      "step": 45000
    },
    {
      "entropy": 6.085718441009521,
      "epoch": 5.193883439122908,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026527289832708014,
      "loss": 5.4535,
      "mean_token_accuracy": 0.21541539281606675,
      "num_tokens": 114062469.0,
      "step": 45005
    },
    {
      "entropy": 5.977302742004395,
      "epoch": 5.194460473167917,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00026523167271796363,
      "loss": 5.2835,
      "mean_token_accuracy": 0.22833088040351868,
      "num_tokens": 114074093.0,
      "step": 45010
    },
    {
      "entropy": 6.058987808227539,
      "epoch": 5.195037507212925,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026519044743739985,
      "loss": 5.4956,
      "mean_token_accuracy": 0.20970632135868073,
      "num_tokens": 114086419.0,
      "step": 45015
    },
    {
      "entropy": 6.06822772026062,
      "epoch": 5.195614541257934,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026514922248677564,
      "loss": 5.4515,
      "mean_token_accuracy": 0.21829158961772918,
      "num_tokens": 114098263.0,
      "step": 45020
    },
    {
      "entropy": 6.098209524154663,
      "epoch": 5.196191575302943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002651079978674773,
      "loss": 5.4619,
      "mean_token_accuracy": 0.21429959088563919,
      "num_tokens": 114109964.0,
      "step": 45025
    },
    {
      "entropy": 6.108616638183594,
      "epoch": 5.196768609347951,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026506677358089167,
      "loss": 5.5027,
      "mean_token_accuracy": 0.2161307543516159,
      "num_tokens": 114121157.0,
      "step": 45030
    },
    {
      "entropy": 6.059918212890625,
      "epoch": 5.197345643392961,
      "grad_norm": 1.015625,
      "learning_rate": 0.000265025549628405,
      "loss": 5.3968,
      "mean_token_accuracy": 0.22774211317300797,
      "num_tokens": 114133413.0,
      "step": 45035
    },
    {
      "entropy": 6.017913436889648,
      "epoch": 5.197922677437969,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002649843260114042,
      "loss": 5.4462,
      "mean_token_accuracy": 0.2191135674715042,
      "num_tokens": 114145293.0,
      "step": 45040
    },
    {
      "entropy": 5.992886972427368,
      "epoch": 5.198499711482977,
      "grad_norm": 1.0,
      "learning_rate": 0.00026494310273127554,
      "loss": 5.3548,
      "mean_token_accuracy": 0.2265837535262108,
      "num_tokens": 114157871.0,
      "step": 45045
    },
    {
      "entropy": 6.089379930496216,
      "epoch": 5.199076745527986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002649018797894056,
      "loss": 5.5359,
      "mean_token_accuracy": 0.2060634970664978,
      "num_tokens": 114170224.0,
      "step": 45050
    },
    {
      "entropy": 6.121094655990601,
      "epoch": 5.199653779572995,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026486065718718093,
      "loss": 5.4878,
      "mean_token_accuracy": 0.21787628680467605,
      "num_tokens": 114182180.0,
      "step": 45055
    },
    {
      "entropy": 6.116237735748291,
      "epoch": 5.2002308136180035,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002648194349259881,
      "loss": 5.5231,
      "mean_token_accuracy": 0.21248628348112106,
      "num_tokens": 114196207.0,
      "step": 45060
    },
    {
      "entropy": 6.0709761619567875,
      "epoch": 5.200807847663012,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002647782130072134,
      "loss": 5.4465,
      "mean_token_accuracy": 0.21946109235286712,
      "num_tokens": 114208413.0,
      "step": 45065
    },
    {
      "entropy": 6.041269111633301,
      "epoch": 5.20138488170802,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002647369914322435,
      "loss": 5.4219,
      "mean_token_accuracy": 0.22239066809415817,
      "num_tokens": 114220419.0,
      "step": 45070
    },
    {
      "entropy": 5.9519435405731205,
      "epoch": 5.20196191575303,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026469577020246475,
      "loss": 5.3752,
      "mean_token_accuracy": 0.22457602620124817,
      "num_tokens": 114232516.0,
      "step": 45075
    },
    {
      "entropy": 6.177624607086182,
      "epoch": 5.202538949798038,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026465454931926376,
      "loss": 5.5322,
      "mean_token_accuracy": 0.2095390170812607,
      "num_tokens": 114245719.0,
      "step": 45080
    },
    {
      "entropy": 6.0798999786376955,
      "epoch": 5.2031159838430465,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002646133287840267,
      "loss": 5.3975,
      "mean_token_accuracy": 0.21895123571157454,
      "num_tokens": 114257329.0,
      "step": 45085
    },
    {
      "entropy": 6.051014947891235,
      "epoch": 5.203693017888055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002645721085981403,
      "loss": 5.4513,
      "mean_token_accuracy": 0.22007804811000825,
      "num_tokens": 114271343.0,
      "step": 45090
    },
    {
      "entropy": 5.978876256942749,
      "epoch": 5.204270051933064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026453088876299074,
      "loss": 5.3819,
      "mean_token_accuracy": 0.22488144785165787,
      "num_tokens": 114284809.0,
      "step": 45095
    },
    {
      "entropy": 6.070902538299561,
      "epoch": 5.204847085978073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026448966927996464,
      "loss": 5.4654,
      "mean_token_accuracy": 0.21619082391262054,
      "num_tokens": 114296513.0,
      "step": 45100
    },
    {
      "entropy": 6.078530263900757,
      "epoch": 5.205424120023081,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002644484501504482,
      "loss": 5.4495,
      "mean_token_accuracy": 0.21460527032613755,
      "num_tokens": 114307899.0,
      "step": 45105
    },
    {
      "entropy": 6.129922199249267,
      "epoch": 5.20600115406809,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000264407231375828,
      "loss": 5.5275,
      "mean_token_accuracy": 0.2129510149359703,
      "num_tokens": 114321279.0,
      "step": 45110
    },
    {
      "entropy": 6.056185102462768,
      "epoch": 5.206578188113099,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002643660129574902,
      "loss": 5.4078,
      "mean_token_accuracy": 0.2148822322487831,
      "num_tokens": 114333125.0,
      "step": 45115
    },
    {
      "entropy": 6.039197063446045,
      "epoch": 5.207155222158107,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00026432479489682135,
      "loss": 5.4342,
      "mean_token_accuracy": 0.21803785413503646,
      "num_tokens": 114345378.0,
      "step": 45120
    },
    {
      "entropy": 6.014359712600708,
      "epoch": 5.207732256203116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002642835771952077,
      "loss": 5.4194,
      "mean_token_accuracy": 0.22385931760072708,
      "num_tokens": 114359177.0,
      "step": 45125
    },
    {
      "entropy": 6.118495273590088,
      "epoch": 5.208309290248125,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002642423598540356,
      "loss": 5.4692,
      "mean_token_accuracy": 0.21097501516342163,
      "num_tokens": 114372276.0,
      "step": 45130
    },
    {
      "entropy": 5.990920782089233,
      "epoch": 5.208886324293133,
      "grad_norm": 1.0,
      "learning_rate": 0.0002642011428746914,
      "loss": 5.3842,
      "mean_token_accuracy": 0.22625180929899216,
      "num_tokens": 114384309.0,
      "step": 45135
    },
    {
      "entropy": 6.005770683288574,
      "epoch": 5.209463358338142,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00026415992625856136,
      "loss": 5.4326,
      "mean_token_accuracy": 0.22662307173013688,
      "num_tokens": 114396947.0,
      "step": 45140
    },
    {
      "entropy": 6.15014910697937,
      "epoch": 5.21004039238315,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002641187100070319,
      "loss": 5.5391,
      "mean_token_accuracy": 0.2087978184223175,
      "num_tokens": 114408781.0,
      "step": 45145
    },
    {
      "entropy": 6.003570985794068,
      "epoch": 5.2106174264281595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002640774941214891,
      "loss": 5.3817,
      "mean_token_accuracy": 0.23329296112060546,
      "num_tokens": 114422642.0,
      "step": 45150
    },
    {
      "entropy": 5.985604095458984,
      "epoch": 5.211194460473168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002640362786033194,
      "loss": 5.3926,
      "mean_token_accuracy": 0.22371471375226976,
      "num_tokens": 114435965.0,
      "step": 45155
    },
    {
      "entropy": 6.086255407333374,
      "epoch": 5.211771494518176,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002639950634539091,
      "loss": 5.4629,
      "mean_token_accuracy": 0.21352626830339433,
      "num_tokens": 114447836.0,
      "step": 45160
    },
    {
      "entropy": 6.0727424144744875,
      "epoch": 5.212348528563186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026395384867464427,
      "loss": 5.454,
      "mean_token_accuracy": 0.2201738879084587,
      "num_tokens": 114460108.0,
      "step": 45165
    },
    {
      "entropy": 6.070657539367676,
      "epoch": 5.212925562608194,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026391263426691133,
      "loss": 5.4588,
      "mean_token_accuracy": 0.21811470091342927,
      "num_tokens": 114472732.0,
      "step": 45170
    },
    {
      "entropy": 6.054652452468872,
      "epoch": 5.2135025966532025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026387142023209635,
      "loss": 5.4401,
      "mean_token_accuracy": 0.2167653888463974,
      "num_tokens": 114485047.0,
      "step": 45175
    },
    {
      "entropy": 6.084554004669189,
      "epoch": 5.214079630698211,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026383020657158574,
      "loss": 5.4792,
      "mean_token_accuracy": 0.2225968673825264,
      "num_tokens": 114497512.0,
      "step": 45180
    },
    {
      "entropy": 6.039117527008057,
      "epoch": 5.21465666474322,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002637889932867655,
      "loss": 5.4051,
      "mean_token_accuracy": 0.22163076102733612,
      "num_tokens": 114511633.0,
      "step": 45185
    },
    {
      "entropy": 6.075606870651245,
      "epoch": 5.215233698788229,
      "grad_norm": 0.9375,
      "learning_rate": 0.000263747780379022,
      "loss": 5.376,
      "mean_token_accuracy": 0.2312086835503578,
      "num_tokens": 114524717.0,
      "step": 45190
    },
    {
      "entropy": 6.0291016578674315,
      "epoch": 5.215810732833237,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00026370656784974113,
      "loss": 5.4307,
      "mean_token_accuracy": 0.2245592087507248,
      "num_tokens": 114537531.0,
      "step": 45195
    },
    {
      "entropy": 5.994745206832886,
      "epoch": 5.2163877668782455,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00026366535570030943,
      "loss": 5.4346,
      "mean_token_accuracy": 0.22165384590625764,
      "num_tokens": 114551554.0,
      "step": 45200
    },
    {
      "entropy": 6.053288745880127,
      "epoch": 5.216964800923255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002636241439321127,
      "loss": 5.4485,
      "mean_token_accuracy": 0.22053719460964202,
      "num_tokens": 114566059.0,
      "step": 45205
    },
    {
      "entropy": 5.948758935928344,
      "epoch": 5.217541834968263,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002635829325465374,
      "loss": 5.3114,
      "mean_token_accuracy": 0.23304682224988937,
      "num_tokens": 114579157.0,
      "step": 45210
    },
    {
      "entropy": 6.009933614730835,
      "epoch": 5.218118869013272,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026354172154496934,
      "loss": 5.3887,
      "mean_token_accuracy": 0.2248021423816681,
      "num_tokens": 114591227.0,
      "step": 45215
    },
    {
      "entropy": 5.981044578552246,
      "epoch": 5.21869590305828,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002635005109287949,
      "loss": 5.4187,
      "mean_token_accuracy": 0.21806665062904357,
      "num_tokens": 114602649.0,
      "step": 45220
    },
    {
      "entropy": 6.039129543304443,
      "epoch": 5.219272937103289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002634593006994,
      "loss": 5.4102,
      "mean_token_accuracy": 0.22095916867256166,
      "num_tokens": 114615128.0,
      "step": 45225
    },
    {
      "entropy": 6.0418919086456295,
      "epoch": 5.219849971148298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026341809085817083,
      "loss": 5.4016,
      "mean_token_accuracy": 0.22474279701709748,
      "num_tokens": 114627901.0,
      "step": 45230
    },
    {
      "entropy": 6.0987687587738035,
      "epoch": 5.220427005193306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026337688140649337,
      "loss": 5.4566,
      "mean_token_accuracy": 0.22007470428943635,
      "num_tokens": 114639580.0,
      "step": 45235
    },
    {
      "entropy": 6.10085015296936,
      "epoch": 5.2210040392383155,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002633356723457538,
      "loss": 5.5598,
      "mean_token_accuracy": 0.21457237303256987,
      "num_tokens": 114651724.0,
      "step": 45240
    },
    {
      "entropy": 6.180582857131958,
      "epoch": 5.221581073283324,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000263294463677338,
      "loss": 5.5101,
      "mean_token_accuracy": 0.21046366393566132,
      "num_tokens": 114663642.0,
      "step": 45245
    },
    {
      "entropy": 6.084267807006836,
      "epoch": 5.222158107328332,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002632532554026322,
      "loss": 5.4316,
      "mean_token_accuracy": 0.2243875578045845,
      "num_tokens": 114675504.0,
      "step": 45250
    },
    {
      "entropy": 6.0650732040405275,
      "epoch": 5.222735141373341,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002632120475230222,
      "loss": 5.4879,
      "mean_token_accuracy": 0.21120632886886598,
      "num_tokens": 114687217.0,
      "step": 45255
    },
    {
      "entropy": 6.097438478469849,
      "epoch": 5.22331217541835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002631708400398942,
      "loss": 5.4815,
      "mean_token_accuracy": 0.22111869156360625,
      "num_tokens": 114699574.0,
      "step": 45260
    },
    {
      "entropy": 6.0115748882293705,
      "epoch": 5.223889209463358,
      "grad_norm": 0.890625,
      "learning_rate": 0.00026312963295463415,
      "loss": 5.3801,
      "mean_token_accuracy": 0.2280147925019264,
      "num_tokens": 114714023.0,
      "step": 45265
    },
    {
      "entropy": 5.943530797958374,
      "epoch": 5.224466243508367,
      "grad_norm": 1.015625,
      "learning_rate": 0.000263088426268628,
      "loss": 5.4008,
      "mean_token_accuracy": 0.22573857009410858,
      "num_tokens": 114726221.0,
      "step": 45270
    },
    {
      "entropy": 6.060222911834717,
      "epoch": 5.225043277553375,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002630472199832616,
      "loss": 5.3801,
      "mean_token_accuracy": 0.2247287943959236,
      "num_tokens": 114739699.0,
      "step": 45275
    },
    {
      "entropy": 5.961628532409668,
      "epoch": 5.225620311598385,
      "grad_norm": 0.90625,
      "learning_rate": 0.00026300601409992114,
      "loss": 5.334,
      "mean_token_accuracy": 0.22202873975038528,
      "num_tokens": 114752814.0,
      "step": 45280
    },
    {
      "entropy": 6.015863943099975,
      "epoch": 5.226197345643393,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002629648086199924,
      "loss": 5.3766,
      "mean_token_accuracy": 0.221244478225708,
      "num_tokens": 114766238.0,
      "step": 45285
    },
    {
      "entropy": 6.008779382705688,
      "epoch": 5.226774379688401,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026292360354486136,
      "loss": 5.4437,
      "mean_token_accuracy": 0.21783802956342696,
      "num_tokens": 114778576.0,
      "step": 45290
    },
    {
      "entropy": 6.04606614112854,
      "epoch": 5.227351413733411,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002628823988759139,
      "loss": 5.4661,
      "mean_token_accuracy": 0.21952925771474838,
      "num_tokens": 114790657.0,
      "step": 45295
    },
    {
      "entropy": 5.973862743377685,
      "epoch": 5.227928447778419,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026284119461453595,
      "loss": 5.4014,
      "mean_token_accuracy": 0.22855849862098693,
      "num_tokens": 114805053.0,
      "step": 45300
    },
    {
      "entropy": 6.011316919326783,
      "epoch": 5.228505481823428,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026279999076211333,
      "loss": 5.3663,
      "mean_token_accuracy": 0.23429108709096907,
      "num_tokens": 114818377.0,
      "step": 45305
    },
    {
      "entropy": 5.965339851379395,
      "epoch": 5.229082515868436,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026275878732003207,
      "loss": 5.3582,
      "mean_token_accuracy": 0.22599933594465255,
      "num_tokens": 114830692.0,
      "step": 45310
    },
    {
      "entropy": 5.955980682373047,
      "epoch": 5.229659549913445,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026271758428967777,
      "loss": 5.3138,
      "mean_token_accuracy": 0.23254485428333282,
      "num_tokens": 114843518.0,
      "step": 45315
    },
    {
      "entropy": 5.981691122055054,
      "epoch": 5.230236583958454,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026267638167243654,
      "loss": 5.3846,
      "mean_token_accuracy": 0.22669032961130142,
      "num_tokens": 114856226.0,
      "step": 45320
    },
    {
      "entropy": 6.090505123138428,
      "epoch": 5.230813618003462,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026263517946969393,
      "loss": 5.5583,
      "mean_token_accuracy": 0.21266670078039168,
      "num_tokens": 114867775.0,
      "step": 45325
    },
    {
      "entropy": 6.058550119400024,
      "epoch": 5.2313906520484705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026259397768283615,
      "loss": 5.3965,
      "mean_token_accuracy": 0.22703013718128204,
      "num_tokens": 114879156.0,
      "step": 45330
    },
    {
      "entropy": 6.0846270561218265,
      "epoch": 5.23196768609348,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002625527763132486,
      "loss": 5.4585,
      "mean_token_accuracy": 0.22403827160596848,
      "num_tokens": 114891144.0,
      "step": 45335
    },
    {
      "entropy": 5.919776439666748,
      "epoch": 5.232544720138488,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00026251157536231735,
      "loss": 5.2769,
      "mean_token_accuracy": 0.2378045827150345,
      "num_tokens": 114905460.0,
      "step": 45340
    },
    {
      "entropy": 6.04665732383728,
      "epoch": 5.233121754183497,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002624703748314279,
      "loss": 5.4511,
      "mean_token_accuracy": 0.22117620706558228,
      "num_tokens": 114918261.0,
      "step": 45345
    },
    {
      "entropy": 6.071588945388794,
      "epoch": 5.233698788228505,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002624291747219663,
      "loss": 5.4599,
      "mean_token_accuracy": 0.2253224402666092,
      "num_tokens": 114930339.0,
      "step": 45350
    },
    {
      "entropy": 6.014683437347412,
      "epoch": 5.234275822273514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026238797503531816,
      "loss": 5.4018,
      "mean_token_accuracy": 0.22180467396974562,
      "num_tokens": 114942985.0,
      "step": 45355
    },
    {
      "entropy": 6.12723388671875,
      "epoch": 5.234852856318523,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002623467757728691,
      "loss": 5.4964,
      "mean_token_accuracy": 0.21537711322307587,
      "num_tokens": 114955270.0,
      "step": 45360
    },
    {
      "entropy": 6.043020296096802,
      "epoch": 5.235429890363531,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026230557693600505,
      "loss": 5.4814,
      "mean_token_accuracy": 0.217724771797657,
      "num_tokens": 114967585.0,
      "step": 45365
    },
    {
      "entropy": 6.023756313323974,
      "epoch": 5.2360069244085405,
      "grad_norm": 0.984375,
      "learning_rate": 0.00026226437852611156,
      "loss": 5.4766,
      "mean_token_accuracy": 0.2226881802082062,
      "num_tokens": 114980714.0,
      "step": 45370
    },
    {
      "entropy": 6.091431999206543,
      "epoch": 5.236583958453549,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002622231805445745,
      "loss": 5.4456,
      "mean_token_accuracy": 0.21698398888111115,
      "num_tokens": 114992453.0,
      "step": 45375
    },
    {
      "entropy": 6.051722526550293,
      "epoch": 5.237160992498557,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002621819829927793,
      "loss": 5.3861,
      "mean_token_accuracy": 0.22146398574113846,
      "num_tokens": 115004147.0,
      "step": 45380
    },
    {
      "entropy": 5.998739957809448,
      "epoch": 5.237738026543566,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002621407858721117,
      "loss": 5.4289,
      "mean_token_accuracy": 0.22249038517475128,
      "num_tokens": 115017131.0,
      "step": 45385
    },
    {
      "entropy": 5.991521787643433,
      "epoch": 5.238315060588575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026209958918395744,
      "loss": 5.4288,
      "mean_token_accuracy": 0.2191956713795662,
      "num_tokens": 115030872.0,
      "step": 45390
    },
    {
      "entropy": 6.0833639144897464,
      "epoch": 5.2388920946335835,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002620583929297022,
      "loss": 5.5547,
      "mean_token_accuracy": 0.20820025205612183,
      "num_tokens": 115044181.0,
      "step": 45395
    },
    {
      "entropy": 5.99387035369873,
      "epoch": 5.239469128678592,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002620171971107313,
      "loss": 5.2937,
      "mean_token_accuracy": 0.23721488416194916,
      "num_tokens": 115057066.0,
      "step": 45400
    },
    {
      "entropy": 6.179176092147827,
      "epoch": 5.2400461627236,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026197600172843063,
      "loss": 5.6402,
      "mean_token_accuracy": 0.204765285551548,
      "num_tokens": 115070292.0,
      "step": 45405
    },
    {
      "entropy": 6.0735204219818115,
      "epoch": 5.24062319676861,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002619348067841856,
      "loss": 5.49,
      "mean_token_accuracy": 0.21091417372226715,
      "num_tokens": 115083101.0,
      "step": 45410
    },
    {
      "entropy": 6.049989318847656,
      "epoch": 5.241200230813618,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026189361227938195,
      "loss": 5.4554,
      "mean_token_accuracy": 0.2197912648320198,
      "num_tokens": 115095615.0,
      "step": 45415
    },
    {
      "entropy": 5.99624195098877,
      "epoch": 5.2417772648586265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026185241821540504,
      "loss": 5.4355,
      "mean_token_accuracy": 0.2310975581407547,
      "num_tokens": 115108347.0,
      "step": 45420
    },
    {
      "entropy": 6.043016529083252,
      "epoch": 5.242354298903635,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002618112245936406,
      "loss": 5.404,
      "mean_token_accuracy": 0.22040394991636275,
      "num_tokens": 115120022.0,
      "step": 45425
    },
    {
      "entropy": 5.9708325386047365,
      "epoch": 5.242931332948644,
      "grad_norm": 1.015625,
      "learning_rate": 0.00026177003141547394,
      "loss": 5.433,
      "mean_token_accuracy": 0.21992162764072418,
      "num_tokens": 115133471.0,
      "step": 45430
    },
    {
      "entropy": 6.0564188957214355,
      "epoch": 5.243508366993653,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026172883868229085,
      "loss": 5.4263,
      "mean_token_accuracy": 0.22573087513446807,
      "num_tokens": 115145027.0,
      "step": 45435
    },
    {
      "entropy": 6.149946308135986,
      "epoch": 5.244085401038661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002616876463954766,
      "loss": 5.5686,
      "mean_token_accuracy": 0.21413047313690187,
      "num_tokens": 115156420.0,
      "step": 45440
    },
    {
      "entropy": 6.017841339111328,
      "epoch": 5.24466243508367,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026164645455641685,
      "loss": 5.4071,
      "mean_token_accuracy": 0.22935985326766967,
      "num_tokens": 115169888.0,
      "step": 45445
    },
    {
      "entropy": 6.000409507751465,
      "epoch": 5.245239469128679,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002616052631664968,
      "loss": 5.3867,
      "mean_token_accuracy": 0.22754311710596084,
      "num_tokens": 115182899.0,
      "step": 45450
    },
    {
      "entropy": 6.005731439590454,
      "epoch": 5.245816503173687,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002615640722271023,
      "loss": 5.4608,
      "mean_token_accuracy": 0.22232909202575685,
      "num_tokens": 115196270.0,
      "step": 45455
    },
    {
      "entropy": 5.947475147247315,
      "epoch": 5.246393537218696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002615228817396183,
      "loss": 5.315,
      "mean_token_accuracy": 0.23969982117414473,
      "num_tokens": 115208846.0,
      "step": 45460
    },
    {
      "entropy": 6.056642198562622,
      "epoch": 5.246970571263705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002614816917054306,
      "loss": 5.4162,
      "mean_token_accuracy": 0.22341176569461824,
      "num_tokens": 115221442.0,
      "step": 45465
    },
    {
      "entropy": 6.0510334968566895,
      "epoch": 5.247547605308713,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002614405021259245,
      "loss": 5.4568,
      "mean_token_accuracy": 0.21845282912254332,
      "num_tokens": 115233965.0,
      "step": 45470
    },
    {
      "entropy": 6.004269647598266,
      "epoch": 5.248124639353722,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00026139931300248536,
      "loss": 5.4203,
      "mean_token_accuracy": 0.21801410913467406,
      "num_tokens": 115246907.0,
      "step": 45475
    },
    {
      "entropy": 6.0250917911529545,
      "epoch": 5.24870167339873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026135812433649856,
      "loss": 5.4083,
      "mean_token_accuracy": 0.21793183833360671,
      "num_tokens": 115260161.0,
      "step": 45480
    },
    {
      "entropy": 6.0069538116455075,
      "epoch": 5.2492787074437395,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002613169361293495,
      "loss": 5.3858,
      "mean_token_accuracy": 0.23049605786800384,
      "num_tokens": 115274029.0,
      "step": 45485
    },
    {
      "entropy": 5.968023443222046,
      "epoch": 5.249855741488748,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002612757483824234,
      "loss": 5.36,
      "mean_token_accuracy": 0.22484962195158004,
      "num_tokens": 115286899.0,
      "step": 45490
    },
    {
      "entropy": 6.05524377822876,
      "epoch": 5.250432775533756,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026123456109710584,
      "loss": 5.4819,
      "mean_token_accuracy": 0.21921216994524,
      "num_tokens": 115299435.0,
      "step": 45495
    },
    {
      "entropy": 5.948203039169312,
      "epoch": 5.251009809578765,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002611933742747819,
      "loss": 5.2605,
      "mean_token_accuracy": 0.23185878694057466,
      "num_tokens": 115311916.0,
      "step": 45500
    },
    {
      "entropy": 5.944322347640991,
      "epoch": 5.251586843623774,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026115218791683694,
      "loss": 5.41,
      "mean_token_accuracy": 0.22838055789470674,
      "num_tokens": 115324772.0,
      "step": 45505
    },
    {
      "entropy": 6.032688188552856,
      "epoch": 5.2521638776687825,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002611110020246563,
      "loss": 5.4389,
      "mean_token_accuracy": 0.21994860470294952,
      "num_tokens": 115337056.0,
      "step": 45510
    },
    {
      "entropy": 5.986717987060547,
      "epoch": 5.252740911713791,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002610698165996252,
      "loss": 5.3927,
      "mean_token_accuracy": 0.2265930250287056,
      "num_tokens": 115349137.0,
      "step": 45515
    },
    {
      "entropy": 6.063390827178955,
      "epoch": 5.2533179457588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026102863164312887,
      "loss": 5.4233,
      "mean_token_accuracy": 0.2270804077386856,
      "num_tokens": 115362738.0,
      "step": 45520
    },
    {
      "entropy": 6.02864031791687,
      "epoch": 5.253894979803809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002609874471565526,
      "loss": 5.4479,
      "mean_token_accuracy": 0.2173398479819298,
      "num_tokens": 115374894.0,
      "step": 45525
    },
    {
      "entropy": 6.051260948181152,
      "epoch": 5.254472013848817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002609462631412815,
      "loss": 5.3779,
      "mean_token_accuracy": 0.22599852234125137,
      "num_tokens": 115387528.0,
      "step": 45530
    },
    {
      "entropy": 5.980574750900269,
      "epoch": 5.255049047893825,
      "grad_norm": 1.046875,
      "learning_rate": 0.000260905079598701,
      "loss": 5.3579,
      "mean_token_accuracy": 0.22937343865633011,
      "num_tokens": 115401002.0,
      "step": 45535
    },
    {
      "entropy": 6.044390439987183,
      "epoch": 5.255626081938835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026086389653019607,
      "loss": 5.457,
      "mean_token_accuracy": 0.21695838868618011,
      "num_tokens": 115411941.0,
      "step": 45540
    },
    {
      "entropy": 6.036340951919556,
      "epoch": 5.256203115983843,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000260822713937152,
      "loss": 5.3961,
      "mean_token_accuracy": 0.22130239605903626,
      "num_tokens": 115424426.0,
      "step": 45545
    },
    {
      "entropy": 6.075245380401611,
      "epoch": 5.256780150028852,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00026078153182095383,
      "loss": 5.4382,
      "mean_token_accuracy": 0.21842439323663712,
      "num_tokens": 115437329.0,
      "step": 45550
    },
    {
      "entropy": 6.058662080764771,
      "epoch": 5.25735718407386,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026074035018298686,
      "loss": 5.466,
      "mean_token_accuracy": 0.2246864840388298,
      "num_tokens": 115449879.0,
      "step": 45555
    },
    {
      "entropy": 6.060898113250732,
      "epoch": 5.257934218118869,
      "grad_norm": 1.078125,
      "learning_rate": 0.000260699169024636,
      "loss": 5.4304,
      "mean_token_accuracy": 0.22123443335294724,
      "num_tokens": 115461854.0,
      "step": 45560
    },
    {
      "entropy": 6.010336875915527,
      "epoch": 5.258511252163878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00026065798834728674,
      "loss": 5.4342,
      "mean_token_accuracy": 0.2219907432794571,
      "num_tokens": 115475899.0,
      "step": 45565
    },
    {
      "entropy": 5.982123565673828,
      "epoch": 5.259088286208886,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026061680815232357,
      "loss": 5.3028,
      "mean_token_accuracy": 0.2310809910297394,
      "num_tokens": 115489532.0,
      "step": 45570
    },
    {
      "entropy": 6.12529673576355,
      "epoch": 5.2596653202538945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002605756284411322,
      "loss": 5.5408,
      "mean_token_accuracy": 0.20360140949487687,
      "num_tokens": 115502188.0,
      "step": 45575
    },
    {
      "entropy": 6.005586051940918,
      "epoch": 5.260242354298904,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026053444921509723,
      "loss": 5.4329,
      "mean_token_accuracy": 0.22544600069522858,
      "num_tokens": 115514346.0,
      "step": 45580
    },
    {
      "entropy": 6.1189343452453615,
      "epoch": 5.260819388343912,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000260493270475604,
      "loss": 5.4924,
      "mean_token_accuracy": 0.214700448513031,
      "num_tokens": 115525855.0,
      "step": 45585
    },
    {
      "entropy": 6.070477294921875,
      "epoch": 5.261396422388921,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002604520922240373,
      "loss": 5.4442,
      "mean_token_accuracy": 0.22412650287151337,
      "num_tokens": 115538115.0,
      "step": 45590
    },
    {
      "entropy": 6.091808462142945,
      "epoch": 5.26197345643393,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002604109144617823,
      "loss": 5.5031,
      "mean_token_accuracy": 0.21764739751815795,
      "num_tokens": 115550827.0,
      "step": 45595
    },
    {
      "entropy": 6.004638004302978,
      "epoch": 5.262550490478938,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000260369737190224,
      "loss": 5.3885,
      "mean_token_accuracy": 0.22925753891468048,
      "num_tokens": 115563790.0,
      "step": 45600
    },
    {
      "entropy": 6.1441141128540036,
      "epoch": 5.263127524523947,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002603285604107473,
      "loss": 5.632,
      "mean_token_accuracy": 0.20858521163463592,
      "num_tokens": 115577194.0,
      "step": 45605
    },
    {
      "entropy": 6.0799883842468265,
      "epoch": 5.263704558568955,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002602873841247371,
      "loss": 5.4538,
      "mean_token_accuracy": 0.22601577788591384,
      "num_tokens": 115590385.0,
      "step": 45610
    },
    {
      "entropy": 6.084481906890869,
      "epoch": 5.2642815926139646,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002602462083335785,
      "loss": 5.4373,
      "mean_token_accuracy": 0.2222571238875389,
      "num_tokens": 115602809.0,
      "step": 45615
    },
    {
      "entropy": 5.991938924789428,
      "epoch": 5.264858626658973,
      "grad_norm": 0.96875,
      "learning_rate": 0.00026020503303865634,
      "loss": 5.4214,
      "mean_token_accuracy": 0.2216294676065445,
      "num_tokens": 115616688.0,
      "step": 45620
    },
    {
      "entropy": 6.006976461410522,
      "epoch": 5.265435660703981,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002601638582413556,
      "loss": 5.4335,
      "mean_token_accuracy": 0.2221789687871933,
      "num_tokens": 115629608.0,
      "step": 45625
    },
    {
      "entropy": 6.02233920097351,
      "epoch": 5.26601269474899,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000260122683943061,
      "loss": 5.3138,
      "mean_token_accuracy": 0.2278474673628807,
      "num_tokens": 115641478.0,
      "step": 45630
    },
    {
      "entropy": 6.138080692291259,
      "epoch": 5.266589728793999,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002600815101451576,
      "loss": 5.5255,
      "mean_token_accuracy": 0.20745670199394226,
      "num_tokens": 115653490.0,
      "step": 45635
    },
    {
      "entropy": 6.032846069335937,
      "epoch": 5.2671667628390075,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002600403368490302,
      "loss": 5.4494,
      "mean_token_accuracy": 0.22217857986688613,
      "num_tokens": 115665547.0,
      "step": 45640
    },
    {
      "entropy": 6.134988594055176,
      "epoch": 5.267743796884016,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025999916405606367,
      "loss": 5.4888,
      "mean_token_accuracy": 0.21783046126365663,
      "num_tokens": 115679102.0,
      "step": 45645
    },
    {
      "entropy": 6.056014633178711,
      "epoch": 5.268320830929024,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025995799176764273,
      "loss": 5.4184,
      "mean_token_accuracy": 0.22798133194446563,
      "num_tokens": 115693469.0,
      "step": 45650
    },
    {
      "entropy": 6.077406549453736,
      "epoch": 5.268897864974034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002599168199851523,
      "loss": 5.5079,
      "mean_token_accuracy": 0.21468451917171477,
      "num_tokens": 115706327.0,
      "step": 45655
    },
    {
      "entropy": 5.939985179901123,
      "epoch": 5.269474899019042,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002598756487099771,
      "loss": 5.3166,
      "mean_token_accuracy": 0.23756767064332962,
      "num_tokens": 115719129.0,
      "step": 45660
    },
    {
      "entropy": 6.025173759460449,
      "epoch": 5.2700519330640505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025983447794350207,
      "loss": 5.418,
      "mean_token_accuracy": 0.22448321729898452,
      "num_tokens": 115730117.0,
      "step": 45665
    },
    {
      "entropy": 6.073547601699829,
      "epoch": 5.27062896710906,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00025979330768711165,
      "loss": 5.4996,
      "mean_token_accuracy": 0.21458194106817247,
      "num_tokens": 115743743.0,
      "step": 45670
    },
    {
      "entropy": 6.089326000213623,
      "epoch": 5.271206001154068,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025975213794219096,
      "loss": 5.4931,
      "mean_token_accuracy": 0.2180785819888115,
      "num_tokens": 115756290.0,
      "step": 45675
    },
    {
      "entropy": 6.06765456199646,
      "epoch": 5.271783035199077,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002597109687101244,
      "loss": 5.4337,
      "mean_token_accuracy": 0.22488387078046798,
      "num_tokens": 115769489.0,
      "step": 45680
    },
    {
      "entropy": 6.073848676681519,
      "epoch": 5.272360069244085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002596697999922969,
      "loss": 5.4384,
      "mean_token_accuracy": 0.21921130418777465,
      "num_tokens": 115782557.0,
      "step": 45685
    },
    {
      "entropy": 5.98873119354248,
      "epoch": 5.272937103289094,
      "grad_norm": 0.90625,
      "learning_rate": 0.000259628631790093,
      "loss": 5.3804,
      "mean_token_accuracy": 0.23462460339069366,
      "num_tokens": 115795706.0,
      "step": 45690
    },
    {
      "entropy": 6.069477844238281,
      "epoch": 5.273514137334103,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025958746410489745,
      "loss": 5.4902,
      "mean_token_accuracy": 0.21369931250810623,
      "num_tokens": 115808736.0,
      "step": 45695
    },
    {
      "entropy": 6.03880877494812,
      "epoch": 5.274091171379111,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002595462969380949,
      "loss": 5.4314,
      "mean_token_accuracy": 0.22609738856554032,
      "num_tokens": 115820862.0,
      "step": 45700
    },
    {
      "entropy": 6.087772083282471,
      "epoch": 5.27466820542412,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025950513029106997,
      "loss": 5.4789,
      "mean_token_accuracy": 0.21456007212400435,
      "num_tokens": 115833334.0,
      "step": 45705
    },
    {
      "entropy": 5.966684484481812,
      "epoch": 5.275245239469129,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002594639641652073,
      "loss": 5.3186,
      "mean_token_accuracy": 0.22581247240304947,
      "num_tokens": 115846965.0,
      "step": 45710
    },
    {
      "entropy": 5.989447069168091,
      "epoch": 5.275822273514137,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002594227985618915,
      "loss": 5.4214,
      "mean_token_accuracy": 0.22116950750350953,
      "num_tokens": 115858381.0,
      "step": 45715
    },
    {
      "entropy": 5.9788047790527346,
      "epoch": 5.276399307559146,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025938163348250705,
      "loss": 5.3615,
      "mean_token_accuracy": 0.22843966335058213,
      "num_tokens": 115870523.0,
      "step": 45720
    },
    {
      "entropy": 6.080673551559448,
      "epoch": 5.276976341604154,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002593404689284387,
      "loss": 5.4317,
      "mean_token_accuracy": 0.22181939333677292,
      "num_tokens": 115882938.0,
      "step": 45725
    },
    {
      "entropy": 6.080872821807861,
      "epoch": 5.2775533756491635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00025929930490107086,
      "loss": 5.42,
      "mean_token_accuracy": 0.21461830288171768,
      "num_tokens": 115895364.0,
      "step": 45730
    },
    {
      "entropy": 5.9829041957855225,
      "epoch": 5.278130409694172,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002592581414017881,
      "loss": 5.3655,
      "mean_token_accuracy": 0.22573001086711883,
      "num_tokens": 115908995.0,
      "step": 45735
    },
    {
      "entropy": 5.987601947784424,
      "epoch": 5.27870744373918,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002592169784319749,
      "loss": 5.3298,
      "mean_token_accuracy": 0.22656229734420777,
      "num_tokens": 115921910.0,
      "step": 45740
    },
    {
      "entropy": 6.074593162536621,
      "epoch": 5.27928447778419,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025917581599301583,
      "loss": 5.4642,
      "mean_token_accuracy": 0.22030116319656373,
      "num_tokens": 115936895.0,
      "step": 45745
    },
    {
      "entropy": 6.057330369949341,
      "epoch": 5.279861511829198,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002591346540862953,
      "loss": 5.3785,
      "mean_token_accuracy": 0.23026176244020463,
      "num_tokens": 115948541.0,
      "step": 45750
    },
    {
      "entropy": 6.049112319946289,
      "epoch": 5.2804385458742065,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002590934927131978,
      "loss": 5.4991,
      "mean_token_accuracy": 0.22063111662864685,
      "num_tokens": 115961635.0,
      "step": 45755
    },
    {
      "entropy": 5.915291213989258,
      "epoch": 5.281015579919215,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002590523318751078,
      "loss": 5.3373,
      "mean_token_accuracy": 0.22988480627536773,
      "num_tokens": 115974022.0,
      "step": 45760
    },
    {
      "entropy": 6.0499063491821286,
      "epoch": 5.281592613964224,
      "grad_norm": 1.0,
      "learning_rate": 0.0002590111715734096,
      "loss": 5.4767,
      "mean_token_accuracy": 0.2208188593387604,
      "num_tokens": 115987068.0,
      "step": 45765
    },
    {
      "entropy": 5.997831678390503,
      "epoch": 5.282169648009233,
      "grad_norm": 0.953125,
      "learning_rate": 0.00025897001180948787,
      "loss": 5.3338,
      "mean_token_accuracy": 0.2302880734205246,
      "num_tokens": 116000621.0,
      "step": 45770
    },
    {
      "entropy": 6.053753232955932,
      "epoch": 5.282746682054241,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002589288525847267,
      "loss": 5.4714,
      "mean_token_accuracy": 0.21738363355398177,
      "num_tokens": 116013321.0,
      "step": 45775
    },
    {
      "entropy": 6.013989210128784,
      "epoch": 5.2833237160992494,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025888769390051064,
      "loss": 5.427,
      "mean_token_accuracy": 0.21772615909576415,
      "num_tokens": 116025458.0,
      "step": 45780
    },
    {
      "entropy": 5.96270751953125,
      "epoch": 5.283900750144259,
      "grad_norm": 0.9375,
      "learning_rate": 0.000258846535758224,
      "loss": 5.3868,
      "mean_token_accuracy": 0.22570870220661163,
      "num_tokens": 116039574.0,
      "step": 45785
    },
    {
      "entropy": 6.051803112030029,
      "epoch": 5.284477784189267,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002588053781592511,
      "loss": 5.3861,
      "mean_token_accuracy": 0.2221839427947998,
      "num_tokens": 116052369.0,
      "step": 45790
    },
    {
      "entropy": 6.016679477691651,
      "epoch": 5.285054818234276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002587642211049762,
      "loss": 5.4829,
      "mean_token_accuracy": 0.21720351427793502,
      "num_tokens": 116064688.0,
      "step": 45795
    },
    {
      "entropy": 5.956248140335083,
      "epoch": 5.285631852279285,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025872306459678374,
      "loss": 5.2938,
      "mean_token_accuracy": 0.23636026084423065,
      "num_tokens": 116079254.0,
      "step": 45800
    },
    {
      "entropy": 6.03924789428711,
      "epoch": 5.286208886324293,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025868190863605786,
      "loss": 5.3819,
      "mean_token_accuracy": 0.22168154418468475,
      "num_tokens": 116091845.0,
      "step": 45805
    },
    {
      "entropy": 5.926907062530518,
      "epoch": 5.286785920369302,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000258640753224183,
      "loss": 5.2364,
      "mean_token_accuracy": 0.2368401825428009,
      "num_tokens": 116104111.0,
      "step": 45810
    },
    {
      "entropy": 5.934835004806518,
      "epoch": 5.28736295441431,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025859959836254314,
      "loss": 5.3555,
      "mean_token_accuracy": 0.22716329991817474,
      "num_tokens": 116116894.0,
      "step": 45815
    },
    {
      "entropy": 6.023984956741333,
      "epoch": 5.2879399884593195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002585584440525227,
      "loss": 5.4313,
      "mean_token_accuracy": 0.22012444138526915,
      "num_tokens": 116129045.0,
      "step": 45820
    },
    {
      "entropy": 6.0685387134552,
      "epoch": 5.288517022504328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002585172902955058,
      "loss": 5.4479,
      "mean_token_accuracy": 0.21874248683452607,
      "num_tokens": 116140569.0,
      "step": 45825
    },
    {
      "entropy": 6.094815349578857,
      "epoch": 5.289094056549336,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002584761370928767,
      "loss": 5.5058,
      "mean_token_accuracy": 0.21773842275142669,
      "num_tokens": 116152967.0,
      "step": 45830
    },
    {
      "entropy": 5.977795171737671,
      "epoch": 5.289671090594345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002584349844460194,
      "loss": 5.4079,
      "mean_token_accuracy": 0.22074140161275863,
      "num_tokens": 116164730.0,
      "step": 45835
    },
    {
      "entropy": 5.9802539348602295,
      "epoch": 5.290248124639354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002583938323563183,
      "loss": 5.3109,
      "mean_token_accuracy": 0.22782213240861893,
      "num_tokens": 116177239.0,
      "step": 45840
    },
    {
      "entropy": 6.014317989349365,
      "epoch": 5.290825158684362,
      "grad_norm": 1.0,
      "learning_rate": 0.0002583526808251573,
      "loss": 5.3303,
      "mean_token_accuracy": 0.23677037358283998,
      "num_tokens": 116189313.0,
      "step": 45845
    },
    {
      "entropy": 6.061425256729126,
      "epoch": 5.291402192729371,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002583115298539208,
      "loss": 5.442,
      "mean_token_accuracy": 0.22103138715028764,
      "num_tokens": 116201814.0,
      "step": 45850
    },
    {
      "entropy": 6.003198957443237,
      "epoch": 5.29197922677438,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002582703794439925,
      "loss": 5.4064,
      "mean_token_accuracy": 0.2240051031112671,
      "num_tokens": 116214857.0,
      "step": 45855
    },
    {
      "entropy": 6.0059796333312985,
      "epoch": 5.292556260819389,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025822922959675676,
      "loss": 5.4331,
      "mean_token_accuracy": 0.22569094598293304,
      "num_tokens": 116226528.0,
      "step": 45860
    },
    {
      "entropy": 6.0012462615966795,
      "epoch": 5.293133294864397,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025818808031359753,
      "loss": 5.3859,
      "mean_token_accuracy": 0.2283470779657364,
      "num_tokens": 116239120.0,
      "step": 45865
    },
    {
      "entropy": 5.941596555709839,
      "epoch": 5.293710328909405,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025814693159589895,
      "loss": 5.3297,
      "mean_token_accuracy": 0.22873965352773667,
      "num_tokens": 116251248.0,
      "step": 45870
    },
    {
      "entropy": 6.036790275573731,
      "epoch": 5.294287362954415,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025810578344504485,
      "loss": 5.4725,
      "mean_token_accuracy": 0.21692661941051483,
      "num_tokens": 116263536.0,
      "step": 45875
    },
    {
      "entropy": 6.0626240253448485,
      "epoch": 5.294864396999423,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025806463586241943,
      "loss": 5.4063,
      "mean_token_accuracy": 0.22704166769981385,
      "num_tokens": 116277119.0,
      "step": 45880
    },
    {
      "entropy": 5.9592243194580075,
      "epoch": 5.2954414310444315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002580234888494064,
      "loss": 5.3547,
      "mean_token_accuracy": 0.2287086546421051,
      "num_tokens": 116289315.0,
      "step": 45885
    },
    {
      "entropy": 6.0819557189941404,
      "epoch": 5.29601846508944,
      "grad_norm": 1.0,
      "learning_rate": 0.0002579823424073901,
      "loss": 5.5206,
      "mean_token_accuracy": 0.20929814130067825,
      "num_tokens": 116302322.0,
      "step": 45890
    },
    {
      "entropy": 6.033450937271118,
      "epoch": 5.296595499134449,
      "grad_norm": 1.0,
      "learning_rate": 0.0002579411965377541,
      "loss": 5.3729,
      "mean_token_accuracy": 0.22816126942634582,
      "num_tokens": 116314688.0,
      "step": 45895
    },
    {
      "entropy": 6.015622091293335,
      "epoch": 5.297172533179458,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025790005124188257,
      "loss": 5.3632,
      "mean_token_accuracy": 0.21758892983198166,
      "num_tokens": 116326242.0,
      "step": 45900
    },
    {
      "entropy": 5.946031475067139,
      "epoch": 5.297749567224466,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00025785890652115927,
      "loss": 5.3799,
      "mean_token_accuracy": 0.2278321698307991,
      "num_tokens": 116338794.0,
      "step": 45905
    },
    {
      "entropy": 6.084583282470703,
      "epoch": 5.2983266012694745,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025781776237696817,
      "loss": 5.5029,
      "mean_token_accuracy": 0.21626635491847992,
      "num_tokens": 116351215.0,
      "step": 45910
    },
    {
      "entropy": 6.027676296234131,
      "epoch": 5.298903635314484,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002577766188106931,
      "loss": 5.3917,
      "mean_token_accuracy": 0.2203213930130005,
      "num_tokens": 116363522.0,
      "step": 45915
    },
    {
      "entropy": 6.056950378417969,
      "epoch": 5.299480669359492,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025773547582371787,
      "loss": 5.44,
      "mean_token_accuracy": 0.21464321464300157,
      "num_tokens": 116374778.0,
      "step": 45920
    },
    {
      "entropy": 6.007820749282837,
      "epoch": 5.300057703404501,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002576943334174263,
      "loss": 5.3907,
      "mean_token_accuracy": 0.2243694022297859,
      "num_tokens": 116387546.0,
      "step": 45925
    },
    {
      "entropy": 6.010303354263305,
      "epoch": 5.30063473744951,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002576531915932024,
      "loss": 5.4524,
      "mean_token_accuracy": 0.21511386930942536,
      "num_tokens": 116401907.0,
      "step": 45930
    },
    {
      "entropy": 6.067706298828125,
      "epoch": 5.301211771494518,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00025761205035242964,
      "loss": 5.4194,
      "mean_token_accuracy": 0.22609938234090804,
      "num_tokens": 116414300.0,
      "step": 45935
    },
    {
      "entropy": 6.007501316070557,
      "epoch": 5.301788805539527,
      "grad_norm": 1.046875,
      "learning_rate": 0.000257570909696492,
      "loss": 5.3333,
      "mean_token_accuracy": 0.22426329255104066,
      "num_tokens": 116427313.0,
      "step": 45940
    },
    {
      "entropy": 6.078638362884521,
      "epoch": 5.302365839584535,
      "grad_norm": 1.03125,
      "learning_rate": 0.000257529769626773,
      "loss": 5.5283,
      "mean_token_accuracy": 0.21886655539274216,
      "num_tokens": 116441131.0,
      "step": 45945
    },
    {
      "entropy": 6.100323963165283,
      "epoch": 5.3029428736295445,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025748863014465667,
      "loss": 5.5551,
      "mean_token_accuracy": 0.21367793679237365,
      "num_tokens": 116454164.0,
      "step": 45950
    },
    {
      "entropy": 6.037721967697143,
      "epoch": 5.303519907674553,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002574474912515265,
      "loss": 5.4417,
      "mean_token_accuracy": 0.21962355971336364,
      "num_tokens": 116466836.0,
      "step": 45955
    },
    {
      "entropy": 6.092635583877564,
      "epoch": 5.304096941719561,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025740635294876636,
      "loss": 5.4381,
      "mean_token_accuracy": 0.2162007212638855,
      "num_tokens": 116478371.0,
      "step": 45960
    },
    {
      "entropy": 6.026547908782959,
      "epoch": 5.30467397576457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025736521523775965,
      "loss": 5.43,
      "mean_token_accuracy": 0.22337144315242768,
      "num_tokens": 116490871.0,
      "step": 45965
    },
    {
      "entropy": 6.121276426315307,
      "epoch": 5.305251009809579,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025732407811989024,
      "loss": 5.5556,
      "mean_token_accuracy": 0.2097402811050415,
      "num_tokens": 116503204.0,
      "step": 45970
    },
    {
      "entropy": 5.976197719573975,
      "epoch": 5.3058280438545875,
      "grad_norm": 1.0,
      "learning_rate": 0.00025728294159654163,
      "loss": 5.3238,
      "mean_token_accuracy": 0.229198095202446,
      "num_tokens": 116517020.0,
      "step": 45975
    },
    {
      "entropy": 6.0076884746551515,
      "epoch": 5.306405077899596,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002572418056690976,
      "loss": 5.4166,
      "mean_token_accuracy": 0.22581440806388856,
      "num_tokens": 116529760.0,
      "step": 45980
    },
    {
      "entropy": 6.026073169708252,
      "epoch": 5.306982111944604,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025720067033894144,
      "loss": 5.3895,
      "mean_token_accuracy": 0.22563980519771576,
      "num_tokens": 116542464.0,
      "step": 45985
    },
    {
      "entropy": 5.93888521194458,
      "epoch": 5.307559145989614,
      "grad_norm": 1.03125,
      "learning_rate": 0.000257159535607457,
      "loss": 5.2922,
      "mean_token_accuracy": 0.2282860517501831,
      "num_tokens": 116554550.0,
      "step": 45990
    },
    {
      "entropy": 6.052002620697022,
      "epoch": 5.308136180034622,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025711840147602774,
      "loss": 5.4443,
      "mean_token_accuracy": 0.22774358987808227,
      "num_tokens": 116566242.0,
      "step": 45995
    },
    {
      "entropy": 6.092132043838501,
      "epoch": 5.3087132140796305,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025707726794603714,
      "loss": 5.4895,
      "mean_token_accuracy": 0.21957289427518845,
      "num_tokens": 116579108.0,
      "step": 46000
    },
    {
      "entropy": 5.944806098937988,
      "epoch": 5.30929024812464,
      "grad_norm": 1.125,
      "learning_rate": 0.0002570361350188687,
      "loss": 5.2938,
      "mean_token_accuracy": 0.23232382982969285,
      "num_tokens": 116590918.0,
      "step": 46005
    },
    {
      "entropy": 6.070037364959717,
      "epoch": 5.309867282169648,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025699500269590593,
      "loss": 5.4645,
      "mean_token_accuracy": 0.218390528857708,
      "num_tokens": 116602342.0,
      "step": 46010
    },
    {
      "entropy": 6.0561624526977536,
      "epoch": 5.310444316214657,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002569538709785323,
      "loss": 5.4451,
      "mean_token_accuracy": 0.2190216898918152,
      "num_tokens": 116614737.0,
      "step": 46015
    },
    {
      "entropy": 5.9866249561309814,
      "epoch": 5.311021350259665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025691273986813137,
      "loss": 5.3572,
      "mean_token_accuracy": 0.22852537781000137,
      "num_tokens": 116626859.0,
      "step": 46020
    },
    {
      "entropy": 5.93863263130188,
      "epoch": 5.311598384304674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002568716093660863,
      "loss": 5.3209,
      "mean_token_accuracy": 0.22840829491615294,
      "num_tokens": 116640319.0,
      "step": 46025
    },
    {
      "entropy": 6.06455283164978,
      "epoch": 5.312175418349683,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002568304794737807,
      "loss": 5.4969,
      "mean_token_accuracy": 0.2117247074842453,
      "num_tokens": 116653293.0,
      "step": 46030
    },
    {
      "entropy": 6.054355430603027,
      "epoch": 5.312752452394691,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025678935019259784,
      "loss": 5.3965,
      "mean_token_accuracy": 0.22374956905841828,
      "num_tokens": 116664989.0,
      "step": 46035
    },
    {
      "entropy": 6.085732126235962,
      "epoch": 5.3133294864397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002567482215239213,
      "loss": 5.4923,
      "mean_token_accuracy": 0.22013434767723083,
      "num_tokens": 116677868.0,
      "step": 46040
    },
    {
      "entropy": 6.021775484085083,
      "epoch": 5.313906520484709,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025670709346913413,
      "loss": 5.4562,
      "mean_token_accuracy": 0.22161409854888917,
      "num_tokens": 116691234.0,
      "step": 46045
    },
    {
      "entropy": 6.082256603240967,
      "epoch": 5.314483554529717,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002566659660296198,
      "loss": 5.4822,
      "mean_token_accuracy": 0.21662663519382477,
      "num_tokens": 116703317.0,
      "step": 46050
    },
    {
      "entropy": 6.056056785583496,
      "epoch": 5.315060588574726,
      "grad_norm": 0.96875,
      "learning_rate": 0.00025662483920676154,
      "loss": 5.4602,
      "mean_token_accuracy": 0.21702508628368378,
      "num_tokens": 116716638.0,
      "step": 46055
    },
    {
      "entropy": 6.05317907333374,
      "epoch": 5.315637622619734,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002565837130019428,
      "loss": 5.4115,
      "mean_token_accuracy": 0.22450577318668366,
      "num_tokens": 116730501.0,
      "step": 46060
    },
    {
      "entropy": 5.972889614105225,
      "epoch": 5.3162146566647435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002565425874165467,
      "loss": 5.3748,
      "mean_token_accuracy": 0.22214292138814926,
      "num_tokens": 116743184.0,
      "step": 46065
    },
    {
      "entropy": 6.033910179138184,
      "epoch": 5.316791690709752,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025650146245195647,
      "loss": 5.4857,
      "mean_token_accuracy": 0.21719806641340256,
      "num_tokens": 116755592.0,
      "step": 46070
    },
    {
      "entropy": 5.98660569190979,
      "epoch": 5.31736872475476,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002564603381095554,
      "loss": 5.4535,
      "mean_token_accuracy": 0.2213363915681839,
      "num_tokens": 116768618.0,
      "step": 46075
    },
    {
      "entropy": 6.0146771430969235,
      "epoch": 5.31794575879977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025641921439072666,
      "loss": 5.4594,
      "mean_token_accuracy": 0.21577292084693908,
      "num_tokens": 116782063.0,
      "step": 46080
    },
    {
      "entropy": 6.14362735748291,
      "epoch": 5.318522792844778,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025637809129685336,
      "loss": 5.4821,
      "mean_token_accuracy": 0.21354314535856248,
      "num_tokens": 116793939.0,
      "step": 46085
    },
    {
      "entropy": 6.0629908561706545,
      "epoch": 5.319099826889786,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002563369688293187,
      "loss": 5.4142,
      "mean_token_accuracy": 0.22511038035154343,
      "num_tokens": 116806161.0,
      "step": 46090
    },
    {
      "entropy": 6.060305643081665,
      "epoch": 5.319676860934795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025629584698950585,
      "loss": 5.488,
      "mean_token_accuracy": 0.20859711170196532,
      "num_tokens": 116818365.0,
      "step": 46095
    },
    {
      "entropy": 6.097740030288696,
      "epoch": 5.320253894979804,
      "grad_norm": 1.015625,
      "learning_rate": 0.000256254725778798,
      "loss": 5.4562,
      "mean_token_accuracy": 0.21608344614505767,
      "num_tokens": 116832219.0,
      "step": 46100
    },
    {
      "entropy": 6.0411949634552,
      "epoch": 5.320830929024813,
      "grad_norm": 1.0625,
      "learning_rate": 0.000256213605198578,
      "loss": 5.3689,
      "mean_token_accuracy": 0.2298753410577774,
      "num_tokens": 116844899.0,
      "step": 46105
    },
    {
      "entropy": 6.077473545074463,
      "epoch": 5.321407963069821,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002561724852502291,
      "loss": 5.4568,
      "mean_token_accuracy": 0.22031740248203277,
      "num_tokens": 116856804.0,
      "step": 46110
    },
    {
      "entropy": 5.987808465957642,
      "epoch": 5.321984997114829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002561313659351342,
      "loss": 5.4077,
      "mean_token_accuracy": 0.22933365702629088,
      "num_tokens": 116870369.0,
      "step": 46115
    },
    {
      "entropy": 6.1456560611724855,
      "epoch": 5.322562031159839,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025609024725467664,
      "loss": 5.5458,
      "mean_token_accuracy": 0.21556393802165985,
      "num_tokens": 116883509.0,
      "step": 46120
    },
    {
      "entropy": 6.033176374435425,
      "epoch": 5.323139065204847,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002560491292102391,
      "loss": 5.389,
      "mean_token_accuracy": 0.23220531791448593,
      "num_tokens": 116897247.0,
      "step": 46125
    },
    {
      "entropy": 5.965729904174805,
      "epoch": 5.323716099249856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025600801180320463,
      "loss": 5.3454,
      "mean_token_accuracy": 0.23463781625032426,
      "num_tokens": 116910489.0,
      "step": 46130
    },
    {
      "entropy": 6.099097108840942,
      "epoch": 5.324293133294864,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002559668950349563,
      "loss": 5.4848,
      "mean_token_accuracy": 0.21454005241394042,
      "num_tokens": 116922257.0,
      "step": 46135
    },
    {
      "entropy": 6.0517219543457035,
      "epoch": 5.324870167339873,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025592577890687715,
      "loss": 5.4061,
      "mean_token_accuracy": 0.2247065007686615,
      "num_tokens": 116934634.0,
      "step": 46140
    },
    {
      "entropy": 5.985164785385132,
      "epoch": 5.325447201384882,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025588466342034975,
      "loss": 5.4136,
      "mean_token_accuracy": 0.21729756891727448,
      "num_tokens": 116947566.0,
      "step": 46145
    },
    {
      "entropy": 6.0302205085754395,
      "epoch": 5.32602423542989,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002558435485767572,
      "loss": 5.4068,
      "mean_token_accuracy": 0.22710861414670944,
      "num_tokens": 116960798.0,
      "step": 46150
    },
    {
      "entropy": 5.947607612609863,
      "epoch": 5.326601269474899,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002558024343774825,
      "loss": 5.2952,
      "mean_token_accuracy": 0.22909213453531266,
      "num_tokens": 116973916.0,
      "step": 46155
    },
    {
      "entropy": 6.048489618301391,
      "epoch": 5.327178303519908,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002557613208239083,
      "loss": 5.521,
      "mean_token_accuracy": 0.2122193083167076,
      "num_tokens": 116986681.0,
      "step": 46160
    },
    {
      "entropy": 6.003508710861206,
      "epoch": 5.327755337564916,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025572020791741763,
      "loss": 5.4508,
      "mean_token_accuracy": 0.22063050121068956,
      "num_tokens": 116999787.0,
      "step": 46165
    },
    {
      "entropy": 6.022827243804931,
      "epoch": 5.328332371609925,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002556790956593931,
      "loss": 5.4166,
      "mean_token_accuracy": 0.2288016691803932,
      "num_tokens": 117012542.0,
      "step": 46170
    },
    {
      "entropy": 5.9424480438232425,
      "epoch": 5.328909405654934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002556379840512176,
      "loss": 5.2842,
      "mean_token_accuracy": 0.2203609138727188,
      "num_tokens": 117024383.0,
      "step": 46175
    },
    {
      "entropy": 6.041703891754151,
      "epoch": 5.329486439699942,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002555968730942738,
      "loss": 5.4748,
      "mean_token_accuracy": 0.21999604105949402,
      "num_tokens": 117037289.0,
      "step": 46180
    },
    {
      "entropy": 6.0298840522766115,
      "epoch": 5.330063473744951,
      "grad_norm": 1.0,
      "learning_rate": 0.00025555576278994474,
      "loss": 5.4642,
      "mean_token_accuracy": 0.22020394802093507,
      "num_tokens": 117049692.0,
      "step": 46185
    },
    {
      "entropy": 6.006472730636597,
      "epoch": 5.330640507789959,
      "grad_norm": 0.96875,
      "learning_rate": 0.00025551465313961274,
      "loss": 5.34,
      "mean_token_accuracy": 0.23253780454397202,
      "num_tokens": 117063423.0,
      "step": 46190
    },
    {
      "entropy": 6.056939315795899,
      "epoch": 5.3312175418349685,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00025547354414466085,
      "loss": 5.4388,
      "mean_token_accuracy": 0.22065360695123673,
      "num_tokens": 117076329.0,
      "step": 46195
    },
    {
      "entropy": 6.016411352157593,
      "epoch": 5.331794575879977,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002554324358064715,
      "loss": 5.4071,
      "mean_token_accuracy": 0.2158508449792862,
      "num_tokens": 117088367.0,
      "step": 46200
    },
    {
      "entropy": 6.035438013076782,
      "epoch": 5.332371609924985,
      "grad_norm": 2.25,
      "learning_rate": 0.00025539132812642755,
      "loss": 5.4438,
      "mean_token_accuracy": 0.22043509036302567,
      "num_tokens": 117101134.0,
      "step": 46205
    },
    {
      "entropy": 6.026764965057373,
      "epoch": 5.332948643969994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00025535022110591147,
      "loss": 5.385,
      "mean_token_accuracy": 0.22913145571947097,
      "num_tokens": 117113539.0,
      "step": 46210
    },
    {
      "entropy": 6.050382852554321,
      "epoch": 5.333525678015003,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00025530911474630594,
      "loss": 5.5,
      "mean_token_accuracy": 0.21475670039653777,
      "num_tokens": 117125573.0,
      "step": 46215
    },
    {
      "entropy": 6.005392789840698,
      "epoch": 5.3341027120600115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025526800904899355,
      "loss": 5.376,
      "mean_token_accuracy": 0.21866468489170074,
      "num_tokens": 117138436.0,
      "step": 46220
    },
    {
      "entropy": 6.032098245620728,
      "epoch": 5.33467974610502,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000255226904015357,
      "loss": 5.4484,
      "mean_token_accuracy": 0.2152218922972679,
      "num_tokens": 117151282.0,
      "step": 46225
    },
    {
      "entropy": 6.079346179962158,
      "epoch": 5.335256780150029,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00025518579964677854,
      "loss": 5.4352,
      "mean_token_accuracy": 0.22209523022174835,
      "num_tokens": 117163434.0,
      "step": 46230
    },
    {
      "entropy": 5.990505599975586,
      "epoch": 5.335833814195038,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002551446959446409,
      "loss": 5.3685,
      "mean_token_accuracy": 0.22944666892290116,
      "num_tokens": 117176520.0,
      "step": 46235
    },
    {
      "entropy": 6.000926828384399,
      "epoch": 5.336410848240046,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002551035929103266,
      "loss": 5.4011,
      "mean_token_accuracy": 0.22566193491220474,
      "num_tokens": 117189525.0,
      "step": 46240
    },
    {
      "entropy": 6.082267141342163,
      "epoch": 5.3369878822850545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025506249054521807,
      "loss": 5.4259,
      "mean_token_accuracy": 0.2284848153591156,
      "num_tokens": 117203747.0,
      "step": 46245
    },
    {
      "entropy": 6.033232164382935,
      "epoch": 5.337564916330064,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025502138885069776,
      "loss": 5.4518,
      "mean_token_accuracy": 0.21333421319723128,
      "num_tokens": 117218016.0,
      "step": 46250
    },
    {
      "entropy": 6.078038740158081,
      "epoch": 5.338141950375072,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025498028782814806,
      "loss": 5.4762,
      "mean_token_accuracy": 0.2225479304790497,
      "num_tokens": 117229272.0,
      "step": 46255
    },
    {
      "entropy": 6.080319595336914,
      "epoch": 5.338718984420081,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025493918747895145,
      "loss": 5.4498,
      "mean_token_accuracy": 0.21851986348629,
      "num_tokens": 117241708.0,
      "step": 46260
    },
    {
      "entropy": 6.124478101730347,
      "epoch": 5.339296018465089,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025489808780449044,
      "loss": 5.4256,
      "mean_token_accuracy": 0.21361800730228425,
      "num_tokens": 117253438.0,
      "step": 46265
    },
    {
      "entropy": 6.04870982170105,
      "epoch": 5.339873052510098,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025485698880614707,
      "loss": 5.4593,
      "mean_token_accuracy": 0.21864764839410783,
      "num_tokens": 117267087.0,
      "step": 46270
    },
    {
      "entropy": 6.008804988861084,
      "epoch": 5.340450086555107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000254815890485304,
      "loss": 5.4423,
      "mean_token_accuracy": 0.2245807632803917,
      "num_tokens": 117280013.0,
      "step": 46275
    },
    {
      "entropy": 6.015140438079834,
      "epoch": 5.341027120600115,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002547747928433433,
      "loss": 5.4093,
      "mean_token_accuracy": 0.2231789782643318,
      "num_tokens": 117292634.0,
      "step": 46280
    },
    {
      "entropy": 6.0914304733276365,
      "epoch": 5.341604154645124,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002547336958816476,
      "loss": 5.4927,
      "mean_token_accuracy": 0.2136649966239929,
      "num_tokens": 117305570.0,
      "step": 46285
    },
    {
      "entropy": 5.982283258438111,
      "epoch": 5.342181188690133,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002546925996015988,
      "loss": 5.333,
      "mean_token_accuracy": 0.23774474412202834,
      "num_tokens": 117318991.0,
      "step": 46290
    },
    {
      "entropy": 5.968101930618286,
      "epoch": 5.342758222735141,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002546515040045794,
      "loss": 5.3121,
      "mean_token_accuracy": 0.22973464876413346,
      "num_tokens": 117331994.0,
      "step": 46295
    },
    {
      "entropy": 6.033121585845947,
      "epoch": 5.34333525678015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025461040909197146,
      "loss": 5.4616,
      "mean_token_accuracy": 0.221347576379776,
      "num_tokens": 117344735.0,
      "step": 46300
    },
    {
      "entropy": 6.062983894348145,
      "epoch": 5.343912290825159,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025456931486515745,
      "loss": 5.4874,
      "mean_token_accuracy": 0.21916218847036362,
      "num_tokens": 117357747.0,
      "step": 46305
    },
    {
      "entropy": 6.064911651611328,
      "epoch": 5.3444893248701675,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025452822132551923,
      "loss": 5.4573,
      "mean_token_accuracy": 0.21743026971817017,
      "num_tokens": 117369970.0,
      "step": 46310
    },
    {
      "entropy": 5.976342725753784,
      "epoch": 5.345066358915176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002544871284744392,
      "loss": 5.2765,
      "mean_token_accuracy": 0.23414543122053147,
      "num_tokens": 117381923.0,
      "step": 46315
    },
    {
      "entropy": 6.093671035766602,
      "epoch": 5.345643392960184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002544460363132994,
      "loss": 5.4771,
      "mean_token_accuracy": 0.21469964534044267,
      "num_tokens": 117393395.0,
      "step": 46320
    },
    {
      "entropy": 6.054122734069824,
      "epoch": 5.346220427005194,
      "grad_norm": 0.96875,
      "learning_rate": 0.000254404944843482,
      "loss": 5.4066,
      "mean_token_accuracy": 0.22262776792049407,
      "num_tokens": 117406017.0,
      "step": 46325
    },
    {
      "entropy": 6.048027229309082,
      "epoch": 5.346797461050202,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002543638540663689,
      "loss": 5.4554,
      "mean_token_accuracy": 0.21837498098611832,
      "num_tokens": 117419769.0,
      "step": 46330
    },
    {
      "entropy": 6.0547269821167,
      "epoch": 5.3473744950952105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025432276398334254,
      "loss": 5.3976,
      "mean_token_accuracy": 0.2258692130446434,
      "num_tokens": 117432050.0,
      "step": 46335
    },
    {
      "entropy": 6.002361392974853,
      "epoch": 5.347951529140219,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025428167459578453,
      "loss": 5.445,
      "mean_token_accuracy": 0.21855174750089645,
      "num_tokens": 117444447.0,
      "step": 46340
    },
    {
      "entropy": 6.043699312210083,
      "epoch": 5.348528563185228,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025424058590507725,
      "loss": 5.4072,
      "mean_token_accuracy": 0.21950049102306365,
      "num_tokens": 117457804.0,
      "step": 46345
    },
    {
      "entropy": 5.956660223007202,
      "epoch": 5.349105597230237,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025419949791260246,
      "loss": 5.3147,
      "mean_token_accuracy": 0.2253398835659027,
      "num_tokens": 117471731.0,
      "step": 46350
    },
    {
      "entropy": 6.07152624130249,
      "epoch": 5.349682631275245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002541584106197423,
      "loss": 5.4362,
      "mean_token_accuracy": 0.22245904058218002,
      "num_tokens": 117485290.0,
      "step": 46355
    },
    {
      "entropy": 6.024050855636597,
      "epoch": 5.350259665320253,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002541173240278786,
      "loss": 5.3514,
      "mean_token_accuracy": 0.22893767952919006,
      "num_tokens": 117496469.0,
      "step": 46360
    },
    {
      "entropy": 6.077806758880615,
      "epoch": 5.350836699365263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002540762381383934,
      "loss": 5.5237,
      "mean_token_accuracy": 0.21693225353956222,
      "num_tokens": 117507862.0,
      "step": 46365
    },
    {
      "entropy": 6.030594682693481,
      "epoch": 5.351413733410271,
      "grad_norm": 1.0,
      "learning_rate": 0.00025403515295266844,
      "loss": 5.4154,
      "mean_token_accuracy": 0.22298555076122284,
      "num_tokens": 117520208.0,
      "step": 46370
    },
    {
      "entropy": 5.948260831832886,
      "epoch": 5.35199076745528,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025399406847208576,
      "loss": 5.2917,
      "mean_token_accuracy": 0.23446203172206878,
      "num_tokens": 117532838.0,
      "step": 46375
    },
    {
      "entropy": 5.980932760238647,
      "epoch": 5.352567801500289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025395298469802705,
      "loss": 5.4071,
      "mean_token_accuracy": 0.22854914218187333,
      "num_tokens": 117545765.0,
      "step": 46380
    },
    {
      "entropy": 6.104554319381714,
      "epoch": 5.353144835545297,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025391190163187436,
      "loss": 5.5505,
      "mean_token_accuracy": 0.21074844747781754,
      "num_tokens": 117557844.0,
      "step": 46385
    },
    {
      "entropy": 6.007717180252075,
      "epoch": 5.353721869590306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025387081927500933,
      "loss": 5.42,
      "mean_token_accuracy": 0.2155975356698036,
      "num_tokens": 117569170.0,
      "step": 46390
    },
    {
      "entropy": 5.99181056022644,
      "epoch": 5.354298903635314,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002538297376288139,
      "loss": 5.4058,
      "mean_token_accuracy": 0.21757039278745652,
      "num_tokens": 117581344.0,
      "step": 46395
    },
    {
      "entropy": 6.142331933975219,
      "epoch": 5.354875937680323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002537886566946696,
      "loss": 5.464,
      "mean_token_accuracy": 0.21811507493257523,
      "num_tokens": 117593486.0,
      "step": 46400
    },
    {
      "entropy": 6.095344972610474,
      "epoch": 5.355452971725332,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002537475764739583,
      "loss": 5.4751,
      "mean_token_accuracy": 0.21605441719293594,
      "num_tokens": 117605233.0,
      "step": 46405
    },
    {
      "entropy": 5.996577167510987,
      "epoch": 5.35603000577034,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025370649696806173,
      "loss": 5.3766,
      "mean_token_accuracy": 0.2289374738931656,
      "num_tokens": 117618463.0,
      "step": 46410
    },
    {
      "entropy": 6.009560251235962,
      "epoch": 5.356607039815349,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002536654181783617,
      "loss": 5.4432,
      "mean_token_accuracy": 0.22371635288000108,
      "num_tokens": 117630325.0,
      "step": 46415
    },
    {
      "entropy": 6.171038103103638,
      "epoch": 5.357184073860358,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025362434010623954,
      "loss": 5.5263,
      "mean_token_accuracy": 0.21389693170785903,
      "num_tokens": 117642971.0,
      "step": 46420
    },
    {
      "entropy": 6.057511615753174,
      "epoch": 5.357761107905366,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025358326275307715,
      "loss": 5.4001,
      "mean_token_accuracy": 0.22465806305408478,
      "num_tokens": 117654995.0,
      "step": 46425
    },
    {
      "entropy": 6.073803091049195,
      "epoch": 5.358338141950375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000253542186120256,
      "loss": 5.508,
      "mean_token_accuracy": 0.2089098021388054,
      "num_tokens": 117667671.0,
      "step": 46430
    },
    {
      "entropy": 5.918203926086425,
      "epoch": 5.358915175995384,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00025350111020915793,
      "loss": 5.3106,
      "mean_token_accuracy": 0.23617036640644073,
      "num_tokens": 117681773.0,
      "step": 46435
    },
    {
      "entropy": 6.088286304473877,
      "epoch": 5.359492210040393,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025346003502116425,
      "loss": 5.4485,
      "mean_token_accuracy": 0.22191868871450424,
      "num_tokens": 117695520.0,
      "step": 46440
    },
    {
      "entropy": 6.1003193855285645,
      "epoch": 5.360069244085401,
      "grad_norm": 1.0,
      "learning_rate": 0.00025341896055765663,
      "loss": 5.4696,
      "mean_token_accuracy": 0.2207578808069229,
      "num_tokens": 117707748.0,
      "step": 46445
    },
    {
      "entropy": 6.039443922042847,
      "epoch": 5.360646278130409,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00025337788682001645,
      "loss": 5.4463,
      "mean_token_accuracy": 0.2190927892923355,
      "num_tokens": 117721638.0,
      "step": 46450
    },
    {
      "entropy": 6.018512916564942,
      "epoch": 5.361223312175419,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025333681380962543,
      "loss": 5.4174,
      "mean_token_accuracy": 0.22454916536808014,
      "num_tokens": 117733644.0,
      "step": 46455
    },
    {
      "entropy": 6.111059856414795,
      "epoch": 5.361800346220427,
      "grad_norm": 0.953125,
      "learning_rate": 0.00025329574152786483,
      "loss": 5.4739,
      "mean_token_accuracy": 0.21339171081781388,
      "num_tokens": 117746112.0,
      "step": 46460
    },
    {
      "entropy": 6.07945408821106,
      "epoch": 5.3623773802654355,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002532546699761162,
      "loss": 5.4759,
      "mean_token_accuracy": 0.21625843048095703,
      "num_tokens": 117758864.0,
      "step": 46465
    },
    {
      "entropy": 6.01701545715332,
      "epoch": 5.362954414310444,
      "grad_norm": 1.0,
      "learning_rate": 0.00025321359915576094,
      "loss": 5.4261,
      "mean_token_accuracy": 0.21516769379377365,
      "num_tokens": 117772185.0,
      "step": 46470
    },
    {
      "entropy": 6.083926105499268,
      "epoch": 5.363531448355453,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002531725290681805,
      "loss": 5.4459,
      "mean_token_accuracy": 0.22386471629142762,
      "num_tokens": 117784622.0,
      "step": 46475
    },
    {
      "entropy": 6.090250253677368,
      "epoch": 5.364108482400462,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00025313145971475614,
      "loss": 5.4557,
      "mean_token_accuracy": 0.21698042005300522,
      "num_tokens": 117797029.0,
      "step": 46480
    },
    {
      "entropy": 6.05199875831604,
      "epoch": 5.36468551644547,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025309039109686933,
      "loss": 5.4413,
      "mean_token_accuracy": 0.2256188228726387,
      "num_tokens": 117810291.0,
      "step": 46485
    },
    {
      "entropy": 6.020019292831421,
      "epoch": 5.365262550490479,
      "grad_norm": 1.0,
      "learning_rate": 0.00025304932321590124,
      "loss": 5.4245,
      "mean_token_accuracy": 0.22318026423454285,
      "num_tokens": 117823196.0,
      "step": 46490
    },
    {
      "entropy": 6.064975929260254,
      "epoch": 5.365839584535488,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002530082560732334,
      "loss": 5.4612,
      "mean_token_accuracy": 0.2136770263314247,
      "num_tokens": 117836234.0,
      "step": 46495
    },
    {
      "entropy": 6.034855079650879,
      "epoch": 5.366416618580496,
      "grad_norm": 1.125,
      "learning_rate": 0.00025296718967024684,
      "loss": 5.4701,
      "mean_token_accuracy": 0.22009082287549972,
      "num_tokens": 117848465.0,
      "step": 46500
    },
    {
      "entropy": 6.104257202148437,
      "epoch": 5.366993652625505,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002529261240083229,
      "loss": 5.4963,
      "mean_token_accuracy": 0.2233443319797516,
      "num_tokens": 117860945.0,
      "step": 46505
    },
    {
      "entropy": 6.065873718261718,
      "epoch": 5.367570686670514,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00025288505908884287,
      "loss": 5.5067,
      "mean_token_accuracy": 0.22102196961641313,
      "num_tokens": 117874283.0,
      "step": 46510
    },
    {
      "entropy": 6.017688369750976,
      "epoch": 5.368147720715522,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002528439949131879,
      "loss": 5.43,
      "mean_token_accuracy": 0.22176415026187896,
      "num_tokens": 117886707.0,
      "step": 46515
    },
    {
      "entropy": 6.039252519607544,
      "epoch": 5.368724754760531,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00025280293148273914,
      "loss": 5.3989,
      "mean_token_accuracy": 0.23084511160850524,
      "num_tokens": 117899810.0,
      "step": 46520
    },
    {
      "entropy": 5.997358179092407,
      "epoch": 5.369301788805539,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002527618687988777,
      "loss": 5.3798,
      "mean_token_accuracy": 0.22086113542318345,
      "num_tokens": 117912607.0,
      "step": 46525
    },
    {
      "entropy": 6.046096277236939,
      "epoch": 5.3698788228505485,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00025272080686298474,
      "loss": 5.4071,
      "mean_token_accuracy": 0.22432028353214264,
      "num_tokens": 117927038.0,
      "step": 46530
    },
    {
      "entropy": 6.028098678588867,
      "epoch": 5.370455856895557,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002526797456764415,
      "loss": 5.4021,
      "mean_token_accuracy": 0.22417077720165252,
      "num_tokens": 117939595.0,
      "step": 46535
    },
    {
      "entropy": 6.004332208633423,
      "epoch": 5.371032890940565,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025263868524062885,
      "loss": 5.4281,
      "mean_token_accuracy": 0.23470542579889297,
      "num_tokens": 117952327.0,
      "step": 46540
    },
    {
      "entropy": 5.911421775817871,
      "epoch": 5.371609924985574,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025259762555692795,
      "loss": 5.3212,
      "mean_token_accuracy": 0.23034425526857377,
      "num_tokens": 117964619.0,
      "step": 46545
    },
    {
      "entropy": 6.098479843139648,
      "epoch": 5.372186959030583,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025255656662671975,
      "loss": 5.5728,
      "mean_token_accuracy": 0.21262627542018891,
      "num_tokens": 117978489.0,
      "step": 46550
    },
    {
      "entropy": 6.07009220123291,
      "epoch": 5.3727639930755915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002525155084513852,
      "loss": 5.4331,
      "mean_token_accuracy": 0.21783585846424103,
      "num_tokens": 117993539.0,
      "step": 46555
    },
    {
      "entropy": 6.074240064620971,
      "epoch": 5.3733410271206,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00025247445103230556,
      "loss": 5.4358,
      "mean_token_accuracy": 0.22158272415399552,
      "num_tokens": 118006521.0,
      "step": 46560
    },
    {
      "entropy": 6.0861289501190186,
      "epoch": 5.373918061165609,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025243339437086137,
      "loss": 5.4578,
      "mean_token_accuracy": 0.21503166407346724,
      "num_tokens": 118019237.0,
      "step": 46565
    },
    {
      "entropy": 6.100465202331543,
      "epoch": 5.374495095210618,
      "grad_norm": 0.9375,
      "learning_rate": 0.00025239233846843383,
      "loss": 5.4623,
      "mean_token_accuracy": 0.219967320561409,
      "num_tokens": 118032020.0,
      "step": 46570
    },
    {
      "entropy": 6.11424674987793,
      "epoch": 5.375072129255626,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002523512833264038,
      "loss": 5.4974,
      "mean_token_accuracy": 0.2167445167899132,
      "num_tokens": 118044317.0,
      "step": 46575
    },
    {
      "entropy": 5.948747444152832,
      "epoch": 5.3756491633006345,
      "grad_norm": 0.90625,
      "learning_rate": 0.00025231022894615213,
      "loss": 5.3687,
      "mean_token_accuracy": 0.23021965324878693,
      "num_tokens": 118058537.0,
      "step": 46580
    },
    {
      "entropy": 5.885046672821045,
      "epoch": 5.376226197345644,
      "grad_norm": 1.0,
      "learning_rate": 0.00025226917532905955,
      "loss": 5.3064,
      "mean_token_accuracy": 0.22981446981430054,
      "num_tokens": 118072638.0,
      "step": 46585
    },
    {
      "entropy": 5.983325052261352,
      "epoch": 5.376803231390652,
      "grad_norm": 1.0,
      "learning_rate": 0.000252228122476507,
      "loss": 5.4153,
      "mean_token_accuracy": 0.22390871942043306,
      "num_tokens": 118086200.0,
      "step": 46590
    },
    {
      "entropy": 6.058635711669922,
      "epoch": 5.377380265435661,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00025218707038987515,
      "loss": 5.4038,
      "mean_token_accuracy": 0.22339641451835632,
      "num_tokens": 118099101.0,
      "step": 46595
    },
    {
      "entropy": 6.10756287574768,
      "epoch": 5.377957299480669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000252146019070545,
      "loss": 5.5342,
      "mean_token_accuracy": 0.21020758002996445,
      "num_tokens": 118111724.0,
      "step": 46600
    },
    {
      "entropy": 6.062566757202148,
      "epoch": 5.378534333525678,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025210496851989705,
      "loss": 5.5042,
      "mean_token_accuracy": 0.20720487385988234,
      "num_tokens": 118125034.0,
      "step": 46605
    },
    {
      "entropy": 5.987087631225586,
      "epoch": 5.379111367570687,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002520639187393122,
      "loss": 5.3618,
      "mean_token_accuracy": 0.2269548088312149,
      "num_tokens": 118138460.0,
      "step": 46610
    },
    {
      "entropy": 6.085792970657349,
      "epoch": 5.379688401615695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000252022869730171,
      "loss": 5.4327,
      "mean_token_accuracy": 0.2233018696308136,
      "num_tokens": 118150687.0,
      "step": 46615
    },
    {
      "entropy": 6.003587532043457,
      "epoch": 5.380265435660704,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002519818214938542,
      "loss": 5.3994,
      "mean_token_accuracy": 0.2293952986598015,
      "num_tokens": 118162794.0,
      "step": 46620
    },
    {
      "entropy": 5.954463100433349,
      "epoch": 5.380842469705713,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002519407740317423,
      "loss": 5.3785,
      "mean_token_accuracy": 0.22495761513710022,
      "num_tokens": 118174744.0,
      "step": 46625
    },
    {
      "entropy": 5.950168943405151,
      "epoch": 5.381419503750721,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002518997273452161,
      "loss": 5.3261,
      "mean_token_accuracy": 0.2330923080444336,
      "num_tokens": 118186543.0,
      "step": 46630
    },
    {
      "entropy": 6.031841373443603,
      "epoch": 5.38199653779573,
      "grad_norm": 0.96875,
      "learning_rate": 0.000251858681435656,
      "loss": 5.3508,
      "mean_token_accuracy": 0.2257835417985916,
      "num_tokens": 118198801.0,
      "step": 46635
    },
    {
      "entropy": 6.073264455795288,
      "epoch": 5.382573571840739,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025181763630444283,
      "loss": 5.4428,
      "mean_token_accuracy": 0.22036836296319962,
      "num_tokens": 118212010.0,
      "step": 46640
    },
    {
      "entropy": 6.069288158416748,
      "epoch": 5.3831506058857475,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025177659195295675,
      "loss": 5.4225,
      "mean_token_accuracy": 0.22528850585222243,
      "num_tokens": 118224568.0,
      "step": 46645
    },
    {
      "entropy": 6.032807064056397,
      "epoch": 5.383727639930756,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00025173554838257855,
      "loss": 5.4433,
      "mean_token_accuracy": 0.2188011109828949,
      "num_tokens": 118236575.0,
      "step": 46650
    },
    {
      "entropy": 6.136781072616577,
      "epoch": 5.384304673975764,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002516945055946886,
      "loss": 5.4979,
      "mean_token_accuracy": 0.21484594643115998,
      "num_tokens": 118249523.0,
      "step": 46655
    },
    {
      "entropy": 6.082072019577026,
      "epoch": 5.384881708020774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002516534635906675,
      "loss": 5.4515,
      "mean_token_accuracy": 0.2174357905983925,
      "num_tokens": 118261440.0,
      "step": 46660
    },
    {
      "entropy": 5.928796100616455,
      "epoch": 5.385458742065782,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002516124223718954,
      "loss": 5.4253,
      "mean_token_accuracy": 0.22346831113100052,
      "num_tokens": 118275352.0,
      "step": 46665
    },
    {
      "entropy": 6.01789927482605,
      "epoch": 5.38603577611079,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002515713819397529,
      "loss": 5.4172,
      "mean_token_accuracy": 0.2251511797308922,
      "num_tokens": 118289355.0,
      "step": 46670
    },
    {
      "entropy": 6.146826171875,
      "epoch": 5.386612810155799,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025153034229562037,
      "loss": 5.4533,
      "mean_token_accuracy": 0.21645608693361282,
      "num_tokens": 118302769.0,
      "step": 46675
    },
    {
      "entropy": 5.979436254501342,
      "epoch": 5.387189844200808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002514893034408782,
      "loss": 5.3486,
      "mean_token_accuracy": 0.23287110179662704,
      "num_tokens": 118313908.0,
      "step": 46680
    },
    {
      "entropy": 6.010890007019043,
      "epoch": 5.387766878245817,
      "grad_norm": 0.90625,
      "learning_rate": 0.00025144826537690647,
      "loss": 5.514,
      "mean_token_accuracy": 0.21533904522657393,
      "num_tokens": 118328604.0,
      "step": 46685
    },
    {
      "entropy": 6.003427124023437,
      "epoch": 5.388343912290825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002514072281050857,
      "loss": 5.3631,
      "mean_token_accuracy": 0.2217478260397911,
      "num_tokens": 118340876.0,
      "step": 46690
    },
    {
      "entropy": 6.031282997131347,
      "epoch": 5.388920946335833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000251366191626796,
      "loss": 5.4147,
      "mean_token_accuracy": 0.2243668869137764,
      "num_tokens": 118353746.0,
      "step": 46695
    },
    {
      "entropy": 6.08333158493042,
      "epoch": 5.389497980380843,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002513251559434179,
      "loss": 5.4723,
      "mean_token_accuracy": 0.22129524052143096,
      "num_tokens": 118367459.0,
      "step": 46700
    },
    {
      "entropy": 5.935800457000733,
      "epoch": 5.390075014425851,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002512841210563313,
      "loss": 5.3079,
      "mean_token_accuracy": 0.23677830398082733,
      "num_tokens": 118380481.0,
      "step": 46705
    },
    {
      "entropy": 6.065203666687012,
      "epoch": 5.3906520484708595,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025124308696691653,
      "loss": 5.4283,
      "mean_token_accuracy": 0.22849969565868378,
      "num_tokens": 118393255.0,
      "step": 46710
    },
    {
      "entropy": 5.990964412689209,
      "epoch": 5.391229082515869,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002512020536765536,
      "loss": 5.4001,
      "mean_token_accuracy": 0.2225814312696457,
      "num_tokens": 118406148.0,
      "step": 46715
    },
    {
      "entropy": 6.050089931488037,
      "epoch": 5.391806116560877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025116102118662306,
      "loss": 5.4161,
      "mean_token_accuracy": 0.21961817592382432,
      "num_tokens": 118417775.0,
      "step": 46720
    },
    {
      "entropy": 6.013803005218506,
      "epoch": 5.392383150605886,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025111998949850446,
      "loss": 5.4336,
      "mean_token_accuracy": 0.2194457858800888,
      "num_tokens": 118430760.0,
      "step": 46725
    },
    {
      "entropy": 6.02965669631958,
      "epoch": 5.392960184650894,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025107895861357834,
      "loss": 5.4203,
      "mean_token_accuracy": 0.216534985601902,
      "num_tokens": 118443010.0,
      "step": 46730
    },
    {
      "entropy": 6.018858575820923,
      "epoch": 5.393537218695903,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025103792853322445,
      "loss": 5.3264,
      "mean_token_accuracy": 0.22580075562000274,
      "num_tokens": 118454712.0,
      "step": 46735
    },
    {
      "entropy": 5.959798383712768,
      "epoch": 5.394114252740912,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025099689925882297,
      "loss": 5.3754,
      "mean_token_accuracy": 0.22562704086303711,
      "num_tokens": 118468362.0,
      "step": 46740
    },
    {
      "entropy": 6.018317937850952,
      "epoch": 5.39469128678592,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002509558707917538,
      "loss": 5.438,
      "mean_token_accuracy": 0.22785538882017137,
      "num_tokens": 118481015.0,
      "step": 46745
    },
    {
      "entropy": 6.055947971343994,
      "epoch": 5.395268320830929,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025091484313339706,
      "loss": 5.4383,
      "mean_token_accuracy": 0.22215741276741027,
      "num_tokens": 118493843.0,
      "step": 46750
    },
    {
      "entropy": 6.095442914962769,
      "epoch": 5.395845354875938,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025087381628513257,
      "loss": 5.5172,
      "mean_token_accuracy": 0.21285209506750108,
      "num_tokens": 118506635.0,
      "step": 46755
    },
    {
      "entropy": 6.137464237213135,
      "epoch": 5.396422388920946,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025083279024834033,
      "loss": 5.4979,
      "mean_token_accuracy": 0.2162633016705513,
      "num_tokens": 118518767.0,
      "step": 46760
    },
    {
      "entropy": 6.085399866104126,
      "epoch": 5.396999422965955,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002507917650244001,
      "loss": 5.4318,
      "mean_token_accuracy": 0.219040122628212,
      "num_tokens": 118531347.0,
      "step": 46765
    },
    {
      "entropy": 6.053800201416015,
      "epoch": 5.397576457010963,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025075074061469197,
      "loss": 5.4698,
      "mean_token_accuracy": 0.21922803223133086,
      "num_tokens": 118545076.0,
      "step": 46770
    },
    {
      "entropy": 6.042861080169677,
      "epoch": 5.3981534910559725,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025070971702059546,
      "loss": 5.4149,
      "mean_token_accuracy": 0.2174384132027626,
      "num_tokens": 118557978.0,
      "step": 46775
    },
    {
      "entropy": 6.042640209197998,
      "epoch": 5.398730525100981,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025066869424349066,
      "loss": 5.4293,
      "mean_token_accuracy": 0.22148347347974778,
      "num_tokens": 118569763.0,
      "step": 46780
    },
    {
      "entropy": 6.07701416015625,
      "epoch": 5.399307559145989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002506276722847571,
      "loss": 5.3932,
      "mean_token_accuracy": 0.21393050849437714,
      "num_tokens": 118583189.0,
      "step": 46785
    },
    {
      "entropy": 6.080574226379395,
      "epoch": 5.399884593190999,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002505866511457749,
      "loss": 5.5012,
      "mean_token_accuracy": 0.21454430967569352,
      "num_tokens": 118595532.0,
      "step": 46790
    },
    {
      "entropy": 5.862489509582519,
      "epoch": 5.400461627236007,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025054563082792345,
      "loss": 5.2306,
      "mean_token_accuracy": 0.24630296677351,
      "num_tokens": 118607534.0,
      "step": 46795
    },
    {
      "entropy": 6.084830188751221,
      "epoch": 5.4010386612810155,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002505046113325825,
      "loss": 5.4613,
      "mean_token_accuracy": 0.21699014753103257,
      "num_tokens": 118619264.0,
      "step": 46800
    },
    {
      "entropy": 6.0267627239227295,
      "epoch": 5.401615695326024,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00025046359266113184,
      "loss": 5.4466,
      "mean_token_accuracy": 0.22269827127456665,
      "num_tokens": 118632341.0,
      "step": 46805
    },
    {
      "entropy": 6.027035236358643,
      "epoch": 5.402192729371033,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002504225748149511,
      "loss": 5.467,
      "mean_token_accuracy": 0.22000111639499664,
      "num_tokens": 118645537.0,
      "step": 46810
    },
    {
      "entropy": 6.041680765151978,
      "epoch": 5.402769763416042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025038155779541976,
      "loss": 5.3922,
      "mean_token_accuracy": 0.22586773037910463,
      "num_tokens": 118658031.0,
      "step": 46815
    },
    {
      "entropy": 6.017295837402344,
      "epoch": 5.40334679746105,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002503405416039176,
      "loss": 5.4335,
      "mean_token_accuracy": 0.21777289360761642,
      "num_tokens": 118671521.0,
      "step": 46820
    },
    {
      "entropy": 6.067204761505127,
      "epoch": 5.4039238315060585,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00025029952624182393,
      "loss": 5.4342,
      "mean_token_accuracy": 0.22069837152957916,
      "num_tokens": 118685069.0,
      "step": 46825
    },
    {
      "entropy": 6.059438848495484,
      "epoch": 5.404500865551068,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002502585117105186,
      "loss": 5.3808,
      "mean_token_accuracy": 0.22172050029039383,
      "num_tokens": 118697423.0,
      "step": 46830
    },
    {
      "entropy": 5.96149320602417,
      "epoch": 5.405077899596076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002502174980113808,
      "loss": 5.3576,
      "mean_token_accuracy": 0.2294830098748207,
      "num_tokens": 118710970.0,
      "step": 46835
    },
    {
      "entropy": 6.01069450378418,
      "epoch": 5.405654933641085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025017648514579014,
      "loss": 5.395,
      "mean_token_accuracy": 0.22226380854845046,
      "num_tokens": 118723915.0,
      "step": 46840
    },
    {
      "entropy": 5.9598448276519775,
      "epoch": 5.406231967686093,
      "grad_norm": 0.984375,
      "learning_rate": 0.00025013547311512614,
      "loss": 5.3552,
      "mean_token_accuracy": 0.23078590780496597,
      "num_tokens": 118736917.0,
      "step": 46845
    },
    {
      "entropy": 6.065068340301513,
      "epoch": 5.406809001731102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002500944619207682,
      "loss": 5.4971,
      "mean_token_accuracy": 0.2155863255262375,
      "num_tokens": 118748420.0,
      "step": 46850
    },
    {
      "entropy": 5.963948822021484,
      "epoch": 5.407386035776111,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025005345156409555,
      "loss": 5.3579,
      "mean_token_accuracy": 0.2259705826640129,
      "num_tokens": 118760388.0,
      "step": 46855
    },
    {
      "entropy": 5.962028932571411,
      "epoch": 5.407963069821119,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002500124420464877,
      "loss": 5.3629,
      "mean_token_accuracy": 0.22914329767227173,
      "num_tokens": 118773317.0,
      "step": 46860
    },
    {
      "entropy": 6.106974172592163,
      "epoch": 5.4085401038661285,
      "grad_norm": 1.125,
      "learning_rate": 0.0002499714333693239,
      "loss": 5.5269,
      "mean_token_accuracy": 0.2157278537750244,
      "num_tokens": 118786435.0,
      "step": 46865
    },
    {
      "entropy": 6.147992467880249,
      "epoch": 5.409117137911137,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024993042553398363,
      "loss": 5.5075,
      "mean_token_accuracy": 0.21184257715940474,
      "num_tokens": 118798567.0,
      "step": 46870
    },
    {
      "entropy": 5.9270271301269535,
      "epoch": 5.409694171956145,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002498894185418459,
      "loss": 5.2974,
      "mean_token_accuracy": 0.2363997533917427,
      "num_tokens": 118811199.0,
      "step": 46875
    },
    {
      "entropy": 5.978492498397827,
      "epoch": 5.410271206001154,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024984841239429026,
      "loss": 5.4237,
      "mean_token_accuracy": 0.2273360162973404,
      "num_tokens": 118824544.0,
      "step": 46880
    },
    {
      "entropy": 6.010010290145874,
      "epoch": 5.410848240046163,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002498074070926956,
      "loss": 5.4176,
      "mean_token_accuracy": 0.2197552815079689,
      "num_tokens": 118838687.0,
      "step": 46885
    },
    {
      "entropy": 6.03066782951355,
      "epoch": 5.4114252740911715,
      "grad_norm": 1.125,
      "learning_rate": 0.00024976640263844153,
      "loss": 5.3977,
      "mean_token_accuracy": 0.2246936932206154,
      "num_tokens": 118851122.0,
      "step": 46890
    },
    {
      "entropy": 5.975719356536866,
      "epoch": 5.41200230813618,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024972539903290677,
      "loss": 5.3595,
      "mean_token_accuracy": 0.22878360748291016,
      "num_tokens": 118862870.0,
      "step": 46895
    },
    {
      "entropy": 5.945103025436401,
      "epoch": 5.412579342181188,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002496843962774707,
      "loss": 5.3014,
      "mean_token_accuracy": 0.2252539098262787,
      "num_tokens": 118875085.0,
      "step": 46900
    },
    {
      "entropy": 5.874655103683471,
      "epoch": 5.413156376226198,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00024964339437351233,
      "loss": 5.2425,
      "mean_token_accuracy": 0.23313433825969695,
      "num_tokens": 118887638.0,
      "step": 46905
    },
    {
      "entropy": 5.942875909805298,
      "epoch": 5.413733410271206,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024960239332241096,
      "loss": 5.321,
      "mean_token_accuracy": 0.22468236982822418,
      "num_tokens": 118901156.0,
      "step": 46910
    },
    {
      "entropy": 6.001813459396362,
      "epoch": 5.414310444316214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002495613931255454,
      "loss": 5.3554,
      "mean_token_accuracy": 0.23009416908025743,
      "num_tokens": 118913570.0,
      "step": 46915
    },
    {
      "entropy": 6.005325078964233,
      "epoch": 5.414887478361223,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024952039378429476,
      "loss": 5.4258,
      "mean_token_accuracy": 0.22464391142129897,
      "num_tokens": 118926462.0,
      "step": 46920
    },
    {
      "entropy": 6.036478805541992,
      "epoch": 5.415464512406232,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002494793953000379,
      "loss": 5.4351,
      "mean_token_accuracy": 0.22260603159666062,
      "num_tokens": 118938457.0,
      "step": 46925
    },
    {
      "entropy": 5.873511743545532,
      "epoch": 5.416041546451241,
      "grad_norm": 1.140625,
      "learning_rate": 0.00024943839767415407,
      "loss": 5.2642,
      "mean_token_accuracy": 0.2437032252550125,
      "num_tokens": 118951256.0,
      "step": 46930
    },
    {
      "entropy": 5.981058263778687,
      "epoch": 5.416618580496249,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024939740090802195,
      "loss": 5.3289,
      "mean_token_accuracy": 0.23051441758871077,
      "num_tokens": 118964416.0,
      "step": 46935
    },
    {
      "entropy": 6.091514682769775,
      "epoch": 5.417195614541258,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00024935640500302054,
      "loss": 5.4645,
      "mean_token_accuracy": 0.2147015541791916,
      "num_tokens": 118977723.0,
      "step": 46940
    },
    {
      "entropy": 6.041667222976685,
      "epoch": 5.417772648586267,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024931540996052864,
      "loss": 5.4267,
      "mean_token_accuracy": 0.21578050553798675,
      "num_tokens": 118990050.0,
      "step": 46945
    },
    {
      "entropy": 6.0305774211883545,
      "epoch": 5.418349682631275,
      "grad_norm": 1.0,
      "learning_rate": 0.00024927441578192523,
      "loss": 5.4392,
      "mean_token_accuracy": 0.2179943487048149,
      "num_tokens": 119002388.0,
      "step": 46950
    },
    {
      "entropy": 6.064316082000732,
      "epoch": 5.418926716676284,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024923342246858915,
      "loss": 5.4766,
      "mean_token_accuracy": 0.21512299329042434,
      "num_tokens": 119013058.0,
      "step": 46955
    },
    {
      "entropy": 6.063048267364502,
      "epoch": 5.419503750721293,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024919243002189897,
      "loss": 5.5153,
      "mean_token_accuracy": 0.2197794571518898,
      "num_tokens": 119024280.0,
      "step": 46960
    },
    {
      "entropy": 6.019555902481079,
      "epoch": 5.420080784766301,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002491514384432337,
      "loss": 5.4272,
      "mean_token_accuracy": 0.22230074852705,
      "num_tokens": 119037219.0,
      "step": 46965
    },
    {
      "entropy": 6.047335577011109,
      "epoch": 5.42065781881131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024911044773397184,
      "loss": 5.4507,
      "mean_token_accuracy": 0.21714093387126923,
      "num_tokens": 119050924.0,
      "step": 46970
    },
    {
      "entropy": 6.083111000061035,
      "epoch": 5.421234852856318,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024906945789549234,
      "loss": 5.4488,
      "mean_token_accuracy": 0.22303324788808823,
      "num_tokens": 119063496.0,
      "step": 46975
    },
    {
      "entropy": 5.966890525817871,
      "epoch": 5.421811886901327,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024902846892917364,
      "loss": 5.3495,
      "mean_token_accuracy": 0.21838210970163346,
      "num_tokens": 119076205.0,
      "step": 46980
    },
    {
      "entropy": 6.060811519622803,
      "epoch": 5.422388920946336,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00024898748083639457,
      "loss": 5.4442,
      "mean_token_accuracy": 0.2119682490825653,
      "num_tokens": 119087892.0,
      "step": 46985
    },
    {
      "entropy": 6.006797504425049,
      "epoch": 5.422965954991344,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0002489464936185336,
      "loss": 5.3869,
      "mean_token_accuracy": 0.22254968285560608,
      "num_tokens": 119100533.0,
      "step": 46990
    },
    {
      "entropy": 6.032630777359008,
      "epoch": 5.423542989036353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002489055072769694,
      "loss": 5.4273,
      "mean_token_accuracy": 0.22533710896968842,
      "num_tokens": 119112901.0,
      "step": 46995
    },
    {
      "entropy": 6.030189371109008,
      "epoch": 5.424120023081362,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024886452181308044,
      "loss": 5.378,
      "mean_token_accuracy": 0.22474608123302459,
      "num_tokens": 119125468.0,
      "step": 47000
    },
    {
      "entropy": 6.04746446609497,
      "epoch": 5.42469705712637,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002488235372282454,
      "loss": 5.4246,
      "mean_token_accuracy": 0.2229582354426384,
      "num_tokens": 119137986.0,
      "step": 47005
    },
    {
      "entropy": 6.011452722549438,
      "epoch": 5.425274091171379,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002487825535238426,
      "loss": 5.3885,
      "mean_token_accuracy": 0.22094323486089706,
      "num_tokens": 119150889.0,
      "step": 47010
    },
    {
      "entropy": 5.966135263442993,
      "epoch": 5.425851125216388,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002487415707012507,
      "loss": 5.3353,
      "mean_token_accuracy": 0.22834445238113404,
      "num_tokens": 119162923.0,
      "step": 47015
    },
    {
      "entropy": 6.033320713043213,
      "epoch": 5.4264281592613965,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002487005887618479,
      "loss": 5.4682,
      "mean_token_accuracy": 0.22073763906955718,
      "num_tokens": 119175793.0,
      "step": 47020
    },
    {
      "entropy": 6.045900058746338,
      "epoch": 5.427005193306405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00024865960770701284,
      "loss": 5.4847,
      "mean_token_accuracy": 0.21898182332515717,
      "num_tokens": 119188177.0,
      "step": 47025
    },
    {
      "entropy": 6.0004877090454105,
      "epoch": 5.427582227351413,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002486186275381237,
      "loss": 5.3371,
      "mean_token_accuracy": 0.23148736357688904,
      "num_tokens": 119200928.0,
      "step": 47030
    },
    {
      "entropy": 6.0511674880981445,
      "epoch": 5.428159261396423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024857764825655904,
      "loss": 5.4538,
      "mean_token_accuracy": 0.21596154272556306,
      "num_tokens": 119212585.0,
      "step": 47035
    },
    {
      "entropy": 6.014525270462036,
      "epoch": 5.428736295441431,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024853666986369695,
      "loss": 5.3913,
      "mean_token_accuracy": 0.22385192662477493,
      "num_tokens": 119224794.0,
      "step": 47040
    },
    {
      "entropy": 6.081261444091797,
      "epoch": 5.4293133294864395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024849569236091583,
      "loss": 5.4042,
      "mean_token_accuracy": 0.2265527844429016,
      "num_tokens": 119237606.0,
      "step": 47045
    },
    {
      "entropy": 5.873202133178711,
      "epoch": 5.429890363531449,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024845471574959393,
      "loss": 5.2988,
      "mean_token_accuracy": 0.2256266549229622,
      "num_tokens": 119250883.0,
      "step": 47050
    },
    {
      "entropy": 5.921171760559082,
      "epoch": 5.430467397576457,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024841374003110953,
      "loss": 5.3223,
      "mean_token_accuracy": 0.23034353107213973,
      "num_tokens": 119263044.0,
      "step": 47055
    },
    {
      "entropy": 5.994158124923706,
      "epoch": 5.431044431621466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002483727652068408,
      "loss": 5.3592,
      "mean_token_accuracy": 0.23060242831707,
      "num_tokens": 119275504.0,
      "step": 47060
    },
    {
      "entropy": 6.015261316299439,
      "epoch": 5.431621465666474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002483317912781658,
      "loss": 5.3986,
      "mean_token_accuracy": 0.22417777627706528,
      "num_tokens": 119288144.0,
      "step": 47065
    },
    {
      "entropy": 6.042486000061035,
      "epoch": 5.432198499711483,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024829081824646274,
      "loss": 5.3955,
      "mean_token_accuracy": 0.2259470283985138,
      "num_tokens": 119301897.0,
      "step": 47070
    },
    {
      "entropy": 5.942798328399658,
      "epoch": 5.432775533756492,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024824984611310983,
      "loss": 5.3177,
      "mean_token_accuracy": 0.2392435312271118,
      "num_tokens": 119315401.0,
      "step": 47075
    },
    {
      "entropy": 6.0097898006439205,
      "epoch": 5.4333525678015,
      "grad_norm": 1.109375,
      "learning_rate": 0.000248208874879485,
      "loss": 5.4164,
      "mean_token_accuracy": 0.21938502490520478,
      "num_tokens": 119328329.0,
      "step": 47080
    },
    {
      "entropy": 6.084555006027221,
      "epoch": 5.433929601846509,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024816790454696645,
      "loss": 5.5007,
      "mean_token_accuracy": 0.21589263379573823,
      "num_tokens": 119341056.0,
      "step": 47085
    },
    {
      "entropy": 6.096207046508789,
      "epoch": 5.434506635891518,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024812693511693196,
      "loss": 5.517,
      "mean_token_accuracy": 0.21363582462072372,
      "num_tokens": 119351902.0,
      "step": 47090
    },
    {
      "entropy": 6.0990509510040285,
      "epoch": 5.435083669936526,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024808596659075985,
      "loss": 5.526,
      "mean_token_accuracy": 0.2152347519993782,
      "num_tokens": 119365129.0,
      "step": 47095
    },
    {
      "entropy": 6.109220790863037,
      "epoch": 5.435660703981535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002480449989698277,
      "loss": 5.5087,
      "mean_token_accuracy": 0.21715941578149794,
      "num_tokens": 119376877.0,
      "step": 47100
    },
    {
      "entropy": 6.0271636009216305,
      "epoch": 5.436237738026543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002480040322555137,
      "loss": 5.4222,
      "mean_token_accuracy": 0.2216916099190712,
      "num_tokens": 119388797.0,
      "step": 47105
    },
    {
      "entropy": 6.0596565246582035,
      "epoch": 5.4368147720715525,
      "grad_norm": 1.0,
      "learning_rate": 0.0002479630664491956,
      "loss": 5.4518,
      "mean_token_accuracy": 0.2175812765955925,
      "num_tokens": 119400892.0,
      "step": 47110
    },
    {
      "entropy": 6.068880605697632,
      "epoch": 5.437391806116561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024792210155225145,
      "loss": 5.4357,
      "mean_token_accuracy": 0.21811304092407227,
      "num_tokens": 119412486.0,
      "step": 47115
    },
    {
      "entropy": 6.034269189834594,
      "epoch": 5.437968840161569,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024788113756605884,
      "loss": 5.4436,
      "mean_token_accuracy": 0.2227180063724518,
      "num_tokens": 119425878.0,
      "step": 47120
    },
    {
      "entropy": 5.984864950180054,
      "epoch": 5.438545874206579,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002478401744919958,
      "loss": 5.3774,
      "mean_token_accuracy": 0.2204645276069641,
      "num_tokens": 119438644.0,
      "step": 47125
    },
    {
      "entropy": 6.059547328948975,
      "epoch": 5.439122908251587,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002477992123314399,
      "loss": 5.4077,
      "mean_token_accuracy": 0.21533680707216263,
      "num_tokens": 119451385.0,
      "step": 47130
    },
    {
      "entropy": 6.066489791870117,
      "epoch": 5.4396999422965955,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024775825108576903,
      "loss": 5.4629,
      "mean_token_accuracy": 0.22821397781372071,
      "num_tokens": 119462742.0,
      "step": 47135
    },
    {
      "entropy": 6.0465186595916744,
      "epoch": 5.440276976341604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024771729075636083,
      "loss": 5.4575,
      "mean_token_accuracy": 0.22218911349773407,
      "num_tokens": 119474996.0,
      "step": 47140
    },
    {
      "entropy": 6.062842798233032,
      "epoch": 5.440854010386613,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002476763313445931,
      "loss": 5.4655,
      "mean_token_accuracy": 0.21930796951055526,
      "num_tokens": 119486331.0,
      "step": 47145
    },
    {
      "entropy": 6.011796617507935,
      "epoch": 5.441431044431622,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024763537285184327,
      "loss": 5.4387,
      "mean_token_accuracy": 0.2246139571070671,
      "num_tokens": 119500081.0,
      "step": 47150
    },
    {
      "entropy": 6.094286584854126,
      "epoch": 5.44200807847663,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002475944152794892,
      "loss": 5.4985,
      "mean_token_accuracy": 0.22111781239509581,
      "num_tokens": 119512991.0,
      "step": 47155
    },
    {
      "entropy": 6.101238059997558,
      "epoch": 5.4425851125216385,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002475534586289082,
      "loss": 5.5133,
      "mean_token_accuracy": 0.21180179566144944,
      "num_tokens": 119526051.0,
      "step": 47160
    },
    {
      "entropy": 6.094206428527832,
      "epoch": 5.443162146566648,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002475125029014782,
      "loss": 5.4585,
      "mean_token_accuracy": 0.21571704000234604,
      "num_tokens": 119537609.0,
      "step": 47165
    },
    {
      "entropy": 6.02075343132019,
      "epoch": 5.443739180611656,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002474715480985763,
      "loss": 5.4065,
      "mean_token_accuracy": 0.22066868096590042,
      "num_tokens": 119550082.0,
      "step": 47170
    },
    {
      "entropy": 5.954913234710693,
      "epoch": 5.444316214656665,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024743059422158034,
      "loss": 5.3359,
      "mean_token_accuracy": 0.2281467780470848,
      "num_tokens": 119562304.0,
      "step": 47175
    },
    {
      "entropy": 6.028477954864502,
      "epoch": 5.444893248701673,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002473896412718675,
      "loss": 5.4589,
      "mean_token_accuracy": 0.22400152385234834,
      "num_tokens": 119575040.0,
      "step": 47180
    },
    {
      "entropy": 6.048852777481079,
      "epoch": 5.445470282746682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002473486892508156,
      "loss": 5.5052,
      "mean_token_accuracy": 0.21516649127006532,
      "num_tokens": 119587136.0,
      "step": 47185
    },
    {
      "entropy": 6.073791122436523,
      "epoch": 5.446047316791691,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002473077381598016,
      "loss": 5.4069,
      "mean_token_accuracy": 0.21754737198352814,
      "num_tokens": 119598831.0,
      "step": 47190
    },
    {
      "entropy": 6.111522626876831,
      "epoch": 5.446624350836699,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002472667880002032,
      "loss": 5.4937,
      "mean_token_accuracy": 0.21168259680271148,
      "num_tokens": 119611529.0,
      "step": 47195
    },
    {
      "entropy": 6.0160486698150635,
      "epoch": 5.4472013848817085,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002472258387733975,
      "loss": 5.409,
      "mean_token_accuracy": 0.22990969717502593,
      "num_tokens": 119623964.0,
      "step": 47200
    },
    {
      "entropy": 6.066016721725464,
      "epoch": 5.447778418926717,
      "grad_norm": 0.9375,
      "learning_rate": 0.000247184890480762,
      "loss": 5.5077,
      "mean_token_accuracy": 0.21833514273166657,
      "num_tokens": 119637407.0,
      "step": 47205
    },
    {
      "entropy": 6.018775701522827,
      "epoch": 5.448355452971725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002471439431236738,
      "loss": 5.4155,
      "mean_token_accuracy": 0.21863548308610917,
      "num_tokens": 119650539.0,
      "step": 47210
    },
    {
      "entropy": 6.133881425857544,
      "epoch": 5.448932487016734,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024710299670351033,
      "loss": 5.5008,
      "mean_token_accuracy": 0.21724879145622253,
      "num_tokens": 119662793.0,
      "step": 47215
    },
    {
      "entropy": 6.106074523925781,
      "epoch": 5.449509521061743,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002470620512216486,
      "loss": 5.5232,
      "mean_token_accuracy": 0.21195166260004045,
      "num_tokens": 119675866.0,
      "step": 47220
    },
    {
      "entropy": 6.070609426498413,
      "epoch": 5.450086555106751,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024702110667946603,
      "loss": 5.4365,
      "mean_token_accuracy": 0.2222458690404892,
      "num_tokens": 119688667.0,
      "step": 47225
    },
    {
      "entropy": 6.04173092842102,
      "epoch": 5.45066358915176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002469801630783395,
      "loss": 5.4111,
      "mean_token_accuracy": 0.21891222149133682,
      "num_tokens": 119701432.0,
      "step": 47230
    },
    {
      "entropy": 5.903676891326905,
      "epoch": 5.451240623196768,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002469392204196464,
      "loss": 5.3025,
      "mean_token_accuracy": 0.22939080744981766,
      "num_tokens": 119714824.0,
      "step": 47235
    },
    {
      "entropy": 6.072595071792603,
      "epoch": 5.451817657241778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002468982787047636,
      "loss": 5.4491,
      "mean_token_accuracy": 0.2178298532962799,
      "num_tokens": 119726647.0,
      "step": 47240
    },
    {
      "entropy": 5.997278881072998,
      "epoch": 5.452394691286786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002468573379350684,
      "loss": 5.3815,
      "mean_token_accuracy": 0.2236155554652214,
      "num_tokens": 119738755.0,
      "step": 47245
    },
    {
      "entropy": 5.907790088653565,
      "epoch": 5.452971725331794,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002468163981119375,
      "loss": 5.2945,
      "mean_token_accuracy": 0.22780349850654602,
      "num_tokens": 119751682.0,
      "step": 47250
    },
    {
      "entropy": 6.148678159713745,
      "epoch": 5.453548759376803,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000246775459236748,
      "loss": 5.5637,
      "mean_token_accuracy": 0.2070833995938301,
      "num_tokens": 119764283.0,
      "step": 47255
    },
    {
      "entropy": 6.072327899932861,
      "epoch": 5.454125793421812,
      "grad_norm": 1.0625,
      "learning_rate": 0.000246734521310877,
      "loss": 5.4178,
      "mean_token_accuracy": 0.21845017075538636,
      "num_tokens": 119776605.0,
      "step": 47260
    },
    {
      "entropy": 6.12420072555542,
      "epoch": 5.454702827466821,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024669358433570133,
      "loss": 5.501,
      "mean_token_accuracy": 0.21198708266019822,
      "num_tokens": 119789097.0,
      "step": 47265
    },
    {
      "entropy": 5.979126930236816,
      "epoch": 5.455279861511829,
      "grad_norm": 1.0,
      "learning_rate": 0.00024665264831259787,
      "loss": 5.34,
      "mean_token_accuracy": 0.23209788352251054,
      "num_tokens": 119802007.0,
      "step": 47270
    },
    {
      "entropy": 6.051213169097901,
      "epoch": 5.455856895556838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024661171324294353,
      "loss": 5.4513,
      "mean_token_accuracy": 0.2172643020749092,
      "num_tokens": 119814322.0,
      "step": 47275
    },
    {
      "entropy": 6.064874839782715,
      "epoch": 5.456433929601847,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024657077912811504,
      "loss": 5.5199,
      "mean_token_accuracy": 0.211814284324646,
      "num_tokens": 119827417.0,
      "step": 47280
    },
    {
      "entropy": 6.066554260253906,
      "epoch": 5.457010963646855,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002465298459694894,
      "loss": 5.456,
      "mean_token_accuracy": 0.219671568274498,
      "num_tokens": 119839914.0,
      "step": 47285
    },
    {
      "entropy": 5.995786046981811,
      "epoch": 5.4575879976918635,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024648891376844313,
      "loss": 5.3774,
      "mean_token_accuracy": 0.2309603810310364,
      "num_tokens": 119852685.0,
      "step": 47290
    },
    {
      "entropy": 6.085952377319336,
      "epoch": 5.458165031736873,
      "grad_norm": 1.125,
      "learning_rate": 0.00024644798252635314,
      "loss": 5.5347,
      "mean_token_accuracy": 0.21341227889060974,
      "num_tokens": 119865329.0,
      "step": 47295
    },
    {
      "entropy": 6.077068901062011,
      "epoch": 5.458742065781881,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024640705224459605,
      "loss": 5.5226,
      "mean_token_accuracy": 0.21247559487819673,
      "num_tokens": 119876706.0,
      "step": 47300
    },
    {
      "entropy": 6.052985143661499,
      "epoch": 5.45931909982689,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002463661229245486,
      "loss": 5.409,
      "mean_token_accuracy": 0.2194671928882599,
      "num_tokens": 119890326.0,
      "step": 47305
    },
    {
      "entropy": 6.067802047729492,
      "epoch": 5.459896133871898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024632519456758734,
      "loss": 5.3982,
      "mean_token_accuracy": 0.22458083927631378,
      "num_tokens": 119902502.0,
      "step": 47310
    },
    {
      "entropy": 6.1157200813293455,
      "epoch": 5.460473167916907,
      "grad_norm": 1.0,
      "learning_rate": 0.00024628426717508895,
      "loss": 5.5868,
      "mean_token_accuracy": 0.20857702940702438,
      "num_tokens": 119915454.0,
      "step": 47315
    },
    {
      "entropy": 5.999802494049073,
      "epoch": 5.461050201961916,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002462433407484299,
      "loss": 5.418,
      "mean_token_accuracy": 0.22297434657812118,
      "num_tokens": 119928719.0,
      "step": 47320
    },
    {
      "entropy": 5.989492988586425,
      "epoch": 5.461627236006924,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002462024152889869,
      "loss": 5.3809,
      "mean_token_accuracy": 0.2281783863902092,
      "num_tokens": 119940441.0,
      "step": 47325
    },
    {
      "entropy": 6.043868541717529,
      "epoch": 5.462204270051933,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002461614907981363,
      "loss": 5.4194,
      "mean_token_accuracy": 0.21828266978263855,
      "num_tokens": 119953262.0,
      "step": 47330
    },
    {
      "entropy": 6.0951252460479735,
      "epoch": 5.462781304096942,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024612056727725464,
      "loss": 5.4896,
      "mean_token_accuracy": 0.21910643726587295,
      "num_tokens": 119966154.0,
      "step": 47335
    },
    {
      "entropy": 6.067836284637451,
      "epoch": 5.46335833814195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024607964472771833,
      "loss": 5.4219,
      "mean_token_accuracy": 0.21964521259069442,
      "num_tokens": 119978595.0,
      "step": 47340
    },
    {
      "entropy": 6.003082799911499,
      "epoch": 5.463935372186959,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024603872315090387,
      "loss": 5.3282,
      "mean_token_accuracy": 0.2387758955359459,
      "num_tokens": 119991226.0,
      "step": 47345
    },
    {
      "entropy": 6.026238059997558,
      "epoch": 5.464512406231968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002459978025481875,
      "loss": 5.4525,
      "mean_token_accuracy": 0.21932094395160676,
      "num_tokens": 120004783.0,
      "step": 47350
    },
    {
      "entropy": 6.07522497177124,
      "epoch": 5.4650894402769765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002459568829209456,
      "loss": 5.4665,
      "mean_token_accuracy": 0.2183040276169777,
      "num_tokens": 120017218.0,
      "step": 47355
    },
    {
      "entropy": 6.029198360443115,
      "epoch": 5.465666474321985,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002459159642705546,
      "loss": 5.3945,
      "mean_token_accuracy": 0.23083268254995346,
      "num_tokens": 120030436.0,
      "step": 47360
    },
    {
      "entropy": 6.066330432891846,
      "epoch": 5.466243508366993,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002458750465983907,
      "loss": 5.4748,
      "mean_token_accuracy": 0.21342501789331436,
      "num_tokens": 120042214.0,
      "step": 47365
    },
    {
      "entropy": 6.032370567321777,
      "epoch": 5.466820542412003,
      "grad_norm": 0.953125,
      "learning_rate": 0.00024583412990583013,
      "loss": 5.4469,
      "mean_token_accuracy": 0.21803742051124572,
      "num_tokens": 120055215.0,
      "step": 47370
    },
    {
      "entropy": 6.115511178970337,
      "epoch": 5.467397576457011,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024579321419424904,
      "loss": 5.4174,
      "mean_token_accuracy": 0.21940379440784455,
      "num_tokens": 120068549.0,
      "step": 47375
    },
    {
      "entropy": 6.080186271667481,
      "epoch": 5.4679746105020195,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024575229946502376,
      "loss": 5.4795,
      "mean_token_accuracy": 0.22196500301361083,
      "num_tokens": 120082753.0,
      "step": 47380
    },
    {
      "entropy": 5.95504903793335,
      "epoch": 5.468551644547028,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024571138571953034,
      "loss": 5.4048,
      "mean_token_accuracy": 0.22288473397493364,
      "num_tokens": 120096109.0,
      "step": 47385
    },
    {
      "entropy": 6.061520910263061,
      "epoch": 5.469128678592037,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024567047295914496,
      "loss": 5.4399,
      "mean_token_accuracy": 0.21826471835374833,
      "num_tokens": 120108033.0,
      "step": 47390
    },
    {
      "entropy": 6.005728721618652,
      "epoch": 5.469705712637046,
      "grad_norm": 1.0,
      "learning_rate": 0.00024562956118524354,
      "loss": 5.3914,
      "mean_token_accuracy": 0.22716316133737563,
      "num_tokens": 120121467.0,
      "step": 47395
    },
    {
      "entropy": 6.070115566253662,
      "epoch": 5.470282746682054,
      "grad_norm": 1.0,
      "learning_rate": 0.00024558865039920236,
      "loss": 5.4438,
      "mean_token_accuracy": 0.21918773651123047,
      "num_tokens": 120132979.0,
      "step": 47400
    },
    {
      "entropy": 6.097302722930908,
      "epoch": 5.4708597807270625,
      "grad_norm": 1.0,
      "learning_rate": 0.0002455477406023972,
      "loss": 5.4285,
      "mean_token_accuracy": 0.2207154229283333,
      "num_tokens": 120144743.0,
      "step": 47405
    },
    {
      "entropy": 6.05715045928955,
      "epoch": 5.471436814772072,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00024550683179620433,
      "loss": 5.4517,
      "mean_token_accuracy": 0.2223209857940674,
      "num_tokens": 120157487.0,
      "step": 47410
    },
    {
      "entropy": 6.003508853912353,
      "epoch": 5.47201384881708,
      "grad_norm": 1.109375,
      "learning_rate": 0.00024546592398199936,
      "loss": 5.3884,
      "mean_token_accuracy": 0.22468466758728028,
      "num_tokens": 120170622.0,
      "step": 47415
    },
    {
      "entropy": 5.987598085403443,
      "epoch": 5.472590882862089,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002454250171611584,
      "loss": 5.334,
      "mean_token_accuracy": 0.23570979833602906,
      "num_tokens": 120184215.0,
      "step": 47420
    },
    {
      "entropy": 6.058071231842041,
      "epoch": 5.473167916907098,
      "grad_norm": 1.0,
      "learning_rate": 0.0002453841113350573,
      "loss": 5.4062,
      "mean_token_accuracy": 0.2212895005941391,
      "num_tokens": 120196058.0,
      "step": 47425
    },
    {
      "entropy": 6.015315103530884,
      "epoch": 5.473744950952106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024534320650507197,
      "loss": 5.501,
      "mean_token_accuracy": 0.2167505294084549,
      "num_tokens": 120208823.0,
      "step": 47430
    },
    {
      "entropy": 6.053440952301026,
      "epoch": 5.474321984997115,
      "grad_norm": 1.0625,
      "learning_rate": 0.000245302302672578,
      "loss": 5.3932,
      "mean_token_accuracy": 0.23017980754375458,
      "num_tokens": 120220081.0,
      "step": 47435
    },
    {
      "entropy": 6.083408975601197,
      "epoch": 5.474899019042123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024526139983895153,
      "loss": 5.5177,
      "mean_token_accuracy": 0.2169123575091362,
      "num_tokens": 120232450.0,
      "step": 47440
    },
    {
      "entropy": 6.044406270980835,
      "epoch": 5.4754760530871325,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002452204980055679,
      "loss": 5.4706,
      "mean_token_accuracy": 0.22209678143262862,
      "num_tokens": 120243952.0,
      "step": 47445
    },
    {
      "entropy": 6.012739467620849,
      "epoch": 5.476053087132141,
      "grad_norm": 1.0,
      "learning_rate": 0.00024517959717380324,
      "loss": 5.4239,
      "mean_token_accuracy": 0.22665991336107255,
      "num_tokens": 120257282.0,
      "step": 47450
    },
    {
      "entropy": 6.050950717926026,
      "epoch": 5.476630121177149,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00024513869734503284,
      "loss": 5.376,
      "mean_token_accuracy": 0.2304854065179825,
      "num_tokens": 120268957.0,
      "step": 47455
    },
    {
      "entropy": 6.105955314636231,
      "epoch": 5.477207155222158,
      "grad_norm": 1.0,
      "learning_rate": 0.0002450977985206327,
      "loss": 5.5999,
      "mean_token_accuracy": 0.20874406695365905,
      "num_tokens": 120281600.0,
      "step": 47460
    },
    {
      "entropy": 5.980347681045532,
      "epoch": 5.477784189267167,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024505690070197807,
      "loss": 5.4072,
      "mean_token_accuracy": 0.22288106828927995,
      "num_tokens": 120294725.0,
      "step": 47465
    },
    {
      "entropy": 6.034222841262817,
      "epoch": 5.4783612233121755,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002450160038904448,
      "loss": 5.3452,
      "mean_token_accuracy": 0.22752912789583207,
      "num_tokens": 120306822.0,
      "step": 47470
    },
    {
      "entropy": 6.05337438583374,
      "epoch": 5.478938257357184,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002449751080874083,
      "loss": 5.4105,
      "mean_token_accuracy": 0.21632318198680878,
      "num_tokens": 120319902.0,
      "step": 47475
    },
    {
      "entropy": 5.985331773757935,
      "epoch": 5.479515291402192,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002449342132942443,
      "loss": 5.3615,
      "mean_token_accuracy": 0.22828896194696427,
      "num_tokens": 120333623.0,
      "step": 47480
    },
    {
      "entropy": 5.9836499214172365,
      "epoch": 5.480092325447202,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002448933195123278,
      "loss": 5.3861,
      "mean_token_accuracy": 0.22556522488594055,
      "num_tokens": 120345458.0,
      "step": 47485
    },
    {
      "entropy": 6.062854671478272,
      "epoch": 5.48066935949221,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002448524267430348,
      "loss": 5.4561,
      "mean_token_accuracy": 0.213370643556118,
      "num_tokens": 120358757.0,
      "step": 47490
    },
    {
      "entropy": 6.093607568740845,
      "epoch": 5.481246393537218,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002448115349877403,
      "loss": 5.5078,
      "mean_token_accuracy": 0.21637726724147796,
      "num_tokens": 120371062.0,
      "step": 47495
    },
    {
      "entropy": 6.1550556182861325,
      "epoch": 5.481823427582228,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002447706442478199,
      "loss": 5.5315,
      "mean_token_accuracy": 0.21053755134344102,
      "num_tokens": 120382934.0,
      "step": 47500
    },
    {
      "entropy": 6.062007284164428,
      "epoch": 5.482400461627236,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002447297545246489,
      "loss": 5.4233,
      "mean_token_accuracy": 0.2202950969338417,
      "num_tokens": 120395103.0,
      "step": 47505
    },
    {
      "entropy": 6.056242656707764,
      "epoch": 5.482977495672245,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002446888658196025,
      "loss": 5.43,
      "mean_token_accuracy": 0.22057365775108337,
      "num_tokens": 120406852.0,
      "step": 47510
    },
    {
      "entropy": 5.998345994949341,
      "epoch": 5.483554529717253,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000244647978134056,
      "loss": 5.3725,
      "mean_token_accuracy": 0.22972165644168854,
      "num_tokens": 120419813.0,
      "step": 47515
    },
    {
      "entropy": 5.972995328903198,
      "epoch": 5.484131563762262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024460709146938473,
      "loss": 5.3677,
      "mean_token_accuracy": 0.2235831692814827,
      "num_tokens": 120432103.0,
      "step": 47520
    },
    {
      "entropy": 5.968393087387085,
      "epoch": 5.484708597807271,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00024456620582696386,
      "loss": 5.3884,
      "mean_token_accuracy": 0.222498719394207,
      "num_tokens": 120446458.0,
      "step": 47525
    },
    {
      "entropy": 6.034120893478393,
      "epoch": 5.485285631852279,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024452532120816856,
      "loss": 5.3883,
      "mean_token_accuracy": 0.21942399591207504,
      "num_tokens": 120459456.0,
      "step": 47530
    },
    {
      "entropy": 6.013608598709107,
      "epoch": 5.4858626658972875,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002444844376143739,
      "loss": 5.3238,
      "mean_token_accuracy": 0.22153729647397996,
      "num_tokens": 120472475.0,
      "step": 47535
    },
    {
      "entropy": 5.953195190429687,
      "epoch": 5.486439699942297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00024444355504695507,
      "loss": 5.3311,
      "mean_token_accuracy": 0.23062245845794677,
      "num_tokens": 120486782.0,
      "step": 47540
    },
    {
      "entropy": 5.976576423645019,
      "epoch": 5.487016733987305,
      "grad_norm": 1.015625,
      "learning_rate": 0.000244402673507287,
      "loss": 5.3564,
      "mean_token_accuracy": 0.22411105483770372,
      "num_tokens": 120499866.0,
      "step": 47545
    },
    {
      "entropy": 6.015101480484009,
      "epoch": 5.487593768032314,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002443617929967449,
      "loss": 5.4647,
      "mean_token_accuracy": 0.2205308571457863,
      "num_tokens": 120512920.0,
      "step": 47550
    },
    {
      "entropy": 5.935817289352417,
      "epoch": 5.488170802077322,
      "grad_norm": 0.96875,
      "learning_rate": 0.00024432091351670354,
      "loss": 5.3037,
      "mean_token_accuracy": 0.23296899795532228,
      "num_tokens": 120525262.0,
      "step": 47555
    },
    {
      "entropy": 6.0095116138458256,
      "epoch": 5.488747836122331,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002442800350685382,
      "loss": 5.4062,
      "mean_token_accuracy": 0.22227492183446884,
      "num_tokens": 120537888.0,
      "step": 47560
    },
    {
      "entropy": 6.096463441848755,
      "epoch": 5.48932487016734,
      "grad_norm": 1.0,
      "learning_rate": 0.0002442391576536235,
      "loss": 5.5657,
      "mean_token_accuracy": 0.21110780090093612,
      "num_tokens": 120550826.0,
      "step": 47565
    },
    {
      "entropy": 6.102590322494507,
      "epoch": 5.489901904212348,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002441982812733345,
      "loss": 5.4156,
      "mean_token_accuracy": 0.22429123073816298,
      "num_tokens": 120562801.0,
      "step": 47570
    },
    {
      "entropy": 6.0433797359466555,
      "epoch": 5.4904789382573576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024415740592904597,
      "loss": 5.398,
      "mean_token_accuracy": 0.2184971123933792,
      "num_tokens": 120575080.0,
      "step": 47575
    },
    {
      "entropy": 6.011787557601929,
      "epoch": 5.491055972302366,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024411653162213282,
      "loss": 5.4585,
      "mean_token_accuracy": 0.21966557651758195,
      "num_tokens": 120587460.0,
      "step": 47580
    },
    {
      "entropy": 6.052439594268799,
      "epoch": 5.491633006347374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002440756583539697,
      "loss": 5.4146,
      "mean_token_accuracy": 0.2198610469698906,
      "num_tokens": 120599970.0,
      "step": 47585
    },
    {
      "entropy": 5.9410182476043705,
      "epoch": 5.492210040392383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024403478612593156,
      "loss": 5.3648,
      "mean_token_accuracy": 0.22431038469076156,
      "num_tokens": 120612546.0,
      "step": 47590
    },
    {
      "entropy": 6.047664785385132,
      "epoch": 5.492787074437392,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024399391493939293,
      "loss": 5.4638,
      "mean_token_accuracy": 0.21801050305366515,
      "num_tokens": 120624648.0,
      "step": 47595
    },
    {
      "entropy": 6.113511180877685,
      "epoch": 5.4933641084824005,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002439530447957286,
      "loss": 5.4757,
      "mean_token_accuracy": 0.21548536866903306,
      "num_tokens": 120637530.0,
      "step": 47600
    },
    {
      "entropy": 6.051801538467407,
      "epoch": 5.493941142527409,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024391217569631318,
      "loss": 5.4024,
      "mean_token_accuracy": 0.2266889750957489,
      "num_tokens": 120649836.0,
      "step": 47605
    },
    {
      "entropy": 6.06033411026001,
      "epoch": 5.494518176572417,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002438713076425213,
      "loss": 5.5336,
      "mean_token_accuracy": 0.21252259314060212,
      "num_tokens": 120662535.0,
      "step": 47610
    },
    {
      "entropy": 6.048929929733276,
      "epoch": 5.495095210617427,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002438304406357275,
      "loss": 5.4076,
      "mean_token_accuracy": 0.2256210908293724,
      "num_tokens": 120675289.0,
      "step": 47615
    },
    {
      "entropy": 6.141551446914673,
      "epoch": 5.495672244662435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024378957467730645,
      "loss": 5.4675,
      "mean_token_accuracy": 0.21609950810670853,
      "num_tokens": 120689208.0,
      "step": 47620
    },
    {
      "entropy": 6.0638526439666744,
      "epoch": 5.4962492787074435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002437487097686324,
      "loss": 5.3945,
      "mean_token_accuracy": 0.22044505923986435,
      "num_tokens": 120700660.0,
      "step": 47625
    },
    {
      "entropy": 6.009811305999756,
      "epoch": 5.496826312752452,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024370784591108002,
      "loss": 5.3857,
      "mean_token_accuracy": 0.222472120821476,
      "num_tokens": 120712114.0,
      "step": 47630
    },
    {
      "entropy": 6.029832887649536,
      "epoch": 5.497403346797461,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002436669831060237,
      "loss": 5.3918,
      "mean_token_accuracy": 0.23503619879484178,
      "num_tokens": 120725751.0,
      "step": 47635
    },
    {
      "entropy": 6.019428396224976,
      "epoch": 5.49798038084247,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024362612135483792,
      "loss": 5.4228,
      "mean_token_accuracy": 0.22621234655380248,
      "num_tokens": 120738396.0,
      "step": 47640
    },
    {
      "entropy": 6.02667236328125,
      "epoch": 5.498557414887478,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024358526065889686,
      "loss": 5.3715,
      "mean_token_accuracy": 0.22395648509263993,
      "num_tokens": 120750460.0,
      "step": 47645
    },
    {
      "entropy": 6.022999858856201,
      "epoch": 5.499134448932487,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024354440101957505,
      "loss": 5.4314,
      "mean_token_accuracy": 0.22071062177419662,
      "num_tokens": 120763227.0,
      "step": 47650
    },
    {
      "entropy": 6.132009840011596,
      "epoch": 5.499711482977496,
      "grad_norm": 1.078125,
      "learning_rate": 0.00024350354243824659,
      "loss": 5.5243,
      "mean_token_accuracy": 0.20742686986923217,
      "num_tokens": 120774786.0,
      "step": 47655
    },
    {
      "entropy": 6.013866996765136,
      "epoch": 5.500288517022504,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002434626849162859,
      "loss": 5.4096,
      "mean_token_accuracy": 0.2245435893535614,
      "num_tokens": 120787806.0,
      "step": 47660
    },
    {
      "entropy": 6.13320574760437,
      "epoch": 5.500865551067513,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024342182845506712,
      "loss": 5.5137,
      "mean_token_accuracy": 0.21297620087862015,
      "num_tokens": 120800580.0,
      "step": 47665
    },
    {
      "entropy": 6.0649432182312015,
      "epoch": 5.501442585112522,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024338097305596446,
      "loss": 5.4349,
      "mean_token_accuracy": 0.22557477951049804,
      "num_tokens": 120813684.0,
      "step": 47670
    },
    {
      "entropy": 6.006740236282349,
      "epoch": 5.50201961915753,
      "grad_norm": 0.953125,
      "learning_rate": 0.000243340118720352,
      "loss": 5.429,
      "mean_token_accuracy": 0.2203612744808197,
      "num_tokens": 120828743.0,
      "step": 47675
    },
    {
      "entropy": 6.054507160186768,
      "epoch": 5.502596653202539,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024329926544960402,
      "loss": 5.4541,
      "mean_token_accuracy": 0.21979046165943145,
      "num_tokens": 120841731.0,
      "step": 47680
    },
    {
      "entropy": 6.000650262832641,
      "epoch": 5.503173687247548,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00024325841324509447,
      "loss": 5.3505,
      "mean_token_accuracy": 0.23008284270763396,
      "num_tokens": 120854457.0,
      "step": 47685
    },
    {
      "entropy": 5.9935808181762695,
      "epoch": 5.5037507212925565,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024321756210819734,
      "loss": 5.4077,
      "mean_token_accuracy": 0.22334915846586229,
      "num_tokens": 120866635.0,
      "step": 47690
    },
    {
      "entropy": 6.067073631286621,
      "epoch": 5.504327755337565,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002431767120402868,
      "loss": 5.4607,
      "mean_token_accuracy": 0.21895507723093033,
      "num_tokens": 120879074.0,
      "step": 47695
    },
    {
      "entropy": 6.061027717590332,
      "epoch": 5.504904789382573,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024313586304273676,
      "loss": 5.439,
      "mean_token_accuracy": 0.21975450068712235,
      "num_tokens": 120891468.0,
      "step": 47700
    },
    {
      "entropy": 5.980240106582642,
      "epoch": 5.505481823427582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024309501511692112,
      "loss": 5.3291,
      "mean_token_accuracy": 0.22788507640361785,
      "num_tokens": 120903559.0,
      "step": 47705
    },
    {
      "entropy": 6.059676456451416,
      "epoch": 5.506058857472591,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002430541682642138,
      "loss": 5.4684,
      "mean_token_accuracy": 0.21489742994308472,
      "num_tokens": 120916875.0,
      "step": 47710
    },
    {
      "entropy": 6.099525594711304,
      "epoch": 5.5066358915175995,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002430133224859886,
      "loss": 5.481,
      "mean_token_accuracy": 0.22402574867010117,
      "num_tokens": 120929676.0,
      "step": 47715
    },
    {
      "entropy": 6.073724603652954,
      "epoch": 5.507212925562608,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024297247778361952,
      "loss": 5.4655,
      "mean_token_accuracy": 0.21962770223617553,
      "num_tokens": 120941913.0,
      "step": 47720
    },
    {
      "entropy": 6.069832420349121,
      "epoch": 5.507789959607617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024293163415848012,
      "loss": 5.4931,
      "mean_token_accuracy": 0.22217957973480223,
      "num_tokens": 120953614.0,
      "step": 47725
    },
    {
      "entropy": 6.058898115158081,
      "epoch": 5.508366993652626,
      "grad_norm": 0.9375,
      "learning_rate": 0.00024289079161194437,
      "loss": 5.4537,
      "mean_token_accuracy": 0.22171433568000792,
      "num_tokens": 120967921.0,
      "step": 47730
    },
    {
      "entropy": 6.112684726715088,
      "epoch": 5.508944027697634,
      "grad_norm": 0.9375,
      "learning_rate": 0.00024284995014538585,
      "loss": 5.4763,
      "mean_token_accuracy": 0.22013049125671386,
      "num_tokens": 120980838.0,
      "step": 47735
    },
    {
      "entropy": 6.022745180130005,
      "epoch": 5.5095210617426424,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024280910976017835,
      "loss": 5.4452,
      "mean_token_accuracy": 0.22427271604537963,
      "num_tokens": 120992772.0,
      "step": 47740
    },
    {
      "entropy": 5.993321275711059,
      "epoch": 5.510098095787652,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002427682704576954,
      "loss": 5.4221,
      "mean_token_accuracy": 0.22051352262496948,
      "num_tokens": 121005767.0,
      "step": 47745
    },
    {
      "entropy": 6.042178201675415,
      "epoch": 5.51067512983266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002427274322393106,
      "loss": 5.368,
      "mean_token_accuracy": 0.22791897505521774,
      "num_tokens": 121018137.0,
      "step": 47750
    },
    {
      "entropy": 6.010069799423218,
      "epoch": 5.511252163877669,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002426865951063977,
      "loss": 5.3828,
      "mean_token_accuracy": 0.22456735223531724,
      "num_tokens": 121029980.0,
      "step": 47755
    },
    {
      "entropy": 5.962734651565552,
      "epoch": 5.511829197922678,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00024264575906033,
      "loss": 5.3912,
      "mean_token_accuracy": 0.21795674711465834,
      "num_tokens": 121042270.0,
      "step": 47760
    },
    {
      "entropy": 6.015217018127442,
      "epoch": 5.512406231967686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024260492410248125,
      "loss": 5.4571,
      "mean_token_accuracy": 0.2169657602906227,
      "num_tokens": 121053808.0,
      "step": 47765
    },
    {
      "entropy": 6.045488691329956,
      "epoch": 5.512983266012695,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002425640902342247,
      "loss": 5.4444,
      "mean_token_accuracy": 0.2223682537674904,
      "num_tokens": 121066112.0,
      "step": 47770
    },
    {
      "entropy": 6.067818975448608,
      "epoch": 5.513560300057703,
      "grad_norm": 1.0,
      "learning_rate": 0.0002425232574569339,
      "loss": 5.4313,
      "mean_token_accuracy": 0.22687044590711594,
      "num_tokens": 121079687.0,
      "step": 47775
    },
    {
      "entropy": 6.010495805740357,
      "epoch": 5.514137334102712,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002424824257719822,
      "loss": 5.3328,
      "mean_token_accuracy": 0.22875746637582778,
      "num_tokens": 121091446.0,
      "step": 47780
    },
    {
      "entropy": 6.044916439056396,
      "epoch": 5.514714368147721,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000242441595180743,
      "loss": 5.4493,
      "mean_token_accuracy": 0.21559704095125198,
      "num_tokens": 121104326.0,
      "step": 47785
    },
    {
      "entropy": 6.007753705978393,
      "epoch": 5.515291402192729,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024240076568458947,
      "loss": 5.3847,
      "mean_token_accuracy": 0.2253635972738266,
      "num_tokens": 121116717.0,
      "step": 47790
    },
    {
      "entropy": 6.063485622406006,
      "epoch": 5.515868436237738,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00024235993728489508,
      "loss": 5.4731,
      "mean_token_accuracy": 0.21702780425548554,
      "num_tokens": 121130737.0,
      "step": 47795
    },
    {
      "entropy": 6.06763801574707,
      "epoch": 5.516445470282747,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024231910998303292,
      "loss": 5.496,
      "mean_token_accuracy": 0.21376340389251708,
      "num_tokens": 121142663.0,
      "step": 47800
    },
    {
      "entropy": 5.987742233276367,
      "epoch": 5.517022504327755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024227828378037635,
      "loss": 5.405,
      "mean_token_accuracy": 0.2216850697994232,
      "num_tokens": 121154873.0,
      "step": 47805
    },
    {
      "entropy": 6.020955753326416,
      "epoch": 5.517599538372764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024223745867829838,
      "loss": 5.4195,
      "mean_token_accuracy": 0.22612695693969725,
      "num_tokens": 121167564.0,
      "step": 47810
    },
    {
      "entropy": 5.975015020370483,
      "epoch": 5.518176572417772,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00024219663467817226,
      "loss": 5.4245,
      "mean_token_accuracy": 0.2225605145096779,
      "num_tokens": 121181296.0,
      "step": 47815
    },
    {
      "entropy": 6.1260621547698975,
      "epoch": 5.518753606462782,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00024215581178137097,
      "loss": 5.4409,
      "mean_token_accuracy": 0.22733189910650253,
      "num_tokens": 121194282.0,
      "step": 47820
    },
    {
      "entropy": 6.017096710205078,
      "epoch": 5.51933064050779,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024211498998926773,
      "loss": 5.4058,
      "mean_token_accuracy": 0.22633518129587174,
      "num_tokens": 121207170.0,
      "step": 47825
    },
    {
      "entropy": 6.006448459625244,
      "epoch": 5.519907674552798,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024207416930323539,
      "loss": 5.3562,
      "mean_token_accuracy": 0.2306416615843773,
      "num_tokens": 121220484.0,
      "step": 47830
    },
    {
      "entropy": 5.999253511428833,
      "epoch": 5.520484708597808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002420333497246471,
      "loss": 5.3574,
      "mean_token_accuracy": 0.22625994235277175,
      "num_tokens": 121233302.0,
      "step": 47835
    },
    {
      "entropy": 6.053505754470825,
      "epoch": 5.521061742642816,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024199253125487558,
      "loss": 5.5222,
      "mean_token_accuracy": 0.20680065751075744,
      "num_tokens": 121245633.0,
      "step": 47840
    },
    {
      "entropy": 6.008155679702758,
      "epoch": 5.5216387766878245,
      "grad_norm": 0.953125,
      "learning_rate": 0.00024195171389529402,
      "loss": 5.3799,
      "mean_token_accuracy": 0.22610725164413453,
      "num_tokens": 121258278.0,
      "step": 47845
    },
    {
      "entropy": 6.010450935363769,
      "epoch": 5.522215810732833,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024191089764727503,
      "loss": 5.3871,
      "mean_token_accuracy": 0.22943540066480636,
      "num_tokens": 121272504.0,
      "step": 47850
    },
    {
      "entropy": 6.006346035003662,
      "epoch": 5.522792844777842,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024187008251219167,
      "loss": 5.3207,
      "mean_token_accuracy": 0.22688601911067963,
      "num_tokens": 121284389.0,
      "step": 47855
    },
    {
      "entropy": 6.204504108428955,
      "epoch": 5.523369878822851,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024182926849141646,
      "loss": 5.594,
      "mean_token_accuracy": 0.20778852701187134,
      "num_tokens": 121297039.0,
      "step": 47860
    },
    {
      "entropy": 6.010227537155151,
      "epoch": 5.523946912867859,
      "grad_norm": 1.0,
      "learning_rate": 0.00024178845558632252,
      "loss": 5.4471,
      "mean_token_accuracy": 0.2188406467437744,
      "num_tokens": 121309748.0,
      "step": 47865
    },
    {
      "entropy": 6.105253553390503,
      "epoch": 5.5245239469128675,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024174764379828224,
      "loss": 5.4981,
      "mean_token_accuracy": 0.21365632861852646,
      "num_tokens": 121321103.0,
      "step": 47870
    },
    {
      "entropy": 6.0800862312316895,
      "epoch": 5.525100980957877,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024170683312866854,
      "loss": 5.4297,
      "mean_token_accuracy": 0.21937412917613983,
      "num_tokens": 121332960.0,
      "step": 47875
    },
    {
      "entropy": 6.0757159233093265,
      "epoch": 5.525678015002885,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024166602357885393,
      "loss": 5.4347,
      "mean_token_accuracy": 0.22198549062013626,
      "num_tokens": 121345361.0,
      "step": 47880
    },
    {
      "entropy": 6.073025035858154,
      "epoch": 5.526255049047894,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024162521515021102,
      "loss": 5.4934,
      "mean_token_accuracy": 0.21564815044403077,
      "num_tokens": 121358858.0,
      "step": 47885
    },
    {
      "entropy": 6.130002021789551,
      "epoch": 5.526832083092902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024158440784411244,
      "loss": 5.4811,
      "mean_token_accuracy": 0.21908482015132905,
      "num_tokens": 121372106.0,
      "step": 47890
    },
    {
      "entropy": 6.098454713821411,
      "epoch": 5.527409117137911,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002415436016619308,
      "loss": 5.4684,
      "mean_token_accuracy": 0.22149865925312043,
      "num_tokens": 121385340.0,
      "step": 47895
    },
    {
      "entropy": 6.021191072463989,
      "epoch": 5.52798615118292,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024150279660503838,
      "loss": 5.429,
      "mean_token_accuracy": 0.22179484814405442,
      "num_tokens": 121397962.0,
      "step": 47900
    },
    {
      "entropy": 6.031183338165283,
      "epoch": 5.528563185227928,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024146199267480778,
      "loss": 5.4712,
      "mean_token_accuracy": 0.2198013424873352,
      "num_tokens": 121409678.0,
      "step": 47905
    },
    {
      "entropy": 6.053447198867798,
      "epoch": 5.5291402192729375,
      "grad_norm": 1.0,
      "learning_rate": 0.0002414211898726114,
      "loss": 5.4368,
      "mean_token_accuracy": 0.2185912847518921,
      "num_tokens": 121422499.0,
      "step": 47910
    },
    {
      "entropy": 6.119691371917725,
      "epoch": 5.529717253317946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024138038819982173,
      "loss": 5.5087,
      "mean_token_accuracy": 0.21256650239229202,
      "num_tokens": 121434562.0,
      "step": 47915
    },
    {
      "entropy": 6.003648567199707,
      "epoch": 5.530294287362954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024133958765781084,
      "loss": 5.4237,
      "mean_token_accuracy": 0.22699394673109055,
      "num_tokens": 121447708.0,
      "step": 47920
    },
    {
      "entropy": 6.122247409820557,
      "epoch": 5.530871321407963,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024129878824795127,
      "loss": 5.5456,
      "mean_token_accuracy": 0.2117946833372116,
      "num_tokens": 121459580.0,
      "step": 47925
    },
    {
      "entropy": 5.9470449924469,
      "epoch": 5.531448355452972,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002412579899716152,
      "loss": 5.3503,
      "mean_token_accuracy": 0.23912058919668197,
      "num_tokens": 121472520.0,
      "step": 47930
    },
    {
      "entropy": 6.036983060836792,
      "epoch": 5.5320253894979805,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002412171928301749,
      "loss": 5.3577,
      "mean_token_accuracy": 0.232684625685215,
      "num_tokens": 121485220.0,
      "step": 47935
    },
    {
      "entropy": 5.952143621444702,
      "epoch": 5.532602423542989,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024117639682500253,
      "loss": 5.3743,
      "mean_token_accuracy": 0.2347503498196602,
      "num_tokens": 121499731.0,
      "step": 47940
    },
    {
      "entropy": 5.994808197021484,
      "epoch": 5.533179457587997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00024113560195747024,
      "loss": 5.4141,
      "mean_token_accuracy": 0.2176785558462143,
      "num_tokens": 121512445.0,
      "step": 47945
    },
    {
      "entropy": 6.119512844085693,
      "epoch": 5.533756491633007,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00024109480822895012,
      "loss": 5.4882,
      "mean_token_accuracy": 0.2130196288228035,
      "num_tokens": 121527174.0,
      "step": 47950
    },
    {
      "entropy": 6.003230953216553,
      "epoch": 5.534333525678015,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024105401564081442,
      "loss": 5.3411,
      "mean_token_accuracy": 0.23128325790166854,
      "num_tokens": 121540541.0,
      "step": 47955
    },
    {
      "entropy": 6.017929744720459,
      "epoch": 5.5349105597230235,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002410132241944349,
      "loss": 5.4135,
      "mean_token_accuracy": 0.22472924143075942,
      "num_tokens": 121552536.0,
      "step": 47960
    },
    {
      "entropy": 5.965036535263062,
      "epoch": 5.535487593768032,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024097243389118377,
      "loss": 5.3169,
      "mean_token_accuracy": 0.2277855783700943,
      "num_tokens": 121565737.0,
      "step": 47965
    },
    {
      "entropy": 6.0405150890350345,
      "epoch": 5.536064627813041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024093164473243283,
      "loss": 5.4157,
      "mean_token_accuracy": 0.2299483373761177,
      "num_tokens": 121577382.0,
      "step": 47970
    },
    {
      "entropy": 6.082750415802002,
      "epoch": 5.53664166185805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002408908567195543,
      "loss": 5.5376,
      "mean_token_accuracy": 0.2111786499619484,
      "num_tokens": 121590595.0,
      "step": 47975
    },
    {
      "entropy": 6.0435713768005375,
      "epoch": 5.537218695903058,
      "grad_norm": 1.046875,
      "learning_rate": 0.00024085006985391966,
      "loss": 5.4121,
      "mean_token_accuracy": 0.22380685657262803,
      "num_tokens": 121603237.0,
      "step": 47980
    },
    {
      "entropy": 6.0881125926971436,
      "epoch": 5.537795729948067,
      "grad_norm": 1.0625,
      "learning_rate": 0.00024080928413690105,
      "loss": 5.468,
      "mean_token_accuracy": 0.2131459578871727,
      "num_tokens": 121614377.0,
      "step": 47985
    },
    {
      "entropy": 5.9977563381195065,
      "epoch": 5.538372763993076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00024076849956987013,
      "loss": 5.3402,
      "mean_token_accuracy": 0.23828363418579102,
      "num_tokens": 121627115.0,
      "step": 47990
    },
    {
      "entropy": 6.038291883468628,
      "epoch": 5.538949798038084,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002407277161541988,
      "loss": 5.4412,
      "mean_token_accuracy": 0.22136062532663345,
      "num_tokens": 121641392.0,
      "step": 47995
    },
    {
      "entropy": 5.990301036834717,
      "epoch": 5.539526832083093,
      "grad_norm": 1.03125,
      "learning_rate": 0.00024068693389125867,
      "loss": 5.3982,
      "mean_token_accuracy": 0.22406841069459915,
      "num_tokens": 121654998.0,
      "step": 48000
    },
    {
      "epoch": 5.539526832083093,
      "eval_entropy": 5.838063651606343,
      "eval_loss": 5.613852024078369,
      "eval_mean_token_accuracy": 0.21948676820731075,
      "eval_num_tokens": 121654998.0,
      "eval_runtime": 22.7255,
      "eval_samples_per_second": 1238.08,
      "eval_steps_per_second": 154.76,
      "step": 48000
    },
    {
      "entropy": 6.126865100860596,
      "epoch": 5.540103866128102,
      "grad_norm": 0.984375,
      "learning_rate": 0.00024064615278242147,
      "loss": 5.4786,
      "mean_token_accuracy": 0.21697935461997986,
      "num_tokens": 121666714.0,
      "step": 48005
    },
    {
      "entropy": 6.011840152740478,
      "epoch": 5.54068090017311,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00024060537282905882,
      "loss": 5.3997,
      "mean_token_accuracy": 0.22318132668733598,
      "num_tokens": 121679129.0,
      "step": 48010
    },
    {
      "entropy": 6.008998537063599,
      "epoch": 5.541257934218119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002405645940325425,
      "loss": 5.3898,
      "mean_token_accuracy": 0.22001457512378692,
      "num_tokens": 121692416.0,
      "step": 48015
    },
    {
      "entropy": 6.0934515476226805,
      "epoch": 5.541834968263127,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00024052381639424382,
      "loss": 5.4936,
      "mean_token_accuracy": 0.21239448338747025,
      "num_tokens": 121705956.0,
      "step": 48020
    },
    {
      "entropy": 6.075184869766235,
      "epoch": 5.5424120023081365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024048303991553455,
      "loss": 5.4762,
      "mean_token_accuracy": 0.22139879018068315,
      "num_tokens": 121718009.0,
      "step": 48025
    },
    {
      "entropy": 6.03062162399292,
      "epoch": 5.542989036353145,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00024044226459778596,
      "loss": 5.4662,
      "mean_token_accuracy": 0.21916611194610597,
      "num_tokens": 121730719.0,
      "step": 48030
    },
    {
      "entropy": 6.0477649688720705,
      "epoch": 5.543566070398153,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002404014904423698,
      "loss": 5.4828,
      "mean_token_accuracy": 0.21260360330343248,
      "num_tokens": 121743720.0,
      "step": 48035
    },
    {
      "entropy": 6.088364171981811,
      "epoch": 5.544143104443162,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002403607174506572,
      "loss": 5.452,
      "mean_token_accuracy": 0.22121359258890153,
      "num_tokens": 121755883.0,
      "step": 48040
    },
    {
      "entropy": 6.091207790374756,
      "epoch": 5.544720138488171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00024031994562401972,
      "loss": 5.4622,
      "mean_token_accuracy": 0.21255310326814653,
      "num_tokens": 121768832.0,
      "step": 48045
    },
    {
      "entropy": 6.112627649307251,
      "epoch": 5.545297172533179,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024027917496382856,
      "loss": 5.4902,
      "mean_token_accuracy": 0.2171428233385086,
      "num_tokens": 121781396.0,
      "step": 48050
    },
    {
      "entropy": 6.060887622833252,
      "epoch": 5.545874206578188,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002402384054714552,
      "loss": 5.4583,
      "mean_token_accuracy": 0.20708019584417342,
      "num_tokens": 121795758.0,
      "step": 48055
    },
    {
      "entropy": 5.846577215194702,
      "epoch": 5.546451240623197,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00024019763714827077,
      "loss": 5.1755,
      "mean_token_accuracy": 0.24399437606334687,
      "num_tokens": 121808114.0,
      "step": 48060
    },
    {
      "entropy": 6.042523765563965,
      "epoch": 5.547028274668206,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002401568699956465,
      "loss": 5.4221,
      "mean_token_accuracy": 0.21863186806440355,
      "num_tokens": 121820130.0,
      "step": 48065
    },
    {
      "entropy": 6.059017610549927,
      "epoch": 5.547605308713214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002401161040149536,
      "loss": 5.4687,
      "mean_token_accuracy": 0.21921880543231964,
      "num_tokens": 121833344.0,
      "step": 48070
    },
    {
      "entropy": 6.051173877716065,
      "epoch": 5.548182342758222,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002400753392075633,
      "loss": 5.4172,
      "mean_token_accuracy": 0.22536935210227965,
      "num_tokens": 121846322.0,
      "step": 48075
    },
    {
      "entropy": 5.9434576511383055,
      "epoch": 5.548759376803232,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00024003457557484655,
      "loss": 5.3034,
      "mean_token_accuracy": 0.22719025611877441,
      "num_tokens": 121859334.0,
      "step": 48080
    },
    {
      "entropy": 6.103307247161865,
      "epoch": 5.54933641084824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023999381311817447,
      "loss": 5.416,
      "mean_token_accuracy": 0.21675145030021667,
      "num_tokens": 121871421.0,
      "step": 48085
    },
    {
      "entropy": 6.0599720001220705,
      "epoch": 5.549913444893249,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023995305183891808,
      "loss": 5.4923,
      "mean_token_accuracy": 0.21771734207868576,
      "num_tokens": 121883335.0,
      "step": 48090
    },
    {
      "entropy": 6.0812835693359375,
      "epoch": 5.550490478938257,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023991229173844848,
      "loss": 5.507,
      "mean_token_accuracy": 0.21519697159528733,
      "num_tokens": 121895875.0,
      "step": 48095
    },
    {
      "entropy": 5.980572080612182,
      "epoch": 5.551067512983266,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002398715328181364,
      "loss": 5.381,
      "mean_token_accuracy": 0.2272232487797737,
      "num_tokens": 121909044.0,
      "step": 48100
    },
    {
      "entropy": 6.041309118270874,
      "epoch": 5.551644547028275,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00023983077507935298,
      "loss": 5.3919,
      "mean_token_accuracy": 0.22771453708410264,
      "num_tokens": 121922050.0,
      "step": 48105
    },
    {
      "entropy": 6.047741365432739,
      "epoch": 5.552221581073283,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023979001852346882,
      "loss": 5.4519,
      "mean_token_accuracy": 0.22750690281391145,
      "num_tokens": 121935808.0,
      "step": 48110
    },
    {
      "entropy": 6.018998479843139,
      "epoch": 5.5527986151182915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023974926315185508,
      "loss": 5.3671,
      "mean_token_accuracy": 0.2309811681509018,
      "num_tokens": 121948511.0,
      "step": 48115
    },
    {
      "entropy": 6.071750593185425,
      "epoch": 5.553375649163301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023970850896588226,
      "loss": 5.4913,
      "mean_token_accuracy": 0.22058721780776977,
      "num_tokens": 121960540.0,
      "step": 48120
    },
    {
      "entropy": 5.987504005432129,
      "epoch": 5.553952683208309,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002396677559669212,
      "loss": 5.3577,
      "mean_token_accuracy": 0.22762912660837173,
      "num_tokens": 121973034.0,
      "step": 48125
    },
    {
      "entropy": 6.048416376113892,
      "epoch": 5.554529717253318,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023962700415634258,
      "loss": 5.4507,
      "mean_token_accuracy": 0.22076354175806046,
      "num_tokens": 121985189.0,
      "step": 48130
    },
    {
      "entropy": 6.059376525878906,
      "epoch": 5.555106751298327,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023958625353551728,
      "loss": 5.4102,
      "mean_token_accuracy": 0.22135140746831894,
      "num_tokens": 121996187.0,
      "step": 48135
    },
    {
      "entropy": 5.993706512451172,
      "epoch": 5.555683785343335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023954550410581555,
      "loss": 5.4293,
      "mean_token_accuracy": 0.22049858719110488,
      "num_tokens": 122008211.0,
      "step": 48140
    },
    {
      "entropy": 5.986296558380127,
      "epoch": 5.556260819388344,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023950475586860832,
      "loss": 5.3992,
      "mean_token_accuracy": 0.22768102288246156,
      "num_tokens": 122019779.0,
      "step": 48145
    },
    {
      "entropy": 6.114831733703613,
      "epoch": 5.556837853433352,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023946400882526587,
      "loss": 5.5148,
      "mean_token_accuracy": 0.21687258034944534,
      "num_tokens": 122031698.0,
      "step": 48150
    },
    {
      "entropy": 6.109345579147339,
      "epoch": 5.5574148874783615,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023942326297715888,
      "loss": 5.4646,
      "mean_token_accuracy": 0.21639580726623536,
      "num_tokens": 122043273.0,
      "step": 48155
    },
    {
      "entropy": 6.071987152099609,
      "epoch": 5.55799192152337,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023938251832565782,
      "loss": 5.4708,
      "mean_token_accuracy": 0.2200428992509842,
      "num_tokens": 122055848.0,
      "step": 48160
    },
    {
      "entropy": 6.062953472137451,
      "epoch": 5.558568955568378,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023934177487213305,
      "loss": 5.4559,
      "mean_token_accuracy": 0.21780041009187698,
      "num_tokens": 122069717.0,
      "step": 48165
    },
    {
      "entropy": 6.013491249084472,
      "epoch": 5.559145989613388,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023930103261795495,
      "loss": 5.4165,
      "mean_token_accuracy": 0.2225350931286812,
      "num_tokens": 122081739.0,
      "step": 48170
    },
    {
      "entropy": 6.058135652542115,
      "epoch": 5.559723023658396,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023926029156449386,
      "loss": 5.4846,
      "mean_token_accuracy": 0.2168823555111885,
      "num_tokens": 122094232.0,
      "step": 48175
    },
    {
      "entropy": 5.928902292251587,
      "epoch": 5.5603000577034045,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023921955171312016,
      "loss": 5.2735,
      "mean_token_accuracy": 0.23009622246026992,
      "num_tokens": 122107609.0,
      "step": 48180
    },
    {
      "entropy": 5.976681661605835,
      "epoch": 5.560877091748413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023917881306520405,
      "loss": 5.3341,
      "mean_token_accuracy": 0.23436030000448227,
      "num_tokens": 122119873.0,
      "step": 48185
    },
    {
      "entropy": 6.118724298477173,
      "epoch": 5.561454125793421,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023913807562211576,
      "loss": 5.516,
      "mean_token_accuracy": 0.21213323324918748,
      "num_tokens": 122133495.0,
      "step": 48190
    },
    {
      "entropy": 6.023125839233399,
      "epoch": 5.562031159838431,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023909733938522543,
      "loss": 5.474,
      "mean_token_accuracy": 0.2208007350564003,
      "num_tokens": 122146517.0,
      "step": 48195
    },
    {
      "entropy": 6.031644868850708,
      "epoch": 5.562608193883439,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023905660435590337,
      "loss": 5.44,
      "mean_token_accuracy": 0.2191479504108429,
      "num_tokens": 122158397.0,
      "step": 48200
    },
    {
      "entropy": 6.070740509033203,
      "epoch": 5.5631852279284475,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023901587053551949,
      "loss": 5.4682,
      "mean_token_accuracy": 0.215529964864254,
      "num_tokens": 122170564.0,
      "step": 48205
    },
    {
      "entropy": 6.123462390899658,
      "epoch": 5.563762261973457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023897513792544396,
      "loss": 5.4975,
      "mean_token_accuracy": 0.21910422146320344,
      "num_tokens": 122183678.0,
      "step": 48210
    },
    {
      "entropy": 6.0267009258270265,
      "epoch": 5.564339296018465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023893440652704673,
      "loss": 5.4326,
      "mean_token_accuracy": 0.22356181740760803,
      "num_tokens": 122196605.0,
      "step": 48215
    },
    {
      "entropy": 6.009166622161866,
      "epoch": 5.564916330063474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002388936763416979,
      "loss": 5.375,
      "mean_token_accuracy": 0.22460194379091264,
      "num_tokens": 122208553.0,
      "step": 48220
    },
    {
      "entropy": 6.091078090667724,
      "epoch": 5.565493364108482,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023885294737076724,
      "loss": 5.5185,
      "mean_token_accuracy": 0.21584667414426803,
      "num_tokens": 122221369.0,
      "step": 48225
    },
    {
      "entropy": 6.115704011917114,
      "epoch": 5.566070398153491,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023881221961562487,
      "loss": 5.5135,
      "mean_token_accuracy": 0.21113484352827072,
      "num_tokens": 122233834.0,
      "step": 48230
    },
    {
      "entropy": 6.061873817443848,
      "epoch": 5.5666474321985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023877149307764036,
      "loss": 5.4167,
      "mean_token_accuracy": 0.22458745986223222,
      "num_tokens": 122246931.0,
      "step": 48235
    },
    {
      "entropy": 6.025534534454346,
      "epoch": 5.567224466243508,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002387307677581838,
      "loss": 5.4283,
      "mean_token_accuracy": 0.21867556869983673,
      "num_tokens": 122260708.0,
      "step": 48240
    },
    {
      "entropy": 6.0145917415618895,
      "epoch": 5.5678015002885175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002386900436586248,
      "loss": 5.3744,
      "mean_token_accuracy": 0.22899825423955916,
      "num_tokens": 122272923.0,
      "step": 48245
    },
    {
      "entropy": 6.0559875011444095,
      "epoch": 5.568378534333526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002386493207803332,
      "loss": 5.4755,
      "mean_token_accuracy": 0.21398269087076188,
      "num_tokens": 122285175.0,
      "step": 48250
    },
    {
      "entropy": 6.02157187461853,
      "epoch": 5.568955568378534,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023860859912467851,
      "loss": 5.3589,
      "mean_token_accuracy": 0.2268316239118576,
      "num_tokens": 122298639.0,
      "step": 48255
    },
    {
      "entropy": 6.121217584609985,
      "epoch": 5.569532602423543,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023856787869303066,
      "loss": 5.4914,
      "mean_token_accuracy": 0.20833466947078705,
      "num_tokens": 122312177.0,
      "step": 48260
    },
    {
      "entropy": 5.960158586502075,
      "epoch": 5.570109636468551,
      "grad_norm": 1.171875,
      "learning_rate": 0.00023852715948675897,
      "loss": 5.4139,
      "mean_token_accuracy": 0.2261803850531578,
      "num_tokens": 122324287.0,
      "step": 48265
    },
    {
      "entropy": 6.02413272857666,
      "epoch": 5.5706866705135605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002384864415072333,
      "loss": 5.4153,
      "mean_token_accuracy": 0.22435270249843597,
      "num_tokens": 122335579.0,
      "step": 48270
    },
    {
      "entropy": 6.08406982421875,
      "epoch": 5.571263704558569,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023844572475582293,
      "loss": 5.4232,
      "mean_token_accuracy": 0.21796109676361083,
      "num_tokens": 122349095.0,
      "step": 48275
    },
    {
      "entropy": 6.0598142623901365,
      "epoch": 5.571840738603577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023840500923389752,
      "loss": 5.4363,
      "mean_token_accuracy": 0.22123051583766937,
      "num_tokens": 122360017.0,
      "step": 48280
    },
    {
      "entropy": 5.98394284248352,
      "epoch": 5.572417772648587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023836429494282634,
      "loss": 5.4147,
      "mean_token_accuracy": 0.21494778096675873,
      "num_tokens": 122372627.0,
      "step": 48285
    },
    {
      "entropy": 6.123559045791626,
      "epoch": 5.572994806693595,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023832358188397907,
      "loss": 5.4563,
      "mean_token_accuracy": 0.22776223123073577,
      "num_tokens": 122385446.0,
      "step": 48290
    },
    {
      "entropy": 6.039320373535157,
      "epoch": 5.5735718407386035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023828287005872473,
      "loss": 5.418,
      "mean_token_accuracy": 0.22595112174749374,
      "num_tokens": 122398432.0,
      "step": 48295
    },
    {
      "entropy": 6.050453233718872,
      "epoch": 5.574148874783612,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023824215946843293,
      "loss": 5.4418,
      "mean_token_accuracy": 0.22706850469112397,
      "num_tokens": 122412324.0,
      "step": 48300
    },
    {
      "entropy": 5.9739116668701175,
      "epoch": 5.574725908828621,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023820145011447274,
      "loss": 5.3668,
      "mean_token_accuracy": 0.22244611978530884,
      "num_tokens": 122424733.0,
      "step": 48305
    },
    {
      "entropy": 5.960130262374878,
      "epoch": 5.57530294287363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023816074199821364,
      "loss": 5.3941,
      "mean_token_accuracy": 0.22321381121873857,
      "num_tokens": 122436543.0,
      "step": 48310
    },
    {
      "entropy": 5.989117527008057,
      "epoch": 5.575879976918638,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023812003512102453,
      "loss": 5.3786,
      "mean_token_accuracy": 0.2229514792561531,
      "num_tokens": 122448898.0,
      "step": 48315
    },
    {
      "entropy": 6.013726902008057,
      "epoch": 5.576457010963647,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023807932948427486,
      "loss": 5.3739,
      "mean_token_accuracy": 0.2266082987189293,
      "num_tokens": 122461727.0,
      "step": 48320
    },
    {
      "entropy": 6.029468297958374,
      "epoch": 5.577034045008656,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002380386250893335,
      "loss": 5.4154,
      "mean_token_accuracy": 0.2189100831747055,
      "num_tokens": 122473656.0,
      "step": 48325
    },
    {
      "entropy": 6.0521844863891605,
      "epoch": 5.577611079053664,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002379979219375697,
      "loss": 5.4806,
      "mean_token_accuracy": 0.2151751697063446,
      "num_tokens": 122485240.0,
      "step": 48330
    },
    {
      "entropy": 6.068121099472046,
      "epoch": 5.578188113098673,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002379572200303523,
      "loss": 5.4473,
      "mean_token_accuracy": 0.22772873044013978,
      "num_tokens": 122498452.0,
      "step": 48335
    },
    {
      "entropy": 6.1030010223388675,
      "epoch": 5.578765147143681,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023791651936905058,
      "loss": 5.4599,
      "mean_token_accuracy": 0.21725086718797684,
      "num_tokens": 122511165.0,
      "step": 48340
    },
    {
      "entropy": 6.05709867477417,
      "epoch": 5.57934218118869,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002378758199550331,
      "loss": 5.4454,
      "mean_token_accuracy": 0.2280835524201393,
      "num_tokens": 122525312.0,
      "step": 48345
    },
    {
      "entropy": 6.036635017395019,
      "epoch": 5.579919215233699,
      "grad_norm": 1.0,
      "learning_rate": 0.00023783512178966916,
      "loss": 5.449,
      "mean_token_accuracy": 0.21545152515172958,
      "num_tokens": 122537629.0,
      "step": 48350
    },
    {
      "entropy": 6.073492383956909,
      "epoch": 5.580496249278707,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023779442487432734,
      "loss": 5.4622,
      "mean_token_accuracy": 0.2206231728196144,
      "num_tokens": 122551170.0,
      "step": 48355
    },
    {
      "entropy": 5.982183933258057,
      "epoch": 5.581073283323716,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023775372921037657,
      "loss": 5.3645,
      "mean_token_accuracy": 0.2302156463265419,
      "num_tokens": 122563161.0,
      "step": 48360
    },
    {
      "entropy": 6.017331552505493,
      "epoch": 5.581650317368725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002377130347991856,
      "loss": 5.3934,
      "mean_token_accuracy": 0.22583483010530472,
      "num_tokens": 122575652.0,
      "step": 48365
    },
    {
      "entropy": 6.043704605102539,
      "epoch": 5.582227351413733,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002376723416421233,
      "loss": 5.4272,
      "mean_token_accuracy": 0.21747861057519913,
      "num_tokens": 122587781.0,
      "step": 48370
    },
    {
      "entropy": 5.964584684371948,
      "epoch": 5.582804385458742,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023763164974055812,
      "loss": 5.4008,
      "mean_token_accuracy": 0.22738104313611984,
      "num_tokens": 122600600.0,
      "step": 48375
    },
    {
      "entropy": 6.088488388061523,
      "epoch": 5.583381419503751,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00023759095909585886,
      "loss": 5.4726,
      "mean_token_accuracy": 0.2209407016634941,
      "num_tokens": 122613160.0,
      "step": 48380
    },
    {
      "entropy": 6.084336805343628,
      "epoch": 5.583958453548759,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00023755026970939414,
      "loss": 5.4479,
      "mean_token_accuracy": 0.22390652298927308,
      "num_tokens": 122625104.0,
      "step": 48385
    },
    {
      "entropy": 5.947153186798095,
      "epoch": 5.584535487593768,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002375095815825325,
      "loss": 5.3215,
      "mean_token_accuracy": 0.23465052247047424,
      "num_tokens": 122638991.0,
      "step": 48390
    },
    {
      "entropy": 5.9636482238769535,
      "epoch": 5.585112521638777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002374688947166424,
      "loss": 5.3457,
      "mean_token_accuracy": 0.234133517742157,
      "num_tokens": 122652947.0,
      "step": 48395
    },
    {
      "entropy": 6.020062780380249,
      "epoch": 5.585689555683786,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002374282091130925,
      "loss": 5.397,
      "mean_token_accuracy": 0.22638286352157594,
      "num_tokens": 122667323.0,
      "step": 48400
    },
    {
      "entropy": 6.0355278015136715,
      "epoch": 5.586266589728794,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00023738752477325099,
      "loss": 5.3982,
      "mean_token_accuracy": 0.2267741620540619,
      "num_tokens": 122679972.0,
      "step": 48405
    },
    {
      "entropy": 5.960158586502075,
      "epoch": 5.586843623773802,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023734684169848648,
      "loss": 5.3831,
      "mean_token_accuracy": 0.23085542470216752,
      "num_tokens": 122693616.0,
      "step": 48410
    },
    {
      "entropy": 6.063619136810303,
      "epoch": 5.587420657818812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023730615989016725,
      "loss": 5.438,
      "mean_token_accuracy": 0.21528619825839995,
      "num_tokens": 122705596.0,
      "step": 48415
    },
    {
      "entropy": 6.012936019897461,
      "epoch": 5.58799769186382,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002372654793496616,
      "loss": 5.3741,
      "mean_token_accuracy": 0.22619224190711976,
      "num_tokens": 122718406.0,
      "step": 48420
    },
    {
      "entropy": 6.049465560913086,
      "epoch": 5.5885747259088285,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023722480007833775,
      "loss": 5.4947,
      "mean_token_accuracy": 0.22253556847572326,
      "num_tokens": 122730693.0,
      "step": 48425
    },
    {
      "entropy": 6.002363586425782,
      "epoch": 5.589151759953837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023718412207756408,
      "loss": 5.4214,
      "mean_token_accuracy": 0.22305867224931716,
      "num_tokens": 122742353.0,
      "step": 48430
    },
    {
      "entropy": 6.089612340927124,
      "epoch": 5.589728793998846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023714344534870858,
      "loss": 5.4666,
      "mean_token_accuracy": 0.22018018513917922,
      "num_tokens": 122754993.0,
      "step": 48435
    },
    {
      "entropy": 6.06230354309082,
      "epoch": 5.590305828043855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023710276989313955,
      "loss": 5.4404,
      "mean_token_accuracy": 0.21110750436782838,
      "num_tokens": 122768154.0,
      "step": 48440
    },
    {
      "entropy": 6.028749275207519,
      "epoch": 5.590882862088863,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023706209571222497,
      "loss": 5.4508,
      "mean_token_accuracy": 0.2165164276957512,
      "num_tokens": 122780895.0,
      "step": 48445
    },
    {
      "entropy": 5.8899320602417,
      "epoch": 5.5914598961338715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023702142280733308,
      "loss": 5.2928,
      "mean_token_accuracy": 0.23407047092914582,
      "num_tokens": 122794439.0,
      "step": 48450
    },
    {
      "entropy": 5.992277908325195,
      "epoch": 5.592036930178881,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002369807511798317,
      "loss": 5.3148,
      "mean_token_accuracy": 0.2304818481206894,
      "num_tokens": 122807089.0,
      "step": 48455
    },
    {
      "entropy": 6.074002027511597,
      "epoch": 5.592613964223889,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023694008083108886,
      "loss": 5.4756,
      "mean_token_accuracy": 0.2115398555994034,
      "num_tokens": 122819767.0,
      "step": 48460
    },
    {
      "entropy": 6.094053745269775,
      "epoch": 5.593190998268898,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023689941176247242,
      "loss": 5.4612,
      "mean_token_accuracy": 0.21423859596252443,
      "num_tokens": 122831260.0,
      "step": 48465
    },
    {
      "entropy": 6.1231413841247555,
      "epoch": 5.593768032313907,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002368587439753505,
      "loss": 5.5252,
      "mean_token_accuracy": 0.22157291918992997,
      "num_tokens": 122843516.0,
      "step": 48470
    },
    {
      "entropy": 5.909885120391846,
      "epoch": 5.594345066358915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023681807747109068,
      "loss": 5.3158,
      "mean_token_accuracy": 0.23961429595947265,
      "num_tokens": 122857791.0,
      "step": 48475
    },
    {
      "entropy": 5.983511352539063,
      "epoch": 5.594922100403924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023677741225106092,
      "loss": 5.3859,
      "mean_token_accuracy": 0.2277877390384674,
      "num_tokens": 122870179.0,
      "step": 48480
    },
    {
      "entropy": 6.062080049514771,
      "epoch": 5.595499134448932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023673674831662877,
      "loss": 5.459,
      "mean_token_accuracy": 0.22217997759580613,
      "num_tokens": 122883450.0,
      "step": 48485
    },
    {
      "entropy": 6.010832786560059,
      "epoch": 5.5960761684939415,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00023669608566916227,
      "loss": 5.4137,
      "mean_token_accuracy": 0.2213813528418541,
      "num_tokens": 122896381.0,
      "step": 48490
    },
    {
      "entropy": 6.075430154800415,
      "epoch": 5.59665320253895,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002366554243100288,
      "loss": 5.4337,
      "mean_token_accuracy": 0.2199198216199875,
      "num_tokens": 122908822.0,
      "step": 48495
    },
    {
      "entropy": 6.124392461776734,
      "epoch": 5.597230236583958,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023661476424059614,
      "loss": 5.5358,
      "mean_token_accuracy": 0.20775512903928756,
      "num_tokens": 122920551.0,
      "step": 48500
    },
    {
      "entropy": 6.135513544082642,
      "epoch": 5.597807270628967,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023657410546223175,
      "loss": 5.5224,
      "mean_token_accuracy": 0.20922305434942245,
      "num_tokens": 122933180.0,
      "step": 48505
    },
    {
      "entropy": 6.0686439037322994,
      "epoch": 5.598384304673976,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023653344797630333,
      "loss": 5.4966,
      "mean_token_accuracy": 0.20996746122837068,
      "num_tokens": 122944701.0,
      "step": 48510
    },
    {
      "entropy": 6.074875593185425,
      "epoch": 5.5989613387189845,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023649279178417816,
      "loss": 5.4703,
      "mean_token_accuracy": 0.2156743362545967,
      "num_tokens": 122957748.0,
      "step": 48515
    },
    {
      "entropy": 6.031651639938355,
      "epoch": 5.599538372763993,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002364521368872239,
      "loss": 5.3483,
      "mean_token_accuracy": 0.23394380062818526,
      "num_tokens": 122970789.0,
      "step": 48520
    },
    {
      "entropy": 6.035295200347901,
      "epoch": 5.600115406809001,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023641148328680777,
      "loss": 5.4263,
      "mean_token_accuracy": 0.22339330315589906,
      "num_tokens": 122983900.0,
      "step": 48525
    },
    {
      "entropy": 6.081436395645142,
      "epoch": 5.600692440854011,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023637083098429734,
      "loss": 5.4952,
      "mean_token_accuracy": 0.22286250740289687,
      "num_tokens": 122997531.0,
      "step": 48530
    },
    {
      "entropy": 6.044224786758423,
      "epoch": 5.601269474899019,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00023633017998105977,
      "loss": 5.4771,
      "mean_token_accuracy": 0.22007400095462798,
      "num_tokens": 123010568.0,
      "step": 48535
    },
    {
      "entropy": 6.055957651138305,
      "epoch": 5.6018465089440275,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002362895302784624,
      "loss": 5.5211,
      "mean_token_accuracy": 0.21879362165927888,
      "num_tokens": 123023645.0,
      "step": 48540
    },
    {
      "entropy": 6.060488843917847,
      "epoch": 5.602423542989037,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023624888187787242,
      "loss": 5.4451,
      "mean_token_accuracy": 0.21792462319135666,
      "num_tokens": 123036529.0,
      "step": 48545
    },
    {
      "entropy": 5.933151292800903,
      "epoch": 5.603000577034045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023620823478065706,
      "loss": 5.2873,
      "mean_token_accuracy": 0.23199762254953385,
      "num_tokens": 123049746.0,
      "step": 48550
    },
    {
      "entropy": 5.95906662940979,
      "epoch": 5.603577611079054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023616758898818352,
      "loss": 5.3554,
      "mean_token_accuracy": 0.23741820305585862,
      "num_tokens": 123061320.0,
      "step": 48555
    },
    {
      "entropy": 6.048046970367432,
      "epoch": 5.604154645124062,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023612694450181877,
      "loss": 5.4331,
      "mean_token_accuracy": 0.22897976040840148,
      "num_tokens": 123073113.0,
      "step": 48560
    },
    {
      "entropy": 6.043319511413574,
      "epoch": 5.604731679169071,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023608630132292992,
      "loss": 5.4742,
      "mean_token_accuracy": 0.217577263712883,
      "num_tokens": 123085820.0,
      "step": 48565
    },
    {
      "entropy": 6.075659322738647,
      "epoch": 5.60530871321408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00023604565945288403,
      "loss": 5.491,
      "mean_token_accuracy": 0.22011861801147461,
      "num_tokens": 123099676.0,
      "step": 48570
    },
    {
      "entropy": 6.0593139171600345,
      "epoch": 5.605885747259088,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023600501889304808,
      "loss": 5.4581,
      "mean_token_accuracy": 0.2181072399020195,
      "num_tokens": 123111866.0,
      "step": 48575
    },
    {
      "entropy": 6.079110336303711,
      "epoch": 5.606462781304097,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00023596437964478885,
      "loss": 5.4577,
      "mean_token_accuracy": 0.21434439569711686,
      "num_tokens": 123123177.0,
      "step": 48580
    },
    {
      "entropy": 5.996688508987427,
      "epoch": 5.607039815349106,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002359237417094734,
      "loss": 5.4172,
      "mean_token_accuracy": 0.22447736114263533,
      "num_tokens": 123135934.0,
      "step": 48585
    },
    {
      "entropy": 6.040232706069946,
      "epoch": 5.607616849394114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023588310508846844,
      "loss": 5.3918,
      "mean_token_accuracy": 0.22378274351358413,
      "num_tokens": 123147920.0,
      "step": 48590
    },
    {
      "entropy": 6.031450891494751,
      "epoch": 5.608193883439123,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023584246978314088,
      "loss": 5.4343,
      "mean_token_accuracy": 0.22294647097587586,
      "num_tokens": 123161903.0,
      "step": 48595
    },
    {
      "entropy": 6.045877885818482,
      "epoch": 5.608770917484131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023580183579485726,
      "loss": 5.4228,
      "mean_token_accuracy": 0.2190336748957634,
      "num_tokens": 123175438.0,
      "step": 48600
    },
    {
      "entropy": 6.0033563613891605,
      "epoch": 5.6093479515291405,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002357612031249846,
      "loss": 5.4301,
      "mean_token_accuracy": 0.22248134016990662,
      "num_tokens": 123188039.0,
      "step": 48605
    },
    {
      "entropy": 6.026626873016357,
      "epoch": 5.609924985574149,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002357205717748892,
      "loss": 5.4696,
      "mean_token_accuracy": 0.21874606609344482,
      "num_tokens": 123200202.0,
      "step": 48610
    },
    {
      "entropy": 6.088913059234619,
      "epoch": 5.610502019619157,
      "grad_norm": 1.03125,
      "learning_rate": 0.000235679941745938,
      "loss": 5.4973,
      "mean_token_accuracy": 0.21340277493000032,
      "num_tokens": 123212000.0,
      "step": 48615
    },
    {
      "entropy": 6.115221071243286,
      "epoch": 5.611079053664167,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023563931303949732,
      "loss": 5.534,
      "mean_token_accuracy": 0.21489989012479782,
      "num_tokens": 123225007.0,
      "step": 48620
    },
    {
      "entropy": 6.033494377136231,
      "epoch": 5.611656087709175,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023559868565693393,
      "loss": 5.3796,
      "mean_token_accuracy": 0.2395218625664711,
      "num_tokens": 123239252.0,
      "step": 48625
    },
    {
      "entropy": 5.988898801803589,
      "epoch": 5.612233121754183,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023555805959961406,
      "loss": 5.3729,
      "mean_token_accuracy": 0.22880322784185408,
      "num_tokens": 123251505.0,
      "step": 48630
    },
    {
      "entropy": 6.0421360492706295,
      "epoch": 5.612810155799192,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023551743486890432,
      "loss": 5.3978,
      "mean_token_accuracy": 0.22480324059724807,
      "num_tokens": 123262607.0,
      "step": 48635
    },
    {
      "entropy": 5.971423530578614,
      "epoch": 5.613387189844201,
      "grad_norm": 1.140625,
      "learning_rate": 0.000235476811466171,
      "loss": 5.3022,
      "mean_token_accuracy": 0.23616798222064972,
      "num_tokens": 123273674.0,
      "step": 48640
    },
    {
      "entropy": 6.050618457794189,
      "epoch": 5.61396422388921,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023543618939278057,
      "loss": 5.4549,
      "mean_token_accuracy": 0.22302118837833404,
      "num_tokens": 123286638.0,
      "step": 48645
    },
    {
      "entropy": 6.0823108673095705,
      "epoch": 5.614541257934218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002353955686500992,
      "loss": 5.4453,
      "mean_token_accuracy": 0.22052307277917862,
      "num_tokens": 123299587.0,
      "step": 48650
    },
    {
      "entropy": 6.052453517913818,
      "epoch": 5.615118291979226,
      "grad_norm": 1.0,
      "learning_rate": 0.0002353549492394932,
      "loss": 5.4798,
      "mean_token_accuracy": 0.22392387986183165,
      "num_tokens": 123312511.0,
      "step": 48655
    },
    {
      "entropy": 6.020852565765381,
      "epoch": 5.615695326024236,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023531433116232876,
      "loss": 5.3954,
      "mean_token_accuracy": 0.23296763896942138,
      "num_tokens": 123324711.0,
      "step": 48660
    },
    {
      "entropy": 6.041205835342407,
      "epoch": 5.616272360069244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002352737144199722,
      "loss": 5.382,
      "mean_token_accuracy": 0.22759367674589157,
      "num_tokens": 123336940.0,
      "step": 48665
    },
    {
      "entropy": 6.0020036697387695,
      "epoch": 5.6168493941142525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002352330990137894,
      "loss": 5.356,
      "mean_token_accuracy": 0.22978801131248475,
      "num_tokens": 123349107.0,
      "step": 48670
    },
    {
      "entropy": 6.009737396240235,
      "epoch": 5.617426428159261,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002351924849451466,
      "loss": 5.436,
      "mean_token_accuracy": 0.22415084689855574,
      "num_tokens": 123361710.0,
      "step": 48675
    },
    {
      "entropy": 6.0294764041900635,
      "epoch": 5.61800346220427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002351518722154098,
      "loss": 5.4043,
      "mean_token_accuracy": 0.21810402572155,
      "num_tokens": 123373716.0,
      "step": 48680
    },
    {
      "entropy": 6.059708738327027,
      "epoch": 5.618580496249279,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023511126082594502,
      "loss": 5.5482,
      "mean_token_accuracy": 0.21546269804239274,
      "num_tokens": 123389052.0,
      "step": 48685
    },
    {
      "entropy": 5.957034683227539,
      "epoch": 5.619157530294287,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023507065077811812,
      "loss": 5.2727,
      "mean_token_accuracy": 0.229182368516922,
      "num_tokens": 123401716.0,
      "step": 48690
    },
    {
      "entropy": 6.053309059143066,
      "epoch": 5.619734564339296,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023503004207329505,
      "loss": 5.4568,
      "mean_token_accuracy": 0.2248503476381302,
      "num_tokens": 123414462.0,
      "step": 48695
    },
    {
      "entropy": 6.066244411468506,
      "epoch": 5.620311598384305,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023498943471284163,
      "loss": 5.3851,
      "mean_token_accuracy": 0.2312533065676689,
      "num_tokens": 123426369.0,
      "step": 48700
    },
    {
      "entropy": 6.064543104171753,
      "epoch": 5.620888632429313,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023494882869812378,
      "loss": 5.4579,
      "mean_token_accuracy": 0.21964575946331025,
      "num_tokens": 123438404.0,
      "step": 48705
    },
    {
      "entropy": 5.984299039840698,
      "epoch": 5.621465666474322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023490822403050715,
      "loss": 5.3476,
      "mean_token_accuracy": 0.22685596495866775,
      "num_tokens": 123450552.0,
      "step": 48710
    },
    {
      "entropy": 5.9882887840271,
      "epoch": 5.622042700519331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002348676207113575,
      "loss": 5.3744,
      "mean_token_accuracy": 0.22609336525201798,
      "num_tokens": 123465061.0,
      "step": 48715
    },
    {
      "entropy": 5.98761215209961,
      "epoch": 5.622619734564339,
      "grad_norm": 0.9375,
      "learning_rate": 0.00023482701874204042,
      "loss": 5.3811,
      "mean_token_accuracy": 0.2279409572482109,
      "num_tokens": 123477189.0,
      "step": 48720
    },
    {
      "entropy": 6.022925519943238,
      "epoch": 5.623196768609348,
      "grad_norm": 1.0,
      "learning_rate": 0.0002347864181239217,
      "loss": 5.4072,
      "mean_token_accuracy": 0.21891028881073,
      "num_tokens": 123489334.0,
      "step": 48725
    },
    {
      "entropy": 6.086496257781983,
      "epoch": 5.623773802654356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023474581885836676,
      "loss": 5.5844,
      "mean_token_accuracy": 0.21108290553092957,
      "num_tokens": 123502187.0,
      "step": 48730
    },
    {
      "entropy": 6.0198976516723635,
      "epoch": 5.6243508366993655,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002347052209467413,
      "loss": 5.4007,
      "mean_token_accuracy": 0.22978236377239228,
      "num_tokens": 123516912.0,
      "step": 48735
    },
    {
      "entropy": 6.13423719406128,
      "epoch": 5.624927870744374,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002346646243904106,
      "loss": 5.5218,
      "mean_token_accuracy": 0.2146674707531929,
      "num_tokens": 123528649.0,
      "step": 48740
    },
    {
      "entropy": 5.993073463439941,
      "epoch": 5.625504904789382,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023462402919074033,
      "loss": 5.2912,
      "mean_token_accuracy": 0.23484388887882232,
      "num_tokens": 123540881.0,
      "step": 48745
    },
    {
      "entropy": 5.997282218933106,
      "epoch": 5.626081938834391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002345834353490957,
      "loss": 5.3545,
      "mean_token_accuracy": 0.22546696215867995,
      "num_tokens": 123553725.0,
      "step": 48750
    },
    {
      "entropy": 5.987742280960083,
      "epoch": 5.6266589728794,
      "grad_norm": 1.28125,
      "learning_rate": 0.00023454284286684218,
      "loss": 5.353,
      "mean_token_accuracy": 0.23213544338941575,
      "num_tokens": 123566078.0,
      "step": 48755
    },
    {
      "entropy": 5.9443034648895265,
      "epoch": 5.6272360069244085,
      "grad_norm": 0.890625,
      "learning_rate": 0.00023450225174534501,
      "loss": 5.3991,
      "mean_token_accuracy": 0.2314183384180069,
      "num_tokens": 123579767.0,
      "step": 48760
    },
    {
      "entropy": 5.916303968429565,
      "epoch": 5.627813040969417,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00023446166198596964,
      "loss": 5.3162,
      "mean_token_accuracy": 0.23134373128414154,
      "num_tokens": 123591940.0,
      "step": 48765
    },
    {
      "entropy": 6.0680896759033205,
      "epoch": 5.628390075014426,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023442107359008093,
      "loss": 5.4147,
      "mean_token_accuracy": 0.22324239313602448,
      "num_tokens": 123604110.0,
      "step": 48770
    },
    {
      "entropy": 6.071094942092896,
      "epoch": 5.628967109059435,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023438048655904437,
      "loss": 5.3931,
      "mean_token_accuracy": 0.22426164895296097,
      "num_tokens": 123616506.0,
      "step": 48775
    },
    {
      "entropy": 6.012036228179932,
      "epoch": 5.629544143104443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002343399008942249,
      "loss": 5.3416,
      "mean_token_accuracy": 0.2312915414571762,
      "num_tokens": 123628790.0,
      "step": 48780
    },
    {
      "entropy": 6.047964763641358,
      "epoch": 5.6301211771494515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002342993165969878,
      "loss": 5.4354,
      "mean_token_accuracy": 0.21887692511081697,
      "num_tokens": 123641395.0,
      "step": 48785
    },
    {
      "entropy": 6.126016521453858,
      "epoch": 5.630698211194461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023425873366869775,
      "loss": 5.5492,
      "mean_token_accuracy": 0.21036146432161332,
      "num_tokens": 123652433.0,
      "step": 48790
    },
    {
      "entropy": 6.082267427444458,
      "epoch": 5.631275245239469,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023421815211072016,
      "loss": 5.4921,
      "mean_token_accuracy": 0.20746401697397232,
      "num_tokens": 123664021.0,
      "step": 48795
    },
    {
      "entropy": 6.096015357971192,
      "epoch": 5.631852279284478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023417757192441963,
      "loss": 5.4937,
      "mean_token_accuracy": 0.21198517233133315,
      "num_tokens": 123676476.0,
      "step": 48800
    },
    {
      "entropy": 6.104717159271241,
      "epoch": 5.632429313329487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002341369931111613,
      "loss": 5.4843,
      "mean_token_accuracy": 0.21767588704824448,
      "num_tokens": 123688232.0,
      "step": 48805
    },
    {
      "entropy": 6.084763431549073,
      "epoch": 5.633006347374495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023409641567230982,
      "loss": 5.5039,
      "mean_token_accuracy": 0.22262796461582185,
      "num_tokens": 123700693.0,
      "step": 48810
    },
    {
      "entropy": 6.023290729522705,
      "epoch": 5.633583381419504,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002340558396092301,
      "loss": 5.3816,
      "mean_token_accuracy": 0.22340599000453948,
      "num_tokens": 123712874.0,
      "step": 48815
    },
    {
      "entropy": 6.000479030609131,
      "epoch": 5.634160415464512,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002340152649232869,
      "loss": 5.3362,
      "mean_token_accuracy": 0.2256360560655594,
      "num_tokens": 123725012.0,
      "step": 48820
    },
    {
      "entropy": 6.072154712677002,
      "epoch": 5.634737449509521,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023397469161584494,
      "loss": 5.4333,
      "mean_token_accuracy": 0.21912608593702315,
      "num_tokens": 123737339.0,
      "step": 48825
    },
    {
      "entropy": 5.912534475326538,
      "epoch": 5.63531448355453,
      "grad_norm": 1.0,
      "learning_rate": 0.00023393411968826878,
      "loss": 5.305,
      "mean_token_accuracy": 0.23952589631080629,
      "num_tokens": 123750024.0,
      "step": 48830
    },
    {
      "entropy": 6.008257913589477,
      "epoch": 5.635891517599538,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023389354914192311,
      "loss": 5.4507,
      "mean_token_accuracy": 0.22470761835575104,
      "num_tokens": 123764051.0,
      "step": 48835
    },
    {
      "entropy": 6.007625865936279,
      "epoch": 5.636468551644547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002338529799781725,
      "loss": 5.3723,
      "mean_token_accuracy": 0.2204154297709465,
      "num_tokens": 123776952.0,
      "step": 48840
    },
    {
      "entropy": 5.956822156906128,
      "epoch": 5.637045585689556,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00023381241219838156,
      "loss": 5.3695,
      "mean_token_accuracy": 0.22407499998807906,
      "num_tokens": 123790476.0,
      "step": 48845
    },
    {
      "entropy": 6.05647463798523,
      "epoch": 5.6376226197345645,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023377184580391453,
      "loss": 5.45,
      "mean_token_accuracy": 0.21693085134029388,
      "num_tokens": 123802390.0,
      "step": 48850
    },
    {
      "entropy": 5.997595596313476,
      "epoch": 5.638199653779573,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002337312807961361,
      "loss": 5.451,
      "mean_token_accuracy": 0.21662611216306688,
      "num_tokens": 123816434.0,
      "step": 48855
    },
    {
      "entropy": 6.021866750717163,
      "epoch": 5.638776687824581,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023369071717641048,
      "loss": 5.3808,
      "mean_token_accuracy": 0.23188189715147017,
      "num_tokens": 123829036.0,
      "step": 48860
    },
    {
      "entropy": 5.8233500003814695,
      "epoch": 5.639353721869591,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00023365015494610213,
      "loss": 5.2539,
      "mean_token_accuracy": 0.23892807513475417,
      "num_tokens": 123843321.0,
      "step": 48865
    },
    {
      "entropy": 6.051720094680786,
      "epoch": 5.639930755914599,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023360959410657523,
      "loss": 5.4926,
      "mean_token_accuracy": 0.2142215132713318,
      "num_tokens": 123856834.0,
      "step": 48870
    },
    {
      "entropy": 6.060619688034057,
      "epoch": 5.640507789959607,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023356903465919404,
      "loss": 5.4241,
      "mean_token_accuracy": 0.2278833732008934,
      "num_tokens": 123869209.0,
      "step": 48875
    },
    {
      "entropy": 6.083366394042969,
      "epoch": 5.641084824004617,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023352847660532283,
      "loss": 5.4145,
      "mean_token_accuracy": 0.2221358209848404,
      "num_tokens": 123881464.0,
      "step": 48880
    },
    {
      "entropy": 6.039099597930909,
      "epoch": 5.641661858049625,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023348791994632574,
      "loss": 5.4163,
      "mean_token_accuracy": 0.22200752496719361,
      "num_tokens": 123893215.0,
      "step": 48885
    },
    {
      "entropy": 6.003465032577514,
      "epoch": 5.642238892094634,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002334473646835668,
      "loss": 5.3195,
      "mean_token_accuracy": 0.23744122982025145,
      "num_tokens": 123906836.0,
      "step": 48890
    },
    {
      "entropy": 6.01914348602295,
      "epoch": 5.642815926139642,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023340681081841016,
      "loss": 5.4539,
      "mean_token_accuracy": 0.21955888867378234,
      "num_tokens": 123920008.0,
      "step": 48895
    },
    {
      "entropy": 6.072041034698486,
      "epoch": 5.64339296018465,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002333662583522197,
      "loss": 5.4592,
      "mean_token_accuracy": 0.22103455364704133,
      "num_tokens": 123932715.0,
      "step": 48900
    },
    {
      "entropy": 6.104324197769165,
      "epoch": 5.64396999422966,
      "grad_norm": 1.0,
      "learning_rate": 0.00023332570728635955,
      "loss": 5.4926,
      "mean_token_accuracy": 0.218172025680542,
      "num_tokens": 123945791.0,
      "step": 48905
    },
    {
      "entropy": 6.018582010269165,
      "epoch": 5.644547028274668,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023328515762219344,
      "loss": 5.4619,
      "mean_token_accuracy": 0.22064395546913146,
      "num_tokens": 123957889.0,
      "step": 48910
    },
    {
      "entropy": 6.029186105728149,
      "epoch": 5.645124062319677,
      "grad_norm": 1.0,
      "learning_rate": 0.00023324460936108537,
      "loss": 5.3675,
      "mean_token_accuracy": 0.22603900730609894,
      "num_tokens": 123969970.0,
      "step": 48915
    },
    {
      "entropy": 6.056659746170044,
      "epoch": 5.645701096364686,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023320406250439913,
      "loss": 5.446,
      "mean_token_accuracy": 0.22046137005090713,
      "num_tokens": 123982008.0,
      "step": 48920
    },
    {
      "entropy": 6.073990678787231,
      "epoch": 5.646278130409694,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023316351705349853,
      "loss": 5.4677,
      "mean_token_accuracy": 0.21452061086893082,
      "num_tokens": 123994539.0,
      "step": 48925
    },
    {
      "entropy": 6.073710250854492,
      "epoch": 5.646855164454703,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00023312297300974715,
      "loss": 5.5101,
      "mean_token_accuracy": 0.21114504039287568,
      "num_tokens": 124007875.0,
      "step": 48930
    },
    {
      "entropy": 6.011294507980347,
      "epoch": 5.647432198499711,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002330824303745089,
      "loss": 5.3979,
      "mean_token_accuracy": 0.22747056931257248,
      "num_tokens": 124020689.0,
      "step": 48935
    },
    {
      "entropy": 6.06277871131897,
      "epoch": 5.64800923254472,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023304188914914715,
      "loss": 5.3898,
      "mean_token_accuracy": 0.22909726053476334,
      "num_tokens": 124033055.0,
      "step": 48940
    },
    {
      "entropy": 6.006246089935303,
      "epoch": 5.648586266589729,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002330013493350257,
      "loss": 5.4385,
      "mean_token_accuracy": 0.2278904914855957,
      "num_tokens": 124046328.0,
      "step": 48945
    },
    {
      "entropy": 6.0035881996154785,
      "epoch": 5.649163300634737,
      "grad_norm": 1.0625,
      "learning_rate": 0.00023296081093350797,
      "loss": 5.4012,
      "mean_token_accuracy": 0.22369564920663834,
      "num_tokens": 124059864.0,
      "step": 48950
    },
    {
      "entropy": 6.016777229309082,
      "epoch": 5.649740334679747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023292027394595743,
      "loss": 5.3591,
      "mean_token_accuracy": 0.23101101368665694,
      "num_tokens": 124072005.0,
      "step": 48955
    },
    {
      "entropy": 6.050408124923706,
      "epoch": 5.650317368724755,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023287973837373766,
      "loss": 5.4565,
      "mean_token_accuracy": 0.22186203449964523,
      "num_tokens": 124085338.0,
      "step": 48960
    },
    {
      "entropy": 6.033575868606567,
      "epoch": 5.650894402769763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023283920421821193,
      "loss": 5.4052,
      "mean_token_accuracy": 0.2284867972135544,
      "num_tokens": 124098735.0,
      "step": 48965
    },
    {
      "entropy": 6.003314685821533,
      "epoch": 5.651471436814772,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002327986714807437,
      "loss": 5.4671,
      "mean_token_accuracy": 0.21685291081666946,
      "num_tokens": 124112452.0,
      "step": 48970
    },
    {
      "entropy": 6.068099641799927,
      "epoch": 5.65204847085978,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023275814016269608,
      "loss": 5.4846,
      "mean_token_accuracy": 0.22210833430290222,
      "num_tokens": 124127107.0,
      "step": 48975
    },
    {
      "entropy": 6.079826211929321,
      "epoch": 5.6526255049047895,
      "grad_norm": 1.0,
      "learning_rate": 0.00023271761026543258,
      "loss": 5.3971,
      "mean_token_accuracy": 0.23116786479949952,
      "num_tokens": 124139866.0,
      "step": 48980
    },
    {
      "entropy": 5.978127098083496,
      "epoch": 5.653202538949798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023267708179031604,
      "loss": 5.3838,
      "mean_token_accuracy": 0.21716877967119216,
      "num_tokens": 124152476.0,
      "step": 48985
    },
    {
      "entropy": 5.948775053024292,
      "epoch": 5.653779572994806,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023263655473871003,
      "loss": 5.3268,
      "mean_token_accuracy": 0.2256606176495552,
      "num_tokens": 124163903.0,
      "step": 48990
    },
    {
      "entropy": 6.05572280883789,
      "epoch": 5.654356607039816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023259602911197735,
      "loss": 5.4437,
      "mean_token_accuracy": 0.21947384178638457,
      "num_tokens": 124175779.0,
      "step": 48995
    },
    {
      "entropy": 6.058055686950683,
      "epoch": 5.654933641084824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023255550491148126,
      "loss": 5.4506,
      "mean_token_accuracy": 0.22696993499994278,
      "num_tokens": 124187877.0,
      "step": 49000
    },
    {
      "entropy": 6.019927215576172,
      "epoch": 5.6555106751298325,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023251498213858456,
      "loss": 5.3285,
      "mean_token_accuracy": 0.23231821805238723,
      "num_tokens": 124200748.0,
      "step": 49005
    },
    {
      "entropy": 6.035115957260132,
      "epoch": 5.656087709174841,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002324744607946504,
      "loss": 5.4808,
      "mean_token_accuracy": 0.21874619424343109,
      "num_tokens": 124212912.0,
      "step": 49010
    },
    {
      "entropy": 6.043200254440308,
      "epoch": 5.65666474321985,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023243394088104158,
      "loss": 5.4261,
      "mean_token_accuracy": 0.21857526600360871,
      "num_tokens": 124224761.0,
      "step": 49015
    },
    {
      "entropy": 6.10180401802063,
      "epoch": 5.657241777264859,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002323934223991211,
      "loss": 5.484,
      "mean_token_accuracy": 0.2132047474384308,
      "num_tokens": 124236482.0,
      "step": 49020
    },
    {
      "entropy": 6.102551460266113,
      "epoch": 5.657818811309867,
      "grad_norm": 1.0,
      "learning_rate": 0.0002323529053502516,
      "loss": 5.5545,
      "mean_token_accuracy": 0.20767866671085358,
      "num_tokens": 124249619.0,
      "step": 49025
    },
    {
      "entropy": 6.080642414093018,
      "epoch": 5.658395845354876,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000232312389735796,
      "loss": 5.4679,
      "mean_token_accuracy": 0.2189650282263756,
      "num_tokens": 124262569.0,
      "step": 49030
    },
    {
      "entropy": 6.040057849884033,
      "epoch": 5.658972879399885,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023227187555711692,
      "loss": 5.3473,
      "mean_token_accuracy": 0.22882839292287827,
      "num_tokens": 124274302.0,
      "step": 49035
    },
    {
      "entropy": 5.952656364440918,
      "epoch": 5.659549913444893,
      "grad_norm": 1.15625,
      "learning_rate": 0.00023223136281557723,
      "loss": 5.3513,
      "mean_token_accuracy": 0.2263977512717247,
      "num_tokens": 124285978.0,
      "step": 49040
    },
    {
      "entropy": 6.098104953765869,
      "epoch": 5.660126947489902,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002321908515125392,
      "loss": 5.4576,
      "mean_token_accuracy": 0.21454819440841674,
      "num_tokens": 124297947.0,
      "step": 49045
    },
    {
      "entropy": 6.087552309036255,
      "epoch": 5.660703981534911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023215034164936576,
      "loss": 5.4298,
      "mean_token_accuracy": 0.2161904200911522,
      "num_tokens": 124310612.0,
      "step": 49050
    },
    {
      "entropy": 5.984824562072754,
      "epoch": 5.661281015579919,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002321098332274192,
      "loss": 5.4039,
      "mean_token_accuracy": 0.22638904750347139,
      "num_tokens": 124322605.0,
      "step": 49055
    },
    {
      "entropy": 5.971964693069458,
      "epoch": 5.661858049624928,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002320693262480622,
      "loss": 5.4085,
      "mean_token_accuracy": 0.2309669479727745,
      "num_tokens": 124334859.0,
      "step": 49060
    },
    {
      "entropy": 6.085665369033814,
      "epoch": 5.662435083669936,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023202882071265702,
      "loss": 5.4456,
      "mean_token_accuracy": 0.22254098802804947,
      "num_tokens": 124346565.0,
      "step": 49065
    },
    {
      "entropy": 6.018671941757202,
      "epoch": 5.6630121177149455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023198831662256614,
      "loss": 5.3427,
      "mean_token_accuracy": 0.22837089002132416,
      "num_tokens": 124360539.0,
      "step": 49070
    },
    {
      "entropy": 6.0850873470306395,
      "epoch": 5.663589151759954,
      "grad_norm": 1.125,
      "learning_rate": 0.00023194781397915178,
      "loss": 5.481,
      "mean_token_accuracy": 0.2100539892911911,
      "num_tokens": 124371414.0,
      "step": 49075
    },
    {
      "entropy": 6.0110148906707765,
      "epoch": 5.664166185804962,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00023190731278377642,
      "loss": 5.4549,
      "mean_token_accuracy": 0.21831038743257522,
      "num_tokens": 124383433.0,
      "step": 49080
    },
    {
      "entropy": 6.000651073455811,
      "epoch": 5.664743219849971,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002318668130378021,
      "loss": 5.3831,
      "mean_token_accuracy": 0.23168915510177612,
      "num_tokens": 124396567.0,
      "step": 49085
    },
    {
      "entropy": 5.98125958442688,
      "epoch": 5.66532025389498,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023182631474259115,
      "loss": 5.3514,
      "mean_token_accuracy": 0.22967492789030075,
      "num_tokens": 124408893.0,
      "step": 49090
    },
    {
      "entropy": 5.910297822952271,
      "epoch": 5.6658972879399885,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00023178581789950564,
      "loss": 5.3682,
      "mean_token_accuracy": 0.23053745180368423,
      "num_tokens": 124422996.0,
      "step": 49095
    },
    {
      "entropy": 6.0870767593383786,
      "epoch": 5.666474321984997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00023174532250990777,
      "loss": 5.4467,
      "mean_token_accuracy": 0.21906241476535798,
      "num_tokens": 124435722.0,
      "step": 49100
    },
    {
      "entropy": 5.959550142288208,
      "epoch": 5.667051356030006,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023170482857515933,
      "loss": 5.2997,
      "mean_token_accuracy": 0.2282679095864296,
      "num_tokens": 124450820.0,
      "step": 49105
    },
    {
      "entropy": 6.088429260253906,
      "epoch": 5.667628390075015,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023166433609662256,
      "loss": 5.4841,
      "mean_token_accuracy": 0.21828382164239885,
      "num_tokens": 124464083.0,
      "step": 49110
    },
    {
      "entropy": 6.0429881572723385,
      "epoch": 5.668205424120023,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002316238450756593,
      "loss": 5.4283,
      "mean_token_accuracy": 0.2175286054611206,
      "num_tokens": 124475730.0,
      "step": 49115
    },
    {
      "entropy": 6.073699903488159,
      "epoch": 5.6687824581650315,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023158335551363152,
      "loss": 5.4501,
      "mean_token_accuracy": 0.221943998336792,
      "num_tokens": 124488901.0,
      "step": 49120
    },
    {
      "entropy": 6.075283765792847,
      "epoch": 5.669359492210041,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00023154286741190096,
      "loss": 5.4947,
      "mean_token_accuracy": 0.2154714584350586,
      "num_tokens": 124501256.0,
      "step": 49125
    },
    {
      "entropy": 6.036823654174805,
      "epoch": 5.669936526255049,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002315023807718295,
      "loss": 5.4528,
      "mean_token_accuracy": 0.22234490811824797,
      "num_tokens": 124513395.0,
      "step": 49130
    },
    {
      "entropy": 6.053715181350708,
      "epoch": 5.670513560300058,
      "grad_norm": 0.953125,
      "learning_rate": 0.00023146189559477876,
      "loss": 5.4162,
      "mean_token_accuracy": 0.2245945706963539,
      "num_tokens": 124525440.0,
      "step": 49135
    },
    {
      "entropy": 6.05609540939331,
      "epoch": 5.671090594345066,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00023142141188211069,
      "loss": 5.4571,
      "mean_token_accuracy": 0.21906668394804002,
      "num_tokens": 124538437.0,
      "step": 49140
    },
    {
      "entropy": 6.041209650039673,
      "epoch": 5.671667628390075,
      "grad_norm": 1.0,
      "learning_rate": 0.00023138092963518664,
      "loss": 5.4005,
      "mean_token_accuracy": 0.22646197080612182,
      "num_tokens": 124551361.0,
      "step": 49145
    },
    {
      "entropy": 6.067145299911499,
      "epoch": 5.672244662435084,
      "grad_norm": 1.109375,
      "learning_rate": 0.00023134044885536847,
      "loss": 5.4715,
      "mean_token_accuracy": 0.21563300341367722,
      "num_tokens": 124564089.0,
      "step": 49150
    },
    {
      "entropy": 5.957214832305908,
      "epoch": 5.672821696480092,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002312999695440175,
      "loss": 5.3184,
      "mean_token_accuracy": 0.22793530374765397,
      "num_tokens": 124576493.0,
      "step": 49155
    },
    {
      "entropy": 6.0703206062316895,
      "epoch": 5.673398730525101,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002312594917024955,
      "loss": 5.5082,
      "mean_token_accuracy": 0.2087271273136139,
      "num_tokens": 124589614.0,
      "step": 49160
    },
    {
      "entropy": 6.090281867980957,
      "epoch": 5.67397576457011,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002312190153321636,
      "loss": 5.4038,
      "mean_token_accuracy": 0.22756052613258362,
      "num_tokens": 124602829.0,
      "step": 49165
    },
    {
      "entropy": 5.981957340240479,
      "epoch": 5.674552798615118,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023117854043438346,
      "loss": 5.3202,
      "mean_token_accuracy": 0.2388827383518219,
      "num_tokens": 124616542.0,
      "step": 49170
    },
    {
      "entropy": 5.908977317810058,
      "epoch": 5.675129832660127,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002311380670105163,
      "loss": 5.3274,
      "mean_token_accuracy": 0.23150137215852737,
      "num_tokens": 124628818.0,
      "step": 49175
    },
    {
      "entropy": 5.969433307647705,
      "epoch": 5.675706866705136,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023109759506192363,
      "loss": 5.3476,
      "mean_token_accuracy": 0.22930223494768143,
      "num_tokens": 124641063.0,
      "step": 49180
    },
    {
      "entropy": 5.971551990509033,
      "epoch": 5.676283900750144,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023105712458996636,
      "loss": 5.4117,
      "mean_token_accuracy": 0.22480900436639786,
      "num_tokens": 124655787.0,
      "step": 49185
    },
    {
      "entropy": 6.0313554286956785,
      "epoch": 5.676860934795153,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00023101665559600593,
      "loss": 5.46,
      "mean_token_accuracy": 0.22012256234884262,
      "num_tokens": 124669093.0,
      "step": 49190
    },
    {
      "entropy": 6.069334840774536,
      "epoch": 5.677437968840161,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002309761880814034,
      "loss": 5.4218,
      "mean_token_accuracy": 0.21948032230138778,
      "num_tokens": 124681413.0,
      "step": 49195
    },
    {
      "entropy": 5.993844938278198,
      "epoch": 5.678015002885171,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00023093572204752007,
      "loss": 5.3545,
      "mean_token_accuracy": 0.23845473378896714,
      "num_tokens": 124695225.0,
      "step": 49200
    },
    {
      "entropy": 6.01864013671875,
      "epoch": 5.678592036930179,
      "grad_norm": 1.125,
      "learning_rate": 0.0002308952574957167,
      "loss": 5.3794,
      "mean_token_accuracy": 0.2287394091486931,
      "num_tokens": 124707924.0,
      "step": 49205
    },
    {
      "entropy": 6.013266324996948,
      "epoch": 5.679169070975187,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00023085479442735448,
      "loss": 5.3522,
      "mean_token_accuracy": 0.22471704035997392,
      "num_tokens": 124721058.0,
      "step": 49210
    },
    {
      "entropy": 6.038667297363281,
      "epoch": 5.679746105020196,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023081433284379428,
      "loss": 5.4632,
      "mean_token_accuracy": 0.21689773648977279,
      "num_tokens": 124734153.0,
      "step": 49215
    },
    {
      "entropy": 5.980361270904541,
      "epoch": 5.680323139065205,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002307738727463971,
      "loss": 5.3139,
      "mean_token_accuracy": 0.22912372797727584,
      "num_tokens": 124747783.0,
      "step": 49220
    },
    {
      "entropy": 6.000739574432373,
      "epoch": 5.680900173110214,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023073341413652371,
      "loss": 5.4258,
      "mean_token_accuracy": 0.22797234058380128,
      "num_tokens": 124760789.0,
      "step": 49225
    },
    {
      "entropy": 6.070571660995483,
      "epoch": 5.681477207155222,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000230692957015535,
      "loss": 5.4604,
      "mean_token_accuracy": 0.21844085156917573,
      "num_tokens": 124773506.0,
      "step": 49230
    },
    {
      "entropy": 6.025353813171387,
      "epoch": 5.68205424120023,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023065250138479154,
      "loss": 5.4385,
      "mean_token_accuracy": 0.22111011445522308,
      "num_tokens": 124787662.0,
      "step": 49235
    },
    {
      "entropy": 6.065636348724365,
      "epoch": 5.68263127524524,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023061204724565432,
      "loss": 5.4192,
      "mean_token_accuracy": 0.2225457862019539,
      "num_tokens": 124800113.0,
      "step": 49240
    },
    {
      "entropy": 5.9465583801269535,
      "epoch": 5.683208309290248,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0002305715945994838,
      "loss": 5.3031,
      "mean_token_accuracy": 0.23550986796617507,
      "num_tokens": 124813677.0,
      "step": 49245
    },
    {
      "entropy": 5.942675304412842,
      "epoch": 5.6837853433352565,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00023053114344764059,
      "loss": 5.3325,
      "mean_token_accuracy": 0.2373376339673996,
      "num_tokens": 124826832.0,
      "step": 49250
    },
    {
      "entropy": 6.108705377578735,
      "epoch": 5.684362377380266,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023049069379148525,
      "loss": 5.5695,
      "mean_token_accuracy": 0.20624623894691468,
      "num_tokens": 124838983.0,
      "step": 49255
    },
    {
      "entropy": 6.065599489212036,
      "epoch": 5.684939411425274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023045024563237844,
      "loss": 5.4244,
      "mean_token_accuracy": 0.2298648998141289,
      "num_tokens": 124852090.0,
      "step": 49260
    },
    {
      "entropy": 6.08922176361084,
      "epoch": 5.685516445470283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00023040979897168034,
      "loss": 5.5427,
      "mean_token_accuracy": 0.21237729340791703,
      "num_tokens": 124866073.0,
      "step": 49265
    },
    {
      "entropy": 5.925222778320313,
      "epoch": 5.686093479515291,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023036935381075163,
      "loss": 5.2939,
      "mean_token_accuracy": 0.23084749281406403,
      "num_tokens": 124879814.0,
      "step": 49270
    },
    {
      "entropy": 6.017553901672363,
      "epoch": 5.6866705135603,
      "grad_norm": 1.046875,
      "learning_rate": 0.00023032891015095243,
      "loss": 5.432,
      "mean_token_accuracy": 0.2248712420463562,
      "num_tokens": 124892915.0,
      "step": 49275
    },
    {
      "entropy": 6.0236412525177006,
      "epoch": 5.687247547605309,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00023028846799364323,
      "loss": 5.4137,
      "mean_token_accuracy": 0.22685618847608566,
      "num_tokens": 124905658.0,
      "step": 49280
    },
    {
      "entropy": 6.039088869094849,
      "epoch": 5.687824581650317,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002302480273401842,
      "loss": 5.4605,
      "mean_token_accuracy": 0.21875010430812836,
      "num_tokens": 124918505.0,
      "step": 49285
    },
    {
      "entropy": 6.031147241592407,
      "epoch": 5.688401615695326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002302075881919355,
      "loss": 5.4336,
      "mean_token_accuracy": 0.21850383579730986,
      "num_tokens": 124931645.0,
      "step": 49290
    },
    {
      "entropy": 6.064820384979248,
      "epoch": 5.688978649740335,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00023016715055025727,
      "loss": 5.4782,
      "mean_token_accuracy": 0.21740976274013518,
      "num_tokens": 124944184.0,
      "step": 49295
    },
    {
      "entropy": 6.003515005111694,
      "epoch": 5.689555683785343,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002301267144165098,
      "loss": 5.3466,
      "mean_token_accuracy": 0.23011250346899031,
      "num_tokens": 124956918.0,
      "step": 49300
    },
    {
      "entropy": 6.043768072128296,
      "epoch": 5.690132717830352,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00023008627979205287,
      "loss": 5.4443,
      "mean_token_accuracy": 0.21665253937244416,
      "num_tokens": 124970080.0,
      "step": 49305
    },
    {
      "entropy": 5.978387022018433,
      "epoch": 5.69070975187536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00023004584667824668,
      "loss": 5.3224,
      "mean_token_accuracy": 0.2366739481687546,
      "num_tokens": 124982416.0,
      "step": 49310
    },
    {
      "entropy": 6.0595170021057125,
      "epoch": 5.6912867859203695,
      "grad_norm": 1.078125,
      "learning_rate": 0.00023000541507645105,
      "loss": 5.4839,
      "mean_token_accuracy": 0.22066294997930527,
      "num_tokens": 124996205.0,
      "step": 49315
    },
    {
      "entropy": 6.082577753067016,
      "epoch": 5.691863819965378,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022996498498802605,
      "loss": 5.5136,
      "mean_token_accuracy": 0.2163488268852234,
      "num_tokens": 125010169.0,
      "step": 49320
    },
    {
      "entropy": 6.094366979598999,
      "epoch": 5.692440854010386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002299245564143313,
      "loss": 5.4489,
      "mean_token_accuracy": 0.22005174160003663,
      "num_tokens": 125022213.0,
      "step": 49325
    },
    {
      "entropy": 6.010295629501343,
      "epoch": 5.693017888055396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022988412935672677,
      "loss": 5.4032,
      "mean_token_accuracy": 0.22347027957439422,
      "num_tokens": 125034355.0,
      "step": 49330
    },
    {
      "entropy": 6.028240203857422,
      "epoch": 5.693594922100404,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000229843703816572,
      "loss": 5.3611,
      "mean_token_accuracy": 0.2266454353928566,
      "num_tokens": 125046525.0,
      "step": 49335
    },
    {
      "entropy": 6.063290119171143,
      "epoch": 5.6941719561454125,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022980327979522702,
      "loss": 5.4696,
      "mean_token_accuracy": 0.224091237783432,
      "num_tokens": 125059164.0,
      "step": 49340
    },
    {
      "entropy": 5.909367990493775,
      "epoch": 5.694748990190421,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002297628572940511,
      "loss": 5.2753,
      "mean_token_accuracy": 0.2388189196586609,
      "num_tokens": 125072068.0,
      "step": 49345
    },
    {
      "entropy": 6.10900764465332,
      "epoch": 5.69532602423543,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002297224363144041,
      "loss": 5.5336,
      "mean_token_accuracy": 0.21222119331359862,
      "num_tokens": 125085192.0,
      "step": 49350
    },
    {
      "entropy": 6.051606845855713,
      "epoch": 5.695903058280439,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022968201685764545,
      "loss": 5.4159,
      "mean_token_accuracy": 0.22583380937576295,
      "num_tokens": 125097883.0,
      "step": 49355
    },
    {
      "entropy": 6.09923906326294,
      "epoch": 5.696480092325447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022964159892513458,
      "loss": 5.4685,
      "mean_token_accuracy": 0.21610894054174423,
      "num_tokens": 125110532.0,
      "step": 49360
    },
    {
      "entropy": 6.132844638824463,
      "epoch": 5.6970571263704555,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002296011825182312,
      "loss": 5.5194,
      "mean_token_accuracy": 0.21789731830358505,
      "num_tokens": 125124315.0,
      "step": 49365
    },
    {
      "entropy": 6.038289594650268,
      "epoch": 5.697634160415465,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002295607676382943,
      "loss": 5.3835,
      "mean_token_accuracy": 0.22340624630451203,
      "num_tokens": 125137353.0,
      "step": 49370
    },
    {
      "entropy": 6.072018480300903,
      "epoch": 5.698211194460473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022952035428668355,
      "loss": 5.5248,
      "mean_token_accuracy": 0.21980332285165788,
      "num_tokens": 125149501.0,
      "step": 49375
    },
    {
      "entropy": 5.977984428405762,
      "epoch": 5.698788228505482,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002294799424647579,
      "loss": 5.3075,
      "mean_token_accuracy": 0.234373015165329,
      "num_tokens": 125162007.0,
      "step": 49380
    },
    {
      "entropy": 6.0247578620910645,
      "epoch": 5.69936526255049,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022943953217387697,
      "loss": 5.3727,
      "mean_token_accuracy": 0.22241763770580292,
      "num_tokens": 125173834.0,
      "step": 49385
    },
    {
      "entropy": 6.060000896453857,
      "epoch": 5.699942296595499,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022939912341539965,
      "loss": 5.4921,
      "mean_token_accuracy": 0.21621426194906235,
      "num_tokens": 125186241.0,
      "step": 49390
    },
    {
      "entropy": 6.044709873199463,
      "epoch": 5.700519330640508,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022935871619068522,
      "loss": 5.3999,
      "mean_token_accuracy": 0.21451074331998826,
      "num_tokens": 125198440.0,
      "step": 49395
    },
    {
      "entropy": 5.998664855957031,
      "epoch": 5.701096364685516,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022931831050109264,
      "loss": 5.3459,
      "mean_token_accuracy": 0.22312902808189392,
      "num_tokens": 125211610.0,
      "step": 49400
    },
    {
      "entropy": 6.083296871185302,
      "epoch": 5.7016733987305255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022927790634798102,
      "loss": 5.4729,
      "mean_token_accuracy": 0.21423638015985488,
      "num_tokens": 125224928.0,
      "step": 49405
    },
    {
      "entropy": 6.023882055282593,
      "epoch": 5.702250432775534,
      "grad_norm": 1.125,
      "learning_rate": 0.00022923750373270925,
      "loss": 5.4793,
      "mean_token_accuracy": 0.2190165251493454,
      "num_tokens": 125236611.0,
      "step": 49410
    },
    {
      "entropy": 5.962625932693482,
      "epoch": 5.702827466820542,
      "grad_norm": 1.0,
      "learning_rate": 0.00022919710265663647,
      "loss": 5.2999,
      "mean_token_accuracy": 0.23331550359725953,
      "num_tokens": 125248934.0,
      "step": 49415
    },
    {
      "entropy": 6.02288122177124,
      "epoch": 5.703404500865551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002291567031211212,
      "loss": 5.4169,
      "mean_token_accuracy": 0.2240486890077591,
      "num_tokens": 125262339.0,
      "step": 49420
    },
    {
      "entropy": 6.056442499160767,
      "epoch": 5.70398153491056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022911630512752258,
      "loss": 5.4355,
      "mean_token_accuracy": 0.22386209964752196,
      "num_tokens": 125275286.0,
      "step": 49425
    },
    {
      "entropy": 6.017352962493897,
      "epoch": 5.7045585689555685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022907590867719913,
      "loss": 5.4129,
      "mean_token_accuracy": 0.21933057755231858,
      "num_tokens": 125287403.0,
      "step": 49430
    },
    {
      "entropy": 6.0339724063873295,
      "epoch": 5.705135603000577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022903551377150984,
      "loss": 5.4168,
      "mean_token_accuracy": 0.22614537179470062,
      "num_tokens": 125299540.0,
      "step": 49435
    },
    {
      "entropy": 6.020222997665405,
      "epoch": 5.705712637045586,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022899512041181307,
      "loss": 5.425,
      "mean_token_accuracy": 0.22286488711833954,
      "num_tokens": 125312440.0,
      "step": 49440
    },
    {
      "entropy": 6.058001184463501,
      "epoch": 5.706289671090595,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022895472859946764,
      "loss": 5.4876,
      "mean_token_accuracy": 0.22050503194332122,
      "num_tokens": 125324834.0,
      "step": 49445
    },
    {
      "entropy": 6.014037609100342,
      "epoch": 5.706866705135603,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022891433833583202,
      "loss": 5.3965,
      "mean_token_accuracy": 0.22351069003343582,
      "num_tokens": 125336807.0,
      "step": 49450
    },
    {
      "entropy": 6.027551651000977,
      "epoch": 5.707443739180611,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022887394962226476,
      "loss": 5.412,
      "mean_token_accuracy": 0.22454663664102553,
      "num_tokens": 125350067.0,
      "step": 49455
    },
    {
      "entropy": 6.030701351165772,
      "epoch": 5.70802077322562,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022883356246012428,
      "loss": 5.4463,
      "mean_token_accuracy": 0.21859916001558305,
      "num_tokens": 125363726.0,
      "step": 49460
    },
    {
      "entropy": 6.082229852676392,
      "epoch": 5.708597807270629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000228793176850769,
      "loss": 5.4687,
      "mean_token_accuracy": 0.21768958419561385,
      "num_tokens": 125376470.0,
      "step": 49465
    },
    {
      "entropy": 6.010114574432373,
      "epoch": 5.709174841315638,
      "grad_norm": 1.0,
      "learning_rate": 0.0002287527927955572,
      "loss": 5.3223,
      "mean_token_accuracy": 0.2245061680674553,
      "num_tokens": 125389294.0,
      "step": 49470
    },
    {
      "entropy": 6.087443542480469,
      "epoch": 5.709751875360646,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022871241029584732,
      "loss": 5.4648,
      "mean_token_accuracy": 0.22029041200876237,
      "num_tokens": 125402476.0,
      "step": 49475
    },
    {
      "entropy": 6.020562028884887,
      "epoch": 5.710328909405655,
      "grad_norm": 1.0,
      "learning_rate": 0.0002286720293529975,
      "loss": 5.4595,
      "mean_token_accuracy": 0.21875909417867662,
      "num_tokens": 125415636.0,
      "step": 49480
    },
    {
      "entropy": 5.9906861782073975,
      "epoch": 5.710905943450664,
      "grad_norm": 1.046875,
      "learning_rate": 0.000228631649968366,
      "loss": 5.3044,
      "mean_token_accuracy": 0.23321988433599472,
      "num_tokens": 125427294.0,
      "step": 49485
    },
    {
      "entropy": 6.011280632019043,
      "epoch": 5.711482977495672,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022859127214331084,
      "loss": 5.4006,
      "mean_token_accuracy": 0.22245244830846786,
      "num_tokens": 125439528.0,
      "step": 49490
    },
    {
      "entropy": 6.032044792175293,
      "epoch": 5.7120600115406805,
      "grad_norm": 1.125,
      "learning_rate": 0.0002285508958791903,
      "loss": 5.4398,
      "mean_token_accuracy": 0.22025610208511354,
      "num_tokens": 125452951.0,
      "step": 49495
    },
    {
      "entropy": 6.099039077758789,
      "epoch": 5.71263704558569,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022851052117736222,
      "loss": 5.5745,
      "mean_token_accuracy": 0.20847016721963882,
      "num_tokens": 125466179.0,
      "step": 49500
    },
    {
      "entropy": 5.992461729049682,
      "epoch": 5.713214079630698,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022847014803918466,
      "loss": 5.3834,
      "mean_token_accuracy": 0.2252698212862015,
      "num_tokens": 125478417.0,
      "step": 49505
    },
    {
      "entropy": 5.958797454833984,
      "epoch": 5.713791113675707,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022842977646601553,
      "loss": 5.4441,
      "mean_token_accuracy": 0.2254386067390442,
      "num_tokens": 125491506.0,
      "step": 49510
    },
    {
      "entropy": 5.992489910125732,
      "epoch": 5.714368147720716,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022838940645921284,
      "loss": 5.3779,
      "mean_token_accuracy": 0.2238729029893875,
      "num_tokens": 125505619.0,
      "step": 49515
    },
    {
      "entropy": 6.117829656600952,
      "epoch": 5.714945181765724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022834903802013425,
      "loss": 5.4618,
      "mean_token_accuracy": 0.21533117145299913,
      "num_tokens": 125519667.0,
      "step": 49520
    },
    {
      "entropy": 6.0782386302948,
      "epoch": 5.715522215810733,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022830867115013765,
      "loss": 5.4128,
      "mean_token_accuracy": 0.22341126948595047,
      "num_tokens": 125532068.0,
      "step": 49525
    },
    {
      "entropy": 5.97263731956482,
      "epoch": 5.716099249855741,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00022826830585058062,
      "loss": 5.3697,
      "mean_token_accuracy": 0.22721271216869354,
      "num_tokens": 125545205.0,
      "step": 49530
    },
    {
      "entropy": 6.001164484024048,
      "epoch": 5.71667628390075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000228227942122821,
      "loss": 5.4092,
      "mean_token_accuracy": 0.22437716722488404,
      "num_tokens": 125557215.0,
      "step": 49535
    },
    {
      "entropy": 6.049753475189209,
      "epoch": 5.717253317945759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002281875799682162,
      "loss": 5.4171,
      "mean_token_accuracy": 0.21761631071567536,
      "num_tokens": 125568496.0,
      "step": 49540
    },
    {
      "entropy": 6.045251655578613,
      "epoch": 5.717830351990767,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022814721938812398,
      "loss": 5.4505,
      "mean_token_accuracy": 0.221990704536438,
      "num_tokens": 125581426.0,
      "step": 49545
    },
    {
      "entropy": 5.998879384994507,
      "epoch": 5.718407386035776,
      "grad_norm": 0.890625,
      "learning_rate": 0.0002281068603839017,
      "loss": 5.3727,
      "mean_token_accuracy": 0.23033248484134675,
      "num_tokens": 125595220.0,
      "step": 49550
    },
    {
      "entropy": 6.022923851013184,
      "epoch": 5.718984420080785,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022806650295690702,
      "loss": 5.3735,
      "mean_token_accuracy": 0.22625367492437362,
      "num_tokens": 125607310.0,
      "step": 49555
    },
    {
      "entropy": 6.019437456130982,
      "epoch": 5.7195614541257935,
      "grad_norm": 1.03125,
      "learning_rate": 0.000228026147108497,
      "loss": 5.3951,
      "mean_token_accuracy": 0.22167592942714692,
      "num_tokens": 125619789.0,
      "step": 49560
    },
    {
      "entropy": 6.043296384811401,
      "epoch": 5.720138488170802,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022798579284002935,
      "loss": 5.5065,
      "mean_token_accuracy": 0.21331021040678025,
      "num_tokens": 125632371.0,
      "step": 49565
    },
    {
      "entropy": 6.021444034576416,
      "epoch": 5.72071552221581,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022794544015286106,
      "loss": 5.3945,
      "mean_token_accuracy": 0.2248529702425003,
      "num_tokens": 125644886.0,
      "step": 49570
    },
    {
      "entropy": 6.043733167648315,
      "epoch": 5.72129255626082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022790508904834967,
      "loss": 5.3897,
      "mean_token_accuracy": 0.22720012813806534,
      "num_tokens": 125656604.0,
      "step": 49575
    },
    {
      "entropy": 6.08525505065918,
      "epoch": 5.721869590305828,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022786473952785214,
      "loss": 5.5247,
      "mean_token_accuracy": 0.21613952964544297,
      "num_tokens": 125668702.0,
      "step": 49580
    },
    {
      "entropy": 6.0983521938323975,
      "epoch": 5.7224466243508365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022782439159272567,
      "loss": 5.4464,
      "mean_token_accuracy": 0.21784875988960267,
      "num_tokens": 125679605.0,
      "step": 49585
    },
    {
      "entropy": 6.077865362167358,
      "epoch": 5.723023658395846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022778404524432733,
      "loss": 5.481,
      "mean_token_accuracy": 0.21043082028627397,
      "num_tokens": 125690820.0,
      "step": 49590
    },
    {
      "entropy": 6.015727472305298,
      "epoch": 5.723600692440854,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022774370048401428,
      "loss": 5.4087,
      "mean_token_accuracy": 0.21982302516698837,
      "num_tokens": 125704560.0,
      "step": 49595
    },
    {
      "entropy": 5.949781084060669,
      "epoch": 5.724177726485863,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002277033573131433,
      "loss": 5.3339,
      "mean_token_accuracy": 0.2331131175160408,
      "num_tokens": 125717281.0,
      "step": 49600
    },
    {
      "entropy": 6.006426763534546,
      "epoch": 5.724754760530871,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022766301573307147,
      "loss": 5.4263,
      "mean_token_accuracy": 0.22441961616277695,
      "num_tokens": 125730281.0,
      "step": 49605
    },
    {
      "entropy": 5.976651763916015,
      "epoch": 5.7253317945758795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002276226757451555,
      "loss": 5.3856,
      "mean_token_accuracy": 0.2272988587617874,
      "num_tokens": 125742157.0,
      "step": 49610
    },
    {
      "entropy": 6.079996061325073,
      "epoch": 5.725908828620889,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002275823373507524,
      "loss": 5.4628,
      "mean_token_accuracy": 0.21890117824077607,
      "num_tokens": 125754860.0,
      "step": 49615
    },
    {
      "entropy": 5.970008659362793,
      "epoch": 5.726485862665897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022754200055121878,
      "loss": 5.2697,
      "mean_token_accuracy": 0.23564553409814834,
      "num_tokens": 125768993.0,
      "step": 49620
    },
    {
      "entropy": 6.055458688735962,
      "epoch": 5.727062896710906,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022750166534791139,
      "loss": 5.487,
      "mean_token_accuracy": 0.21542387455701828,
      "num_tokens": 125780775.0,
      "step": 49625
    },
    {
      "entropy": 6.019528818130493,
      "epoch": 5.727639930755915,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022746133174218687,
      "loss": 5.3689,
      "mean_token_accuracy": 0.2287971258163452,
      "num_tokens": 125793269.0,
      "step": 49630
    },
    {
      "entropy": 6.0886589050292965,
      "epoch": 5.728216964800923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022742099973540193,
      "loss": 5.466,
      "mean_token_accuracy": 0.2193503424525261,
      "num_tokens": 125806386.0,
      "step": 49635
    },
    {
      "entropy": 6.104493427276611,
      "epoch": 5.728793998845932,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00022738066932891294,
      "loss": 5.5403,
      "mean_token_accuracy": 0.21566086858510972,
      "num_tokens": 125819164.0,
      "step": 49640
    },
    {
      "entropy": 5.929534196853638,
      "epoch": 5.72937103289094,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022734034052407655,
      "loss": 5.3415,
      "mean_token_accuracy": 0.22868810147047042,
      "num_tokens": 125831952.0,
      "step": 49645
    },
    {
      "entropy": 5.997420167922973,
      "epoch": 5.7299480669359495,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022730001332224903,
      "loss": 5.3486,
      "mean_token_accuracy": 0.22770389318466186,
      "num_tokens": 125844537.0,
      "step": 49650
    },
    {
      "entropy": 6.0090131759643555,
      "epoch": 5.730525100980958,
      "grad_norm": 1.0625,
      "learning_rate": 0.000227259687724787,
      "loss": 5.3545,
      "mean_token_accuracy": 0.22506375908851622,
      "num_tokens": 125855946.0,
      "step": 49655
    },
    {
      "entropy": 6.107216787338257,
      "epoch": 5.731102135025966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022721936373304658,
      "loss": 5.5474,
      "mean_token_accuracy": 0.21119464933872223,
      "num_tokens": 125868298.0,
      "step": 49660
    },
    {
      "entropy": 6.053257608413697,
      "epoch": 5.731679169070976,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00022717904134838412,
      "loss": 5.5353,
      "mean_token_accuracy": 0.21726712435483933,
      "num_tokens": 125879984.0,
      "step": 49665
    },
    {
      "entropy": 6.09370903968811,
      "epoch": 5.732256203115984,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022713872057215584,
      "loss": 5.5058,
      "mean_token_accuracy": 0.21583452820777893,
      "num_tokens": 125892668.0,
      "step": 49670
    },
    {
      "entropy": 6.109275436401367,
      "epoch": 5.7328332371609925,
      "grad_norm": 1.125,
      "learning_rate": 0.00022709840140571803,
      "loss": 5.4524,
      "mean_token_accuracy": 0.2226664289832115,
      "num_tokens": 125904768.0,
      "step": 49675
    },
    {
      "entropy": 6.014448022842407,
      "epoch": 5.733410271206001,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022705808385042658,
      "loss": 5.4281,
      "mean_token_accuracy": 0.22407831400632858,
      "num_tokens": 125917439.0,
      "step": 49680
    },
    {
      "entropy": 6.064894390106201,
      "epoch": 5.73398730525101,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022701776790763772,
      "loss": 5.4871,
      "mean_token_accuracy": 0.2175563395023346,
      "num_tokens": 125930159.0,
      "step": 49685
    },
    {
      "entropy": 6.058375358581543,
      "epoch": 5.734564339296019,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022697745357870737,
      "loss": 5.4282,
      "mean_token_accuracy": 0.21238384246826172,
      "num_tokens": 125941886.0,
      "step": 49690
    },
    {
      "entropy": 6.04431676864624,
      "epoch": 5.735141373341027,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022693714086499157,
      "loss": 5.461,
      "mean_token_accuracy": 0.22099485397338867,
      "num_tokens": 125954526.0,
      "step": 49695
    },
    {
      "entropy": 6.054618215560913,
      "epoch": 5.7357184073860354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002268968297678461,
      "loss": 5.4897,
      "mean_token_accuracy": 0.2189372271299362,
      "num_tokens": 125966539.0,
      "step": 49700
    },
    {
      "entropy": 6.0252748966217045,
      "epoch": 5.736295441431045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002268565202886269,
      "loss": 5.3762,
      "mean_token_accuracy": 0.22499978840351104,
      "num_tokens": 125978712.0,
      "step": 49705
    },
    {
      "entropy": 6.028599500656128,
      "epoch": 5.736872475476053,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022681621242868967,
      "loss": 5.4316,
      "mean_token_accuracy": 0.22541026324033736,
      "num_tokens": 125990942.0,
      "step": 49710
    },
    {
      "entropy": 5.966779327392578,
      "epoch": 5.737449509521062,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022677590618939031,
      "loss": 5.3434,
      "mean_token_accuracy": 0.2317936524748802,
      "num_tokens": 126003956.0,
      "step": 49715
    },
    {
      "entropy": 6.00062141418457,
      "epoch": 5.73802654356607,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022673560157208435,
      "loss": 5.4257,
      "mean_token_accuracy": 0.22327411770820618,
      "num_tokens": 126016524.0,
      "step": 49720
    },
    {
      "entropy": 6.032732963562012,
      "epoch": 5.738603577611079,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022669529857812743,
      "loss": 5.4939,
      "mean_token_accuracy": 0.21566978991031646,
      "num_tokens": 126028432.0,
      "step": 49725
    },
    {
      "entropy": 6.074692821502685,
      "epoch": 5.739180611656088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022665499720887506,
      "loss": 5.4381,
      "mean_token_accuracy": 0.22237212657928468,
      "num_tokens": 126039476.0,
      "step": 49730
    },
    {
      "entropy": 6.011688375473023,
      "epoch": 5.739757645701096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022661469746568304,
      "loss": 5.3558,
      "mean_token_accuracy": 0.2296280339360237,
      "num_tokens": 126052599.0,
      "step": 49735
    },
    {
      "entropy": 5.901757621765137,
      "epoch": 5.7403346797461055,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022657439934990647,
      "loss": 5.3188,
      "mean_token_accuracy": 0.23965569734573364,
      "num_tokens": 126066176.0,
      "step": 49740
    },
    {
      "entropy": 5.996234798431397,
      "epoch": 5.740911713791114,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022653410286290105,
      "loss": 5.3528,
      "mean_token_accuracy": 0.22932596653699874,
      "num_tokens": 126077773.0,
      "step": 49745
    },
    {
      "entropy": 6.077146911621094,
      "epoch": 5.741488747836122,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022649380800602193,
      "loss": 5.4536,
      "mean_token_accuracy": 0.22362992763519288,
      "num_tokens": 126090167.0,
      "step": 49750
    },
    {
      "entropy": 6.041822719573974,
      "epoch": 5.742065781881131,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022645351478062437,
      "loss": 5.5182,
      "mean_token_accuracy": 0.21482108980417253,
      "num_tokens": 126102075.0,
      "step": 49755
    },
    {
      "entropy": 6.044607210159302,
      "epoch": 5.74264281592614,
      "grad_norm": 1.0,
      "learning_rate": 0.00022641322318806397,
      "loss": 5.4713,
      "mean_token_accuracy": 0.22000051736831666,
      "num_tokens": 126115269.0,
      "step": 49760
    },
    {
      "entropy": 6.038962650299072,
      "epoch": 5.743219849971148,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002263729332296955,
      "loss": 5.4869,
      "mean_token_accuracy": 0.21887048780918122,
      "num_tokens": 126128115.0,
      "step": 49765
    },
    {
      "entropy": 6.084073638916015,
      "epoch": 5.743796884016157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002263326449068744,
      "loss": 5.4642,
      "mean_token_accuracy": 0.22342609018087387,
      "num_tokens": 126139677.0,
      "step": 49770
    },
    {
      "entropy": 6.040108728408813,
      "epoch": 5.744373918061165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002262923582209555,
      "loss": 5.3806,
      "mean_token_accuracy": 0.22168772369623185,
      "num_tokens": 126153393.0,
      "step": 49775
    },
    {
      "entropy": 6.081512212753296,
      "epoch": 5.744950952106175,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000226252073173294,
      "loss": 5.5135,
      "mean_token_accuracy": 0.21374876350164412,
      "num_tokens": 126165172.0,
      "step": 49780
    },
    {
      "entropy": 5.962333106994629,
      "epoch": 5.745527986151183,
      "grad_norm": 1.25,
      "learning_rate": 0.0002262117897652447,
      "loss": 5.2425,
      "mean_token_accuracy": 0.24064727872610092,
      "num_tokens": 126177929.0,
      "step": 49785
    },
    {
      "entropy": 5.981118488311767,
      "epoch": 5.746105020196191,
      "grad_norm": 1.0,
      "learning_rate": 0.00022617150799816277,
      "loss": 5.3662,
      "mean_token_accuracy": 0.22630050778388977,
      "num_tokens": 126192013.0,
      "step": 49790
    },
    {
      "entropy": 6.025222826004028,
      "epoch": 5.7466820542412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00022613122787340279,
      "loss": 5.4736,
      "mean_token_accuracy": 0.2192372128367424,
      "num_tokens": 126205142.0,
      "step": 49795
    },
    {
      "entropy": 6.100870132446289,
      "epoch": 5.747259088286209,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022609094939231978,
      "loss": 5.4296,
      "mean_token_accuracy": 0.22028184235095977,
      "num_tokens": 126217939.0,
      "step": 49800
    },
    {
      "entropy": 5.908008670806884,
      "epoch": 5.7478361223312175,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022605067255626832,
      "loss": 5.3053,
      "mean_token_accuracy": 0.23460683375597,
      "num_tokens": 126230073.0,
      "step": 49805
    },
    {
      "entropy": 5.975973892211914,
      "epoch": 5.748413156376226,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00022601039736660323,
      "loss": 5.4325,
      "mean_token_accuracy": 0.2241601303219795,
      "num_tokens": 126243141.0,
      "step": 49810
    },
    {
      "entropy": 5.9624570369720455,
      "epoch": 5.748990190421235,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022597012382467903,
      "loss": 5.3772,
      "mean_token_accuracy": 0.22081854939460754,
      "num_tokens": 126256389.0,
      "step": 49815
    },
    {
      "entropy": 6.038847589492798,
      "epoch": 5.749567224466244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002259298519318504,
      "loss": 5.3369,
      "mean_token_accuracy": 0.2268427237868309,
      "num_tokens": 126268255.0,
      "step": 49820
    },
    {
      "entropy": 6.035766315460205,
      "epoch": 5.750144258511252,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00022588958168947182,
      "loss": 5.4331,
      "mean_token_accuracy": 0.22694395184516908,
      "num_tokens": 126281250.0,
      "step": 49825
    },
    {
      "entropy": 6.049199008941651,
      "epoch": 5.7507212925562605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022584931309889783,
      "loss": 5.4564,
      "mean_token_accuracy": 0.2150783956050873,
      "num_tokens": 126292572.0,
      "step": 49830
    },
    {
      "entropy": 5.958545875549317,
      "epoch": 5.75129832660127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022580904616148267,
      "loss": 5.3985,
      "mean_token_accuracy": 0.22599672079086303,
      "num_tokens": 126305222.0,
      "step": 49835
    },
    {
      "entropy": 5.969755983352661,
      "epoch": 5.751875360646278,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022576878087858082,
      "loss": 5.3804,
      "mean_token_accuracy": 0.2245454952120781,
      "num_tokens": 126320193.0,
      "step": 49840
    },
    {
      "entropy": 6.08737621307373,
      "epoch": 5.752452394691287,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022572851725154657,
      "loss": 5.4531,
      "mean_token_accuracy": 0.22295423448085785,
      "num_tokens": 126332865.0,
      "step": 49845
    },
    {
      "entropy": 6.078256750106812,
      "epoch": 5.753029428736295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00022568825528173427,
      "loss": 5.4528,
      "mean_token_accuracy": 0.2178967148065567,
      "num_tokens": 126345830.0,
      "step": 49850
    },
    {
      "entropy": 5.994029235839844,
      "epoch": 5.753606462781304,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022564799497049794,
      "loss": 5.299,
      "mean_token_accuracy": 0.23850910663604735,
      "num_tokens": 126357014.0,
      "step": 49855
    },
    {
      "entropy": 6.059012794494629,
      "epoch": 5.754183496826313,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002256077363191918,
      "loss": 5.4323,
      "mean_token_accuracy": 0.22131062299013138,
      "num_tokens": 126368326.0,
      "step": 49860
    },
    {
      "entropy": 6.034923839569092,
      "epoch": 5.754760530871321,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002255674793291698,
      "loss": 5.4652,
      "mean_token_accuracy": 0.219676411151886,
      "num_tokens": 126381647.0,
      "step": 49865
    },
    {
      "entropy": 6.06753945350647,
      "epoch": 5.75533756491633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022552722400178627,
      "loss": 5.4841,
      "mean_token_accuracy": 0.21351805776357652,
      "num_tokens": 126393772.0,
      "step": 49870
    },
    {
      "entropy": 6.103093814849854,
      "epoch": 5.755914598961339,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022548697033839487,
      "loss": 5.5211,
      "mean_token_accuracy": 0.21728209406137466,
      "num_tokens": 126405824.0,
      "step": 49875
    },
    {
      "entropy": 6.115517234802246,
      "epoch": 5.756491633006347,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00022544671834034968,
      "loss": 5.5161,
      "mean_token_accuracy": 0.22121085673570634,
      "num_tokens": 126417663.0,
      "step": 49880
    },
    {
      "entropy": 5.970009708404541,
      "epoch": 5.757068667051356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002254064680090044,
      "loss": 5.3553,
      "mean_token_accuracy": 0.2329029530286789,
      "num_tokens": 126429709.0,
      "step": 49885
    },
    {
      "entropy": 5.913326263427734,
      "epoch": 5.757645701096365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002253662193457131,
      "loss": 5.2207,
      "mean_token_accuracy": 0.23543574362993241,
      "num_tokens": 126441859.0,
      "step": 49890
    },
    {
      "entropy": 5.947325372695923,
      "epoch": 5.7582227351413735,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022532597235182922,
      "loss": 5.3907,
      "mean_token_accuracy": 0.2207629755139351,
      "num_tokens": 126455330.0,
      "step": 49895
    },
    {
      "entropy": 6.047615957260132,
      "epoch": 5.758799769186382,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022528572702870664,
      "loss": 5.3868,
      "mean_token_accuracy": 0.21666864156723023,
      "num_tokens": 126467028.0,
      "step": 49900
    },
    {
      "entropy": 6.074980783462524,
      "epoch": 5.75937680323139,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002252454833776989,
      "loss": 5.4508,
      "mean_token_accuracy": 0.21478413939476013,
      "num_tokens": 126478638.0,
      "step": 49905
    },
    {
      "entropy": 5.99262580871582,
      "epoch": 5.7599538372764,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022520524140015968,
      "loss": 5.402,
      "mean_token_accuracy": 0.22312563061714172,
      "num_tokens": 126490639.0,
      "step": 49910
    },
    {
      "entropy": 5.994853687286377,
      "epoch": 5.760530871321408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022516500109744236,
      "loss": 5.3326,
      "mean_token_accuracy": 0.22468214482069016,
      "num_tokens": 126502968.0,
      "step": 49915
    },
    {
      "entropy": 5.999149560928345,
      "epoch": 5.7611079053664165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002251247624709005,
      "loss": 5.4003,
      "mean_token_accuracy": 0.22592335790395737,
      "num_tokens": 126514931.0,
      "step": 49920
    },
    {
      "entropy": 6.031009101867676,
      "epoch": 5.761684939411425,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00022508452552188752,
      "loss": 5.3771,
      "mean_token_accuracy": 0.22444205582141877,
      "num_tokens": 126528082.0,
      "step": 49925
    },
    {
      "entropy": 6.082283210754395,
      "epoch": 5.762261973456434,
      "grad_norm": 1.0,
      "learning_rate": 0.00022504429025175672,
      "loss": 5.4668,
      "mean_token_accuracy": 0.22562110275030137,
      "num_tokens": 126540194.0,
      "step": 49930
    },
    {
      "entropy": 6.033959293365479,
      "epoch": 5.762839007501443,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022500405666186135,
      "loss": 5.4779,
      "mean_token_accuracy": 0.21441821604967118,
      "num_tokens": 126554694.0,
      "step": 49935
    },
    {
      "entropy": 5.9843250751495365,
      "epoch": 5.763416041546451,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022496382475355477,
      "loss": 5.3555,
      "mean_token_accuracy": 0.22191950678825378,
      "num_tokens": 126567993.0,
      "step": 49940
    },
    {
      "entropy": 5.978922176361084,
      "epoch": 5.7639930755914595,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022492359452819,
      "loss": 5.3791,
      "mean_token_accuracy": 0.2332445651292801,
      "num_tokens": 126580020.0,
      "step": 49945
    },
    {
      "entropy": 5.922740793228149,
      "epoch": 5.764570109636469,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022488336598712032,
      "loss": 5.3329,
      "mean_token_accuracy": 0.23084624111652374,
      "num_tokens": 126593373.0,
      "step": 49950
    },
    {
      "entropy": 6.07316837310791,
      "epoch": 5.765147143681477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022484313913169874,
      "loss": 5.4426,
      "mean_token_accuracy": 0.22318137139081956,
      "num_tokens": 126606247.0,
      "step": 49955
    },
    {
      "entropy": 6.1736674308776855,
      "epoch": 5.765724177726486,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022480291396327818,
      "loss": 5.5727,
      "mean_token_accuracy": 0.21208734214305877,
      "num_tokens": 126618736.0,
      "step": 49960
    },
    {
      "entropy": 5.87519416809082,
      "epoch": 5.766301211771495,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00022476269048321175,
      "loss": 5.1992,
      "mean_token_accuracy": 0.2513745278120041,
      "num_tokens": 126632353.0,
      "step": 49965
    },
    {
      "entropy": 6.012812232971191,
      "epoch": 5.766878245816503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022472246869285228,
      "loss": 5.4111,
      "mean_token_accuracy": 0.22521844655275344,
      "num_tokens": 126644101.0,
      "step": 49970
    },
    {
      "entropy": 5.99951868057251,
      "epoch": 5.767455279861512,
      "grad_norm": 1.0,
      "learning_rate": 0.00022468224859355256,
      "loss": 5.4225,
      "mean_token_accuracy": 0.22927317023277283,
      "num_tokens": 126657234.0,
      "step": 49975
    },
    {
      "entropy": 5.994365692138672,
      "epoch": 5.76803231390652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022464203018666545,
      "loss": 5.3539,
      "mean_token_accuracy": 0.22960476130247115,
      "num_tokens": 126669557.0,
      "step": 49980
    },
    {
      "entropy": 5.967706108093262,
      "epoch": 5.7686093479515295,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022460181347354352,
      "loss": 5.2847,
      "mean_token_accuracy": 0.2337131306529045,
      "num_tokens": 126683203.0,
      "step": 49985
    },
    {
      "entropy": 6.055516147613526,
      "epoch": 5.769186381996538,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022456159845553976,
      "loss": 5.3713,
      "mean_token_accuracy": 0.22442142963409423,
      "num_tokens": 126696446.0,
      "step": 49990
    },
    {
      "entropy": 5.970490169525147,
      "epoch": 5.769763416041546,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022452138513400643,
      "loss": 5.3657,
      "mean_token_accuracy": 0.23229675143957138,
      "num_tokens": 126709302.0,
      "step": 49995
    },
    {
      "entropy": 6.00638427734375,
      "epoch": 5.770340450086556,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022448117351029633,
      "loss": 5.4234,
      "mean_token_accuracy": 0.21543505191802978,
      "num_tokens": 126721495.0,
      "step": 50000
    },
    {
      "entropy": 6.02474889755249,
      "epoch": 5.770917484131564,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022444096358576176,
      "loss": 5.3996,
      "mean_token_accuracy": 0.21535304486751555,
      "num_tokens": 126734102.0,
      "step": 50005
    },
    {
      "entropy": 6.110607290267945,
      "epoch": 5.771494518176572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002244007553617554,
      "loss": 5.4824,
      "mean_token_accuracy": 0.21578304916620256,
      "num_tokens": 126748363.0,
      "step": 50010
    },
    {
      "entropy": 6.066141748428345,
      "epoch": 5.772071552221581,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022436054883962937,
      "loss": 5.4571,
      "mean_token_accuracy": 0.21879150420427323,
      "num_tokens": 126760546.0,
      "step": 50015
    },
    {
      "entropy": 5.972143030166626,
      "epoch": 5.772648586266589,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002243203440207362,
      "loss": 5.3292,
      "mean_token_accuracy": 0.23004869669675826,
      "num_tokens": 126772062.0,
      "step": 50020
    },
    {
      "entropy": 6.010297584533691,
      "epoch": 5.773225620311599,
      "grad_norm": 1.0625,
      "learning_rate": 0.00022428014090642806,
      "loss": 5.4769,
      "mean_token_accuracy": 0.2178845226764679,
      "num_tokens": 126785147.0,
      "step": 50025
    },
    {
      "entropy": 6.048260927200317,
      "epoch": 5.773802654356607,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00022423993949805722,
      "loss": 5.4405,
      "mean_token_accuracy": 0.221363665163517,
      "num_tokens": 126797916.0,
      "step": 50030
    },
    {
      "entropy": 6.134730386734009,
      "epoch": 5.774379688401615,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022419973979697573,
      "loss": 5.5011,
      "mean_token_accuracy": 0.21450162976980208,
      "num_tokens": 126809340.0,
      "step": 50035
    },
    {
      "entropy": 6.079486560821533,
      "epoch": 5.774956722446625,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022415954180453584,
      "loss": 5.5198,
      "mean_token_accuracy": 0.21297764629125596,
      "num_tokens": 126821735.0,
      "step": 50040
    },
    {
      "entropy": 5.96282205581665,
      "epoch": 5.775533756491633,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002241193455220894,
      "loss": 5.3953,
      "mean_token_accuracy": 0.23137205392122268,
      "num_tokens": 126834703.0,
      "step": 50045
    },
    {
      "entropy": 6.089934825897217,
      "epoch": 5.776110790536642,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022407915095098868,
      "loss": 5.435,
      "mean_token_accuracy": 0.2171575039625168,
      "num_tokens": 126845984.0,
      "step": 50050
    },
    {
      "entropy": 6.0216522216796875,
      "epoch": 5.77668782458165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022403895809258522,
      "loss": 5.4092,
      "mean_token_accuracy": 0.22644775956869126,
      "num_tokens": 126860204.0,
      "step": 50055
    },
    {
      "entropy": 6.06054277420044,
      "epoch": 5.777264858626659,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022399876694823123,
      "loss": 5.4196,
      "mean_token_accuracy": 0.21965304017066956,
      "num_tokens": 126871829.0,
      "step": 50060
    },
    {
      "entropy": 5.975581645965576,
      "epoch": 5.777841892671668,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002239585775192783,
      "loss": 5.4145,
      "mean_token_accuracy": 0.22135410606861114,
      "num_tokens": 126885888.0,
      "step": 50065
    },
    {
      "entropy": 5.947051763534546,
      "epoch": 5.778418926716676,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00022391838980707836,
      "loss": 5.3647,
      "mean_token_accuracy": 0.2250747114419937,
      "num_tokens": 126898742.0,
      "step": 50070
    },
    {
      "entropy": 6.101533269882202,
      "epoch": 5.778995960761685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002238782038129829,
      "loss": 5.445,
      "mean_token_accuracy": 0.22581405192613602,
      "num_tokens": 126912988.0,
      "step": 50075
    },
    {
      "entropy": 6.018276643753052,
      "epoch": 5.779572994806694,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002238380195383437,
      "loss": 5.4134,
      "mean_token_accuracy": 0.22227396070957184,
      "num_tokens": 126925513.0,
      "step": 50080
    },
    {
      "entropy": 5.925036144256592,
      "epoch": 5.780150028851702,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022379783698451227,
      "loss": 5.3001,
      "mean_token_accuracy": 0.23089775145053865,
      "num_tokens": 126938355.0,
      "step": 50085
    },
    {
      "entropy": 6.027796936035156,
      "epoch": 5.780727062896711,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022375765615284028,
      "loss": 5.4483,
      "mean_token_accuracy": 0.22097047567367553,
      "num_tokens": 126951000.0,
      "step": 50090
    },
    {
      "entropy": 6.096664381027222,
      "epoch": 5.781304096941719,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00022371747704467892,
      "loss": 5.4939,
      "mean_token_accuracy": 0.2167383164167404,
      "num_tokens": 126962477.0,
      "step": 50095
    },
    {
      "entropy": 6.008701181411743,
      "epoch": 5.781881130986728,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002236772996613799,
      "loss": 5.3954,
      "mean_token_accuracy": 0.22348914444446563,
      "num_tokens": 126974890.0,
      "step": 50100
    },
    {
      "entropy": 5.889964485168457,
      "epoch": 5.782458165031737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002236371240042942,
      "loss": 5.278,
      "mean_token_accuracy": 0.24203188121318817,
      "num_tokens": 126988376.0,
      "step": 50105
    },
    {
      "entropy": 5.979335927963257,
      "epoch": 5.783035199076745,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002235969500747735,
      "loss": 5.3626,
      "mean_token_accuracy": 0.22512841671705247,
      "num_tokens": 127001061.0,
      "step": 50110
    },
    {
      "entropy": 6.014907026290894,
      "epoch": 5.7836122331217545,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022355677787416878,
      "loss": 5.3657,
      "mean_token_accuracy": 0.22747167199850082,
      "num_tokens": 127014368.0,
      "step": 50115
    },
    {
      "entropy": 6.017622566223144,
      "epoch": 5.784189267166763,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022351660740383135,
      "loss": 5.3681,
      "mean_token_accuracy": 0.23234473317861556,
      "num_tokens": 127027207.0,
      "step": 50120
    },
    {
      "entropy": 6.017213296890259,
      "epoch": 5.784766301211771,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022347643866511214,
      "loss": 5.5171,
      "mean_token_accuracy": 0.21721160113811494,
      "num_tokens": 127040276.0,
      "step": 50125
    },
    {
      "entropy": 5.990512418746948,
      "epoch": 5.78534333525678,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002234362716593624,
      "loss": 5.3775,
      "mean_token_accuracy": 0.2281092420220375,
      "num_tokens": 127053622.0,
      "step": 50130
    },
    {
      "entropy": 6.101986694335937,
      "epoch": 5.785920369301789,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000223396106387933,
      "loss": 5.4255,
      "mean_token_accuracy": 0.22069344520568848,
      "num_tokens": 127066009.0,
      "step": 50135
    },
    {
      "entropy": 6.102804660797119,
      "epoch": 5.7864974033467975,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022335594285217503,
      "loss": 5.4435,
      "mean_token_accuracy": 0.21678621023893357,
      "num_tokens": 127078455.0,
      "step": 50140
    },
    {
      "entropy": 6.011951303482055,
      "epoch": 5.787074437391806,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022331578105343918,
      "loss": 5.4157,
      "mean_token_accuracy": 0.2207127407193184,
      "num_tokens": 127092224.0,
      "step": 50145
    },
    {
      "entropy": 6.049638748168945,
      "epoch": 5.787651471436815,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022327562099307634,
      "loss": 5.4251,
      "mean_token_accuracy": 0.2205551117658615,
      "num_tokens": 127104704.0,
      "step": 50150
    },
    {
      "entropy": 6.065674781799316,
      "epoch": 5.788228505481824,
      "grad_norm": 0.921875,
      "learning_rate": 0.00022323546267243732,
      "loss": 5.4496,
      "mean_token_accuracy": 0.21896771639585494,
      "num_tokens": 127117485.0,
      "step": 50155
    },
    {
      "entropy": 6.061332035064697,
      "epoch": 5.788805539526832,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00022319530609287282,
      "loss": 5.4623,
      "mean_token_accuracy": 0.2144940212368965,
      "num_tokens": 127130076.0,
      "step": 50160
    },
    {
      "entropy": 6.018175363540649,
      "epoch": 5.7893825735718405,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022315515125573355,
      "loss": 5.3835,
      "mean_token_accuracy": 0.22282203584909438,
      "num_tokens": 127142077.0,
      "step": 50165
    },
    {
      "entropy": 6.039028263092041,
      "epoch": 5.789959607616849,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002231149981623699,
      "loss": 5.3353,
      "mean_token_accuracy": 0.2287507548928261,
      "num_tokens": 127153863.0,
      "step": 50170
    },
    {
      "entropy": 5.959327173233032,
      "epoch": 5.790536641661858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00022307484681413254,
      "loss": 5.3662,
      "mean_token_accuracy": 0.23061311691999437,
      "num_tokens": 127165802.0,
      "step": 50175
    },
    {
      "entropy": 5.8730137825012205,
      "epoch": 5.791113675706867,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002230346972123719,
      "loss": 5.2725,
      "mean_token_accuracy": 0.23711815029382705,
      "num_tokens": 127179346.0,
      "step": 50180
    },
    {
      "entropy": 6.08215479850769,
      "epoch": 5.791690709751875,
      "grad_norm": 0.953125,
      "learning_rate": 0.00022299454935843845,
      "loss": 5.4655,
      "mean_token_accuracy": 0.21458880603313446,
      "num_tokens": 127191523.0,
      "step": 50185
    },
    {
      "entropy": 6.062656497955322,
      "epoch": 5.792267743796884,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022295440325368243,
      "loss": 5.4392,
      "mean_token_accuracy": 0.2195281684398651,
      "num_tokens": 127205063.0,
      "step": 50190
    },
    {
      "entropy": 6.028362035751343,
      "epoch": 5.792844777841893,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00022291425889945426,
      "loss": 5.4781,
      "mean_token_accuracy": 0.22332883924245833,
      "num_tokens": 127219711.0,
      "step": 50195
    },
    {
      "entropy": 6.000434350967407,
      "epoch": 5.793421811886901,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022287411629710408,
      "loss": 5.3932,
      "mean_token_accuracy": 0.22552696168422698,
      "num_tokens": 127231466.0,
      "step": 50200
    },
    {
      "entropy": 6.020996618270874,
      "epoch": 5.79399884593191,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00022283397544798223,
      "loss": 5.3763,
      "mean_token_accuracy": 0.23099322468042374,
      "num_tokens": 127243936.0,
      "step": 50205
    },
    {
      "entropy": 6.043927240371704,
      "epoch": 5.794575879976919,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022279383635343854,
      "loss": 5.4125,
      "mean_token_accuracy": 0.21853159070014955,
      "num_tokens": 127256444.0,
      "step": 50210
    },
    {
      "entropy": 5.9943962574005125,
      "epoch": 5.795152914021927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022275369901482333,
      "loss": 5.4111,
      "mean_token_accuracy": 0.22530001699924468,
      "num_tokens": 127269298.0,
      "step": 50215
    },
    {
      "entropy": 5.988767194747925,
      "epoch": 5.795729948066936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00022271356343348642,
      "loss": 5.3522,
      "mean_token_accuracy": 0.22620439231395723,
      "num_tokens": 127281061.0,
      "step": 50220
    },
    {
      "entropy": 6.030716753005981,
      "epoch": 5.796306982111945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022267342961077797,
      "loss": 5.445,
      "mean_token_accuracy": 0.2158772960305214,
      "num_tokens": 127294217.0,
      "step": 50225
    },
    {
      "entropy": 6.027329826354981,
      "epoch": 5.7968840161569535,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002226332975480476,
      "loss": 5.4264,
      "mean_token_accuracy": 0.22051745802164077,
      "num_tokens": 127306463.0,
      "step": 50230
    },
    {
      "entropy": 5.963424158096314,
      "epoch": 5.797461050201962,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002225931672466453,
      "loss": 5.3329,
      "mean_token_accuracy": 0.2327040657401085,
      "num_tokens": 127319625.0,
      "step": 50235
    },
    {
      "entropy": 5.9281950950622555,
      "epoch": 5.79803808424697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022255303870792077,
      "loss": 5.3906,
      "mean_token_accuracy": 0.23432537764310837,
      "num_tokens": 127333735.0,
      "step": 50240
    },
    {
      "entropy": 6.138464403152466,
      "epoch": 5.798615118291979,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022251291193322377,
      "loss": 5.5601,
      "mean_token_accuracy": 0.2133669897913933,
      "num_tokens": 127346043.0,
      "step": 50245
    },
    {
      "entropy": 6.015682935714722,
      "epoch": 5.799192152336988,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022247278692390387,
      "loss": 5.3946,
      "mean_token_accuracy": 0.22676908671855928,
      "num_tokens": 127358225.0,
      "step": 50250
    },
    {
      "entropy": 6.0067120552062985,
      "epoch": 5.7997691863819965,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002224326636813107,
      "loss": 5.3755,
      "mean_token_accuracy": 0.22662544697523118,
      "num_tokens": 127370924.0,
      "step": 50255
    },
    {
      "entropy": 6.067869329452515,
      "epoch": 5.800346220427005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022239254220679373,
      "loss": 5.4149,
      "mean_token_accuracy": 0.22250195890665053,
      "num_tokens": 127382060.0,
      "step": 50260
    },
    {
      "entropy": 6.01682596206665,
      "epoch": 5.800923254472014,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022235242250170256,
      "loss": 5.4209,
      "mean_token_accuracy": 0.21754756420850754,
      "num_tokens": 127394559.0,
      "step": 50265
    },
    {
      "entropy": 5.899926567077637,
      "epoch": 5.801500288517023,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022231230456738643,
      "loss": 5.2785,
      "mean_token_accuracy": 0.23758705109357833,
      "num_tokens": 127407072.0,
      "step": 50270
    },
    {
      "entropy": 6.061111736297607,
      "epoch": 5.802077322562031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002222721884051948,
      "loss": 5.5075,
      "mean_token_accuracy": 0.2173176422715187,
      "num_tokens": 127420364.0,
      "step": 50275
    },
    {
      "entropy": 6.086686658859253,
      "epoch": 5.802654356607039,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002222320740164769,
      "loss": 5.5006,
      "mean_token_accuracy": 0.21597716808319092,
      "num_tokens": 127433569.0,
      "step": 50280
    },
    {
      "entropy": 6.040704202651978,
      "epoch": 5.803231390652049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022219196140258203,
      "loss": 5.4181,
      "mean_token_accuracy": 0.22443622797727586,
      "num_tokens": 127445734.0,
      "step": 50285
    },
    {
      "entropy": 5.977840328216553,
      "epoch": 5.803808424697057,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022215185056485926,
      "loss": 5.3559,
      "mean_token_accuracy": 0.23254020214080812,
      "num_tokens": 127458553.0,
      "step": 50290
    },
    {
      "entropy": 6.00077977180481,
      "epoch": 5.804385458742066,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002221117415046578,
      "loss": 5.3898,
      "mean_token_accuracy": 0.2300723522901535,
      "num_tokens": 127470686.0,
      "step": 50295
    },
    {
      "entropy": 6.035689735412598,
      "epoch": 5.804962492787075,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022207163422332653,
      "loss": 5.4023,
      "mean_token_accuracy": 0.22128079682588578,
      "num_tokens": 127482867.0,
      "step": 50300
    },
    {
      "entropy": 5.999207162857056,
      "epoch": 5.805539526832083,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022203152872221467,
      "loss": 5.3665,
      "mean_token_accuracy": 0.225989992916584,
      "num_tokens": 127496237.0,
      "step": 50305
    },
    {
      "entropy": 5.940468645095825,
      "epoch": 5.806116560877092,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022199142500267094,
      "loss": 5.3473,
      "mean_token_accuracy": 0.23692095130681992,
      "num_tokens": 127509833.0,
      "step": 50310
    },
    {
      "entropy": 6.028581428527832,
      "epoch": 5.8066935949221,
      "grad_norm": 1.140625,
      "learning_rate": 0.00022195132306604437,
      "loss": 5.3839,
      "mean_token_accuracy": 0.22225600779056548,
      "num_tokens": 127522967.0,
      "step": 50315
    },
    {
      "entropy": 6.022025108337402,
      "epoch": 5.807270628967109,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022191122291368364,
      "loss": 5.414,
      "mean_token_accuracy": 0.22748849987983705,
      "num_tokens": 127535466.0,
      "step": 50320
    },
    {
      "entropy": 6.047796106338501,
      "epoch": 5.807847663012118,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022187112454693765,
      "loss": 5.419,
      "mean_token_accuracy": 0.2235051855444908,
      "num_tokens": 127548068.0,
      "step": 50325
    },
    {
      "entropy": 5.996612882614135,
      "epoch": 5.808424697057126,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00022183102796715493,
      "loss": 5.3234,
      "mean_token_accuracy": 0.23644556999206542,
      "num_tokens": 127560648.0,
      "step": 50330
    },
    {
      "entropy": 5.920382738113403,
      "epoch": 5.809001731102135,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022179093317568416,
      "loss": 5.3199,
      "mean_token_accuracy": 0.22743216902017593,
      "num_tokens": 127573034.0,
      "step": 50335
    },
    {
      "entropy": 6.035889625549316,
      "epoch": 5.809578765147144,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022175084017387394,
      "loss": 5.4615,
      "mean_token_accuracy": 0.21759170442819595,
      "num_tokens": 127585377.0,
      "step": 50340
    },
    {
      "entropy": 6.012312126159668,
      "epoch": 5.810155799192152,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002217107489630728,
      "loss": 5.3887,
      "mean_token_accuracy": 0.22402003407478333,
      "num_tokens": 127597593.0,
      "step": 50345
    },
    {
      "entropy": 6.122308397293091,
      "epoch": 5.810732833237161,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022167065954462907,
      "loss": 5.4703,
      "mean_token_accuracy": 0.2186146304011345,
      "num_tokens": 127611891.0,
      "step": 50350
    },
    {
      "entropy": 6.0643054962158205,
      "epoch": 5.811309867282169,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022163057191989128,
      "loss": 5.3967,
      "mean_token_accuracy": 0.21922862082719802,
      "num_tokens": 127624512.0,
      "step": 50355
    },
    {
      "entropy": 5.853642177581787,
      "epoch": 5.811886901327179,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002215904860902076,
      "loss": 5.2042,
      "mean_token_accuracy": 0.24209549725055696,
      "num_tokens": 127636857.0,
      "step": 50360
    },
    {
      "entropy": 6.0192639350891115,
      "epoch": 5.812463935372187,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002215504020569265,
      "loss": 5.4108,
      "mean_token_accuracy": 0.22073237746953964,
      "num_tokens": 127649577.0,
      "step": 50365
    },
    {
      "entropy": 6.17706069946289,
      "epoch": 5.813040969417195,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022151031982139602,
      "loss": 5.5159,
      "mean_token_accuracy": 0.21245130896568298,
      "num_tokens": 127661876.0,
      "step": 50370
    },
    {
      "entropy": 5.946478843688965,
      "epoch": 5.813618003462205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002214702393849644,
      "loss": 5.3091,
      "mean_token_accuracy": 0.2342524379491806,
      "num_tokens": 127674786.0,
      "step": 50375
    },
    {
      "entropy": 6.025868988037109,
      "epoch": 5.814195037507213,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022143016074897964,
      "loss": 5.4269,
      "mean_token_accuracy": 0.22212120592594148,
      "num_tokens": 127687041.0,
      "step": 50380
    },
    {
      "entropy": 6.006273889541626,
      "epoch": 5.8147720715522215,
      "grad_norm": 1.09375,
      "learning_rate": 0.00022139008391478988,
      "loss": 5.4211,
      "mean_token_accuracy": 0.22161509841680527,
      "num_tokens": 127698178.0,
      "step": 50385
    },
    {
      "entropy": 6.0416015625,
      "epoch": 5.81534910559723,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022135000888374296,
      "loss": 5.4544,
      "mean_token_accuracy": 0.22033730447292327,
      "num_tokens": 127709422.0,
      "step": 50390
    },
    {
      "entropy": 6.064035367965698,
      "epoch": 5.815926139642239,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022130993565718687,
      "loss": 5.4691,
      "mean_token_accuracy": 0.21578666865825652,
      "num_tokens": 127721259.0,
      "step": 50395
    },
    {
      "entropy": 6.102241563796997,
      "epoch": 5.816503173687248,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022126986423646937,
      "loss": 5.474,
      "mean_token_accuracy": 0.221116741001606,
      "num_tokens": 127733537.0,
      "step": 50400
    },
    {
      "entropy": 6.047968816757202,
      "epoch": 5.817080207732256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00022122979462293842,
      "loss": 5.4421,
      "mean_token_accuracy": 0.22333806604146958,
      "num_tokens": 127745735.0,
      "step": 50405
    },
    {
      "entropy": 6.080864334106446,
      "epoch": 5.8176572417772645,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00022118972681794152,
      "loss": 5.3961,
      "mean_token_accuracy": 0.22876183539628983,
      "num_tokens": 127758406.0,
      "step": 50410
    },
    {
      "entropy": 6.022109603881836,
      "epoch": 5.818234275822274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00022114966082282644,
      "loss": 5.4241,
      "mean_token_accuracy": 0.22243682593107222,
      "num_tokens": 127770049.0,
      "step": 50415
    },
    {
      "entropy": 5.9110535144805905,
      "epoch": 5.818811309867282,
      "grad_norm": 0.96875,
      "learning_rate": 0.00022110959663894077,
      "loss": 5.3702,
      "mean_token_accuracy": 0.23012958467006683,
      "num_tokens": 127784114.0,
      "step": 50420
    },
    {
      "entropy": 6.095863962173462,
      "epoch": 5.819388343912291,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002210695342676321,
      "loss": 5.4865,
      "mean_token_accuracy": 0.21268316209316254,
      "num_tokens": 127797139.0,
      "step": 50425
    },
    {
      "entropy": 6.03691668510437,
      "epoch": 5.819965377957299,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002210294737102478,
      "loss": 5.4564,
      "mean_token_accuracy": 0.22031344920396806,
      "num_tokens": 127809314.0,
      "step": 50430
    },
    {
      "entropy": 5.9681525230407715,
      "epoch": 5.820542412002308,
      "grad_norm": 0.96875,
      "learning_rate": 0.00022098941496813534,
      "loss": 5.3793,
      "mean_token_accuracy": 0.23264460414648055,
      "num_tokens": 127822607.0,
      "step": 50435
    },
    {
      "entropy": 6.0978330135345455,
      "epoch": 5.821119446047317,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022094935804264205,
      "loss": 5.4825,
      "mean_token_accuracy": 0.21497157961130142,
      "num_tokens": 127835428.0,
      "step": 50440
    },
    {
      "entropy": 6.140019083023072,
      "epoch": 5.821696480092325,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002209093029351154,
      "loss": 5.4736,
      "mean_token_accuracy": 0.21492733657360077,
      "num_tokens": 127847351.0,
      "step": 50445
    },
    {
      "entropy": 6.074713563919067,
      "epoch": 5.8222735141373345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002208692496469023,
      "loss": 5.462,
      "mean_token_accuracy": 0.2184465542435646,
      "num_tokens": 127859843.0,
      "step": 50450
    },
    {
      "entropy": 6.044490242004395,
      "epoch": 5.822850548182343,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002208291981793502,
      "loss": 5.363,
      "mean_token_accuracy": 0.22210218012332916,
      "num_tokens": 127871953.0,
      "step": 50455
    },
    {
      "entropy": 6.000745153427124,
      "epoch": 5.823427582227351,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022078914853380604,
      "loss": 5.4085,
      "mean_token_accuracy": 0.23046095222234725,
      "num_tokens": 127885162.0,
      "step": 50460
    },
    {
      "entropy": 5.9967468738555905,
      "epoch": 5.82400461627236,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002207491007116171,
      "loss": 5.4032,
      "mean_token_accuracy": 0.23266995549201966,
      "num_tokens": 127897425.0,
      "step": 50465
    },
    {
      "entropy": 5.904079389572144,
      "epoch": 5.824581650317369,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022070905471413004,
      "loss": 5.3529,
      "mean_token_accuracy": 0.2272264376282692,
      "num_tokens": 127911088.0,
      "step": 50470
    },
    {
      "entropy": 6.0410847663879395,
      "epoch": 5.8251586843623775,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022066901054269207,
      "loss": 5.4237,
      "mean_token_accuracy": 0.22082597762346268,
      "num_tokens": 127923359.0,
      "step": 50475
    },
    {
      "entropy": 5.991813039779663,
      "epoch": 5.825735718407386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002206289681986498,
      "loss": 5.3776,
      "mean_token_accuracy": 0.2273949921131134,
      "num_tokens": 127935450.0,
      "step": 50480
    },
    {
      "entropy": 6.027788400650024,
      "epoch": 5.826312752452394,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00022058892768335038,
      "loss": 5.4406,
      "mean_token_accuracy": 0.21817007511854172,
      "num_tokens": 127948436.0,
      "step": 50485
    },
    {
      "entropy": 5.9902318000793455,
      "epoch": 5.826889786497404,
      "grad_norm": 1.0,
      "learning_rate": 0.00022054888899814012,
      "loss": 5.3533,
      "mean_token_accuracy": 0.22581617385149003,
      "num_tokens": 127961639.0,
      "step": 50490
    },
    {
      "entropy": 6.03393588066101,
      "epoch": 5.827466820542412,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022050885214436618,
      "loss": 5.4297,
      "mean_token_accuracy": 0.22351187020540236,
      "num_tokens": 127974376.0,
      "step": 50495
    },
    {
      "entropy": 6.050661420822143,
      "epoch": 5.8280438545874205,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022046881712337475,
      "loss": 5.4378,
      "mean_token_accuracy": 0.2202516421675682,
      "num_tokens": 127987221.0,
      "step": 50500
    },
    {
      "entropy": 6.01173095703125,
      "epoch": 5.828620888632429,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002204287839365126,
      "loss": 5.3522,
      "mean_token_accuracy": 0.2260069102048874,
      "num_tokens": 127999512.0,
      "step": 50505
    },
    {
      "entropy": 6.0701940059661865,
      "epoch": 5.829197922677438,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002203887525851262,
      "loss": 5.4425,
      "mean_token_accuracy": 0.22250343710184098,
      "num_tokens": 128013408.0,
      "step": 50510
    },
    {
      "entropy": 6.071457099914551,
      "epoch": 5.829774956722447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00022034872307056204,
      "loss": 5.4146,
      "mean_token_accuracy": 0.22135910838842393,
      "num_tokens": 128025839.0,
      "step": 50515
    },
    {
      "entropy": 5.995633602142334,
      "epoch": 5.830351990767455,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00022030869539416638,
      "loss": 5.3519,
      "mean_token_accuracy": 0.2197800561785698,
      "num_tokens": 128039935.0,
      "step": 50520
    },
    {
      "entropy": 6.043131399154663,
      "epoch": 5.830929024812464,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002202686695572856,
      "loss": 5.4312,
      "mean_token_accuracy": 0.2271645188331604,
      "num_tokens": 128052297.0,
      "step": 50525
    },
    {
      "entropy": 6.061482191085815,
      "epoch": 5.831506058857473,
      "grad_norm": 1.078125,
      "learning_rate": 0.00022022864556126587,
      "loss": 5.4615,
      "mean_token_accuracy": 0.2227476179599762,
      "num_tokens": 128064393.0,
      "step": 50530
    },
    {
      "entropy": 5.991846370697021,
      "epoch": 5.832083092902481,
      "grad_norm": 1.015625,
      "learning_rate": 0.00022018862340745355,
      "loss": 5.3873,
      "mean_token_accuracy": 0.2262405723333359,
      "num_tokens": 128077514.0,
      "step": 50535
    },
    {
      "entropy": 6.089515829086304,
      "epoch": 5.83266012694749,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002201486030971946,
      "loss": 5.4218,
      "mean_token_accuracy": 0.2243250235915184,
      "num_tokens": 128089834.0,
      "step": 50540
    },
    {
      "entropy": 6.050447463989258,
      "epoch": 5.833237160992499,
      "grad_norm": 0.96875,
      "learning_rate": 0.00022010858463183508,
      "loss": 5.4456,
      "mean_token_accuracy": 0.22340308278799056,
      "num_tokens": 128103083.0,
      "step": 50545
    },
    {
      "entropy": 6.014553499221802,
      "epoch": 5.833814195037507,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00022006856801272114,
      "loss": 5.3639,
      "mean_token_accuracy": 0.22425753474235535,
      "num_tokens": 128115786.0,
      "step": 50550
    },
    {
      "entropy": 5.9598688125610355,
      "epoch": 5.834391229082516,
      "grad_norm": 1.03125,
      "learning_rate": 0.00022002855324119853,
      "loss": 5.3316,
      "mean_token_accuracy": 0.23018798232078552,
      "num_tokens": 128127485.0,
      "step": 50555
    },
    {
      "entropy": 6.038669490814209,
      "epoch": 5.834968263127524,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021998854031861332,
      "loss": 5.3826,
      "mean_token_accuracy": 0.22737166732549668,
      "num_tokens": 128140282.0,
      "step": 50560
    },
    {
      "entropy": 6.0039464950561525,
      "epoch": 5.8355452971725335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002199485292463112,
      "loss": 5.3891,
      "mean_token_accuracy": 0.23098409175872803,
      "num_tokens": 128153293.0,
      "step": 50565
    },
    {
      "entropy": 6.037856101989746,
      "epoch": 5.836122331217542,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021990852002563794,
      "loss": 5.4717,
      "mean_token_accuracy": 0.21816712617874146,
      "num_tokens": 128166087.0,
      "step": 50570
    },
    {
      "entropy": 6.042622089385986,
      "epoch": 5.83669936526255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021986851265793923,
      "loss": 5.4435,
      "mean_token_accuracy": 0.22034710347652436,
      "num_tokens": 128178642.0,
      "step": 50575
    },
    {
      "entropy": 5.988347959518433,
      "epoch": 5.837276399307559,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002198285071445608,
      "loss": 5.4234,
      "mean_token_accuracy": 0.2197228953242302,
      "num_tokens": 128190713.0,
      "step": 50580
    },
    {
      "entropy": 6.056886100769043,
      "epoch": 5.837853433352568,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021978850348684803,
      "loss": 5.3601,
      "mean_token_accuracy": 0.22639572471380234,
      "num_tokens": 128202030.0,
      "step": 50585
    },
    {
      "entropy": 6.025046253204346,
      "epoch": 5.838430467397576,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021974850168614656,
      "loss": 5.4135,
      "mean_token_accuracy": 0.21961946338415145,
      "num_tokens": 128214240.0,
      "step": 50590
    },
    {
      "entropy": 5.98553318977356,
      "epoch": 5.839007501442585,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00021970850174380174,
      "loss": 5.4151,
      "mean_token_accuracy": 0.21660540848970414,
      "num_tokens": 128227733.0,
      "step": 50595
    },
    {
      "entropy": 5.945986986160278,
      "epoch": 5.839584535487594,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021966850366115915,
      "loss": 5.3422,
      "mean_token_accuracy": 0.22543343305587768,
      "num_tokens": 128239964.0,
      "step": 50600
    },
    {
      "entropy": 6.061747360229492,
      "epoch": 5.840161569532603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00021962850743956382,
      "loss": 5.4699,
      "mean_token_accuracy": 0.21340246498584747,
      "num_tokens": 128252647.0,
      "step": 50605
    },
    {
      "entropy": 6.086827564239502,
      "epoch": 5.840738603577611,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00021958851308036116,
      "loss": 5.4675,
      "mean_token_accuracy": 0.21760924607515336,
      "num_tokens": 128264876.0,
      "step": 50610
    },
    {
      "entropy": 6.120539569854737,
      "epoch": 5.841315637622619,
      "grad_norm": 0.984375,
      "learning_rate": 0.00021954852058489634,
      "loss": 5.4813,
      "mean_token_accuracy": 0.2170608103275299,
      "num_tokens": 128278572.0,
      "step": 50615
    },
    {
      "entropy": 5.974930477142334,
      "epoch": 5.841892671667629,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021950852995451453,
      "loss": 5.3533,
      "mean_token_accuracy": 0.22724754810333253,
      "num_tokens": 128291052.0,
      "step": 50620
    },
    {
      "entropy": 6.099519968032837,
      "epoch": 5.842469705712637,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002194685411905607,
      "loss": 5.4917,
      "mean_token_accuracy": 0.21947124749422073,
      "num_tokens": 128303582.0,
      "step": 50625
    },
    {
      "entropy": 6.02627272605896,
      "epoch": 5.8430467397576455,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002194285542943799,
      "loss": 5.4156,
      "mean_token_accuracy": 0.22575947195291518,
      "num_tokens": 128317367.0,
      "step": 50630
    },
    {
      "entropy": 6.07182240486145,
      "epoch": 5.843623773802655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021938856926731704,
      "loss": 5.5115,
      "mean_token_accuracy": 0.2182227909564972,
      "num_tokens": 128329652.0,
      "step": 50635
    },
    {
      "entropy": 5.958067417144775,
      "epoch": 5.844200807847663,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021934858611071719,
      "loss": 5.3523,
      "mean_token_accuracy": 0.2276764467358589,
      "num_tokens": 128341538.0,
      "step": 50640
    },
    {
      "entropy": 6.075793695449829,
      "epoch": 5.844777841892672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021930860482592484,
      "loss": 5.4548,
      "mean_token_accuracy": 0.21957226991653442,
      "num_tokens": 128353533.0,
      "step": 50645
    },
    {
      "entropy": 6.091213226318359,
      "epoch": 5.84535487593768,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000219268625414285,
      "loss": 5.4527,
      "mean_token_accuracy": 0.2202288642525673,
      "num_tokens": 128367028.0,
      "step": 50650
    },
    {
      "entropy": 5.9958878517150875,
      "epoch": 5.8459319099826885,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021922864787714215,
      "loss": 5.4744,
      "mean_token_accuracy": 0.21656205207109452,
      "num_tokens": 128378759.0,
      "step": 50655
    },
    {
      "entropy": 6.097091245651245,
      "epoch": 5.846508944027698,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002191886722158411,
      "loss": 5.4662,
      "mean_token_accuracy": 0.21781108379364014,
      "num_tokens": 128390689.0,
      "step": 50660
    },
    {
      "entropy": 6.130036067962647,
      "epoch": 5.847085978072706,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002191486984317263,
      "loss": 5.4266,
      "mean_token_accuracy": 0.21853924095630645,
      "num_tokens": 128402774.0,
      "step": 50665
    },
    {
      "entropy": 6.017915105819702,
      "epoch": 5.847663012117715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021910872652614234,
      "loss": 5.4131,
      "mean_token_accuracy": 0.22539464086294175,
      "num_tokens": 128416237.0,
      "step": 50670
    },
    {
      "entropy": 6.010024499893189,
      "epoch": 5.848240046162724,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021906875650043356,
      "loss": 5.3807,
      "mean_token_accuracy": 0.23103055208921433,
      "num_tokens": 128428298.0,
      "step": 50675
    },
    {
      "entropy": 6.102875566482544,
      "epoch": 5.848817080207732,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021902878835594443,
      "loss": 5.376,
      "mean_token_accuracy": 0.22378448098897935,
      "num_tokens": 128439925.0,
      "step": 50680
    },
    {
      "entropy": 5.91007432937622,
      "epoch": 5.849394114252741,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021898882209401915,
      "loss": 5.2412,
      "mean_token_accuracy": 0.23623914569616317,
      "num_tokens": 128453908.0,
      "step": 50685
    },
    {
      "entropy": 6.054044008255005,
      "epoch": 5.849971148297749,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021894885771600204,
      "loss": 5.4946,
      "mean_token_accuracy": 0.21876438558101655,
      "num_tokens": 128466695.0,
      "step": 50690
    },
    {
      "entropy": 6.035588026046753,
      "epoch": 5.8505481823427585,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002189088952232372,
      "loss": 5.3946,
      "mean_token_accuracy": 0.22270541936159133,
      "num_tokens": 128479312.0,
      "step": 50695
    },
    {
      "entropy": 5.870685958862305,
      "epoch": 5.851125216387767,
      "grad_norm": 0.84375,
      "learning_rate": 0.00021886893461706892,
      "loss": 5.2587,
      "mean_token_accuracy": 0.23716987520456315,
      "num_tokens": 128491469.0,
      "step": 50700
    },
    {
      "entropy": 5.998844146728516,
      "epoch": 5.851702250432775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000218828975898841,
      "loss": 5.4141,
      "mean_token_accuracy": 0.2196686178445816,
      "num_tokens": 128503059.0,
      "step": 50705
    },
    {
      "entropy": 6.131736183166504,
      "epoch": 5.852279284477785,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021878901906989767,
      "loss": 5.5001,
      "mean_token_accuracy": 0.2119828701019287,
      "num_tokens": 128517689.0,
      "step": 50710
    },
    {
      "entropy": 6.022342777252197,
      "epoch": 5.852856318522793,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021874906413158268,
      "loss": 5.3884,
      "mean_token_accuracy": 0.23145127147436143,
      "num_tokens": 128530240.0,
      "step": 50715
    },
    {
      "entropy": 6.052708959579467,
      "epoch": 5.8534333525678015,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002187091110852401,
      "loss": 5.4694,
      "mean_token_accuracy": 0.2142168790102005,
      "num_tokens": 128542832.0,
      "step": 50720
    },
    {
      "entropy": 6.0598667621612545,
      "epoch": 5.85401038661281,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00021866915993221353,
      "loss": 5.4693,
      "mean_token_accuracy": 0.21437126845121385,
      "num_tokens": 128556304.0,
      "step": 50725
    },
    {
      "entropy": 6.003561687469483,
      "epoch": 5.854587420657818,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021862921067384683,
      "loss": 5.3894,
      "mean_token_accuracy": 0.23023393005132675,
      "num_tokens": 128569192.0,
      "step": 50730
    },
    {
      "entropy": 6.012569570541382,
      "epoch": 5.855164454702828,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00021858926331148356,
      "loss": 5.4147,
      "mean_token_accuracy": 0.22599702328443527,
      "num_tokens": 128582767.0,
      "step": 50735
    },
    {
      "entropy": 6.026070165634155,
      "epoch": 5.855741488747836,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00021854931784646747,
      "loss": 5.408,
      "mean_token_accuracy": 0.22527667880058289,
      "num_tokens": 128596200.0,
      "step": 50740
    },
    {
      "entropy": 6.038392448425293,
      "epoch": 5.8563185227928445,
      "grad_norm": 1.0,
      "learning_rate": 0.00021850937428014196,
      "loss": 5.4279,
      "mean_token_accuracy": 0.22600363045930863,
      "num_tokens": 128610298.0,
      "step": 50745
    },
    {
      "entropy": 6.003039407730102,
      "epoch": 5.856895556837854,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021846943261385067,
      "loss": 5.3744,
      "mean_token_accuracy": 0.2263938993215561,
      "num_tokens": 128622274.0,
      "step": 50750
    },
    {
      "entropy": 6.075804996490478,
      "epoch": 5.857472590882862,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002184294928489368,
      "loss": 5.4372,
      "mean_token_accuracy": 0.2138909727334976,
      "num_tokens": 128634772.0,
      "step": 50755
    },
    {
      "entropy": 6.0322585105896,
      "epoch": 5.858049624927871,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021838955498674402,
      "loss": 5.4585,
      "mean_token_accuracy": 0.22602907866239547,
      "num_tokens": 128647283.0,
      "step": 50760
    },
    {
      "entropy": 5.969331216812134,
      "epoch": 5.858626658972879,
      "grad_norm": 1.0,
      "learning_rate": 0.00021834961902861527,
      "loss": 5.3436,
      "mean_token_accuracy": 0.23066012412309647,
      "num_tokens": 128659277.0,
      "step": 50765
    },
    {
      "entropy": 6.022809076309204,
      "epoch": 5.859203693017888,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021830968497589404,
      "loss": 5.3272,
      "mean_token_accuracy": 0.23730990290641785,
      "num_tokens": 128670499.0,
      "step": 50770
    },
    {
      "entropy": 5.933892345428466,
      "epoch": 5.859780727062897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021826975282992334,
      "loss": 5.3352,
      "mean_token_accuracy": 0.22655107378959655,
      "num_tokens": 128683153.0,
      "step": 50775
    },
    {
      "entropy": 5.933896589279175,
      "epoch": 5.860357761107905,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021822982259204644,
      "loss": 5.3717,
      "mean_token_accuracy": 0.22985600233078002,
      "num_tokens": 128696067.0,
      "step": 50780
    },
    {
      "entropy": 6.115619468688965,
      "epoch": 5.8609347951529145,
      "grad_norm": 1.015625,
      "learning_rate": 0.00021818989426360613,
      "loss": 5.5854,
      "mean_token_accuracy": 0.21275230199098588,
      "num_tokens": 128709220.0,
      "step": 50785
    },
    {
      "entropy": 6.144386529922485,
      "epoch": 5.861511829197923,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00021814996784594553,
      "loss": 5.5551,
      "mean_token_accuracy": 0.2121775820851326,
      "num_tokens": 128722363.0,
      "step": 50790
    },
    {
      "entropy": 6.10267071723938,
      "epoch": 5.862088863242931,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021811004334040753,
      "loss": 5.4615,
      "mean_token_accuracy": 0.2197597399353981,
      "num_tokens": 128734413.0,
      "step": 50795
    },
    {
      "entropy": 6.015191650390625,
      "epoch": 5.86266589728794,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00021807012074833505,
      "loss": 5.3643,
      "mean_token_accuracy": 0.22646291553974152,
      "num_tokens": 128746333.0,
      "step": 50800
    },
    {
      "entropy": 5.973905420303344,
      "epoch": 5.863242931332948,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021803020007107066,
      "loss": 5.3981,
      "mean_token_accuracy": 0.22852888852357864,
      "num_tokens": 128758889.0,
      "step": 50805
    },
    {
      "entropy": 6.01613335609436,
      "epoch": 5.8638199653779575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021799028130995723,
      "loss": 5.3557,
      "mean_token_accuracy": 0.23281191736459733,
      "num_tokens": 128771089.0,
      "step": 50810
    },
    {
      "entropy": 6.040916776657104,
      "epoch": 5.864396999422966,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021795036446633736,
      "loss": 5.3886,
      "mean_token_accuracy": 0.22952151596546172,
      "num_tokens": 128784590.0,
      "step": 50815
    },
    {
      "entropy": 6.0145426273345945,
      "epoch": 5.864974033467974,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00021791044954155375,
      "loss": 5.4272,
      "mean_token_accuracy": 0.21972869485616683,
      "num_tokens": 128798415.0,
      "step": 50820
    },
    {
      "entropy": 5.94070954322815,
      "epoch": 5.865551067512984,
      "grad_norm": 0.96875,
      "learning_rate": 0.00021787053653694866,
      "loss": 5.3086,
      "mean_token_accuracy": 0.2368646591901779,
      "num_tokens": 128812405.0,
      "step": 50825
    },
    {
      "entropy": 6.076986265182495,
      "epoch": 5.866128101557992,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00021783062545386477,
      "loss": 5.4708,
      "mean_token_accuracy": 0.21962153613567353,
      "num_tokens": 128824665.0,
      "step": 50830
    },
    {
      "entropy": 6.005318927764892,
      "epoch": 5.866705135603,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002177907162936444,
      "loss": 5.397,
      "mean_token_accuracy": 0.22843997776508332,
      "num_tokens": 128837561.0,
      "step": 50835
    },
    {
      "entropy": 6.0911918640136715,
      "epoch": 5.867282169648009,
      "grad_norm": 1.0,
      "learning_rate": 0.0002177508090576299,
      "loss": 5.4723,
      "mean_token_accuracy": 0.21704014986753464,
      "num_tokens": 128849954.0,
      "step": 50840
    },
    {
      "entropy": 6.0029552459716795,
      "epoch": 5.867859203693018,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002177109037471634,
      "loss": 5.3657,
      "mean_token_accuracy": 0.2323867380619049,
      "num_tokens": 128864258.0,
      "step": 50845
    },
    {
      "entropy": 5.99102520942688,
      "epoch": 5.868436237738027,
      "grad_norm": 1.0,
      "learning_rate": 0.0002176710003635873,
      "loss": 5.3429,
      "mean_token_accuracy": 0.23248618543148042,
      "num_tokens": 128877708.0,
      "step": 50850
    },
    {
      "entropy": 6.038684701919555,
      "epoch": 5.869013271783035,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002176310989082435,
      "loss": 5.374,
      "mean_token_accuracy": 0.22527575343847275,
      "num_tokens": 128890039.0,
      "step": 50855
    },
    {
      "entropy": 6.018736982345581,
      "epoch": 5.869590305828044,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021759119938247434,
      "loss": 5.343,
      "mean_token_accuracy": 0.22151255905628203,
      "num_tokens": 128901873.0,
      "step": 50860
    },
    {
      "entropy": 6.012150239944458,
      "epoch": 5.870167339873053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00021755130178762155,
      "loss": 5.4482,
      "mean_token_accuracy": 0.22416732013225554,
      "num_tokens": 128915151.0,
      "step": 50865
    },
    {
      "entropy": 6.021463060379029,
      "epoch": 5.870744373918061,
      "grad_norm": 1.078125,
      "learning_rate": 0.00021751140612502734,
      "loss": 5.3704,
      "mean_token_accuracy": 0.22769350409507752,
      "num_tokens": 128927948.0,
      "step": 50870
    },
    {
      "entropy": 5.9567488670349125,
      "epoch": 5.87132140796307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021747151239603324,
      "loss": 5.3884,
      "mean_token_accuracy": 0.22673175632953643,
      "num_tokens": 128940564.0,
      "step": 50875
    },
    {
      "entropy": 6.0511589527130125,
      "epoch": 5.871898442008078,
      "grad_norm": 1.03125,
      "learning_rate": 0.00021743162060198134,
      "loss": 5.3979,
      "mean_token_accuracy": 0.23280119150877,
      "num_tokens": 128954579.0,
      "step": 50880
    },
    {
      "entropy": 5.927225542068482,
      "epoch": 5.872475476053087,
      "grad_norm": 1.125,
      "learning_rate": 0.00021739173074421331,
      "loss": 5.2451,
      "mean_token_accuracy": 0.24241523295640946,
      "num_tokens": 128967706.0,
      "step": 50885
    },
    {
      "entropy": 6.064105987548828,
      "epoch": 5.873052510098096,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021735184282407077,
      "loss": 5.452,
      "mean_token_accuracy": 0.22285414934158326,
      "num_tokens": 128980416.0,
      "step": 50890
    },
    {
      "entropy": 6.044906520843506,
      "epoch": 5.873629544143104,
      "grad_norm": 1.0,
      "learning_rate": 0.00021731195684289534,
      "loss": 5.4474,
      "mean_token_accuracy": 0.22325989454984665,
      "num_tokens": 128993765.0,
      "step": 50895
    },
    {
      "entropy": 6.016779899597168,
      "epoch": 5.874206578188113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021727207280202866,
      "loss": 5.3659,
      "mean_token_accuracy": 0.22246128171682358,
      "num_tokens": 129005562.0,
      "step": 50900
    },
    {
      "entropy": 5.930756664276123,
      "epoch": 5.874783612233122,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00021723219070281203,
      "loss": 5.3151,
      "mean_token_accuracy": 0.2313905492424965,
      "num_tokens": 129017836.0,
      "step": 50905
    },
    {
      "entropy": 6.047598266601563,
      "epoch": 5.87536064627813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00021719231054658707,
      "loss": 5.4383,
      "mean_token_accuracy": 0.22508041560649872,
      "num_tokens": 129030881.0,
      "step": 50910
    },
    {
      "entropy": 6.0395432472229,
      "epoch": 5.875937680323139,
      "grad_norm": 1.0,
      "learning_rate": 0.00021715243233469495,
      "loss": 5.4367,
      "mean_token_accuracy": 0.22594944685697554,
      "num_tokens": 129043415.0,
      "step": 50915
    },
    {
      "entropy": 6.013824510574341,
      "epoch": 5.876514714368148,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021711255606847706,
      "loss": 5.3929,
      "mean_token_accuracy": 0.22822608351707457,
      "num_tokens": 129057097.0,
      "step": 50920
    },
    {
      "entropy": 5.946033668518067,
      "epoch": 5.877091748413156,
      "grad_norm": 1.09375,
      "learning_rate": 0.00021707268174927459,
      "loss": 5.3636,
      "mean_token_accuracy": 0.22646875828504562,
      "num_tokens": 129072562.0,
      "step": 50925
    },
    {
      "entropy": 5.958529043197632,
      "epoch": 5.877668782458165,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00021703280937842879,
      "loss": 5.3135,
      "mean_token_accuracy": 0.23903025984764098,
      "num_tokens": 129086007.0,
      "step": 50930
    },
    {
      "entropy": 5.991726493835449,
      "epoch": 5.878245816503174,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002169929389572805,
      "loss": 5.4012,
      "mean_token_accuracy": 0.22207655608654023,
      "num_tokens": 129097903.0,
      "step": 50935
    },
    {
      "entropy": 6.144826602935791,
      "epoch": 5.8788228505481825,
      "grad_norm": 1.03125,
      "learning_rate": 0.000216953070487171,
      "loss": 5.572,
      "mean_token_accuracy": 0.21332115530967713,
      "num_tokens": 129111229.0,
      "step": 50940
    },
    {
      "entropy": 6.032774257659912,
      "epoch": 5.879399884593191,
      "grad_norm": 1.0625,
      "learning_rate": 0.00021691320396944103,
      "loss": 5.4081,
      "mean_token_accuracy": 0.22807783782482147,
      "num_tokens": 129123853.0,
      "step": 50945
    },
    {
      "entropy": 6.027386093139649,
      "epoch": 5.879976918638199,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021687333940543173,
      "loss": 5.4157,
      "mean_token_accuracy": 0.2205837145447731,
      "num_tokens": 129135717.0,
      "step": 50950
    },
    {
      "entropy": 6.111490917205811,
      "epoch": 5.880553952683209,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00021683347679648369,
      "loss": 5.4938,
      "mean_token_accuracy": 0.2160240516066551,
      "num_tokens": 129149733.0,
      "step": 50955
    },
    {
      "entropy": 6.070977592468262,
      "epoch": 5.881130986728217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00021679361614393777,
      "loss": 5.4071,
      "mean_token_accuracy": 0.22597229331731797,
      "num_tokens": 129161623.0,
      "step": 50960
    },
    {
      "entropy": 6.02405571937561,
      "epoch": 5.8817080207732255,
      "grad_norm": 1.0,
      "learning_rate": 0.00021675375744913467,
      "loss": 5.3948,
      "mean_token_accuracy": 0.22637299746274947,
      "num_tokens": 129174730.0,
      "step": 50965
    },
    {
      "entropy": 6.020820236206054,
      "epoch": 5.882285054818234,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000216713900713415,
      "loss": 5.3855,
      "mean_token_accuracy": 0.22477019876241683,
      "num_tokens": 129188961.0,
      "step": 50970
    },
    {
      "entropy": 5.972708034515381,
      "epoch": 5.882862088863243,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002166740459381194,
      "loss": 5.331,
      "mean_token_accuracy": 0.2267143651843071,
      "num_tokens": 129202263.0,
      "step": 50975
    },
    {
      "entropy": 6.039251661300659,
      "epoch": 5.883439122908252,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002166341931245882,
      "loss": 5.468,
      "mean_token_accuracy": 0.2169051706790924,
      "num_tokens": 129215631.0,
      "step": 50980
    },
    {
      "entropy": 6.010639667510986,
      "epoch": 5.88401615695326,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021659434227416197,
      "loss": 5.3566,
      "mean_token_accuracy": 0.22541877925395964,
      "num_tokens": 129228004.0,
      "step": 50985
    },
    {
      "entropy": 6.100912475585938,
      "epoch": 5.8845931909982685,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00021655449338818102,
      "loss": 5.4332,
      "mean_token_accuracy": 0.2221224844455719,
      "num_tokens": 129240185.0,
      "step": 50990
    },
    {
      "entropy": 5.853770303726196,
      "epoch": 5.885170225043278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021651464646798568,
      "loss": 5.2237,
      "mean_token_accuracy": 0.2406178742647171,
      "num_tokens": 129252524.0,
      "step": 50995
    },
    {
      "entropy": 6.060685539245606,
      "epoch": 5.885747259088286,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002164748015149161,
      "loss": 5.5067,
      "mean_token_accuracy": 0.21363474279642106,
      "num_tokens": 129265409.0,
      "step": 51000
    },
    {
      "epoch": 5.885747259088286,
      "eval_entropy": 5.845093757549673,
      "eval_loss": 5.596216678619385,
      "eval_mean_token_accuracy": 0.22051587693005553,
      "eval_num_tokens": 129265409.0,
      "eval_runtime": 22.9165,
      "eval_samples_per_second": 1227.762,
      "eval_steps_per_second": 153.47,
      "step": 51000
    }
  ],
  "logging_steps": 5,
  "max_steps": 86650,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9069394614272e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}